8 ETHICON Endo-Surgery [ MECHANICAL Stapling product 1./GACI.IP' 111doscoflic Cli1, Awlicrs ....s_, I'IWXlll!JATPLi11enr Stn1der, PROXlMATE' S/.:111 Stn14crs SAFETY · SECURITY · CONTROL .. S.E., lnc. ETHICON MECHANICAL/ ENDO • SURGERY ·FREDERICK D. HRKAC • THE ESPLANADE• Room 43/44 TONE LOVŠIN Mihanoviceva 1 · 41000 Zagreb · Croatia Dvoržakova 5 · 61000 Ljubljana · Slovenia 'B' 041/277340 · lax:041/277325 'B' 38/611322275·1ax:38/611326223 RADIOLOGY AND ONCOLOGY Established in 1964 as Radiologia Iugoslavica in Ljubljana, Slovenia. Radiology and Oncology is a journal devoted to publication of original contributions in diagnostic and interventional radiology, computerized tomography, ultrasound, magnetic resonance, nuclear medicine, radiotherapy, clinical and experimental oncology, radiophysics and radiation protection. Editor in chief Tomaž Benulic Ljubljana, Slovenia Associate editors Gregor Serša Ljubljana, Slovenia Viljem Kovac · Ljubljana, Slovenia Editorial board Marija Auersperg Andrija Hebrang Branko Palcic Ljubljana, Slovenia Zagreb, Croatia Vancouver, Canada !urica Papa Matija Bistrovic Durila Horvat Zagreb, Croatia Zagreb, Croatia Zagreb, Croatia Dušan Pavcnik Haris Boko Berta Jereb Ljubljana, Slovenia Zagreb, Croatia Ljubljana, Slovenia Stojan Plesnicar Malte Clausen Vladimir Jevtic Ljubljana, Slovenia Kiel, Germany Ljubljana, Slovenia Ervin B. Podgoršak Christoph Clemm H. Dieter Kogelnik Montreal, Canada Munchen, Germany Salzburg, Austria Miran Porenta Ljubljana, Slovenia Christian Dittric/1 Ivan Lovasic Jan C. Roos Vienna, Austria Rijeka, Croatia Amsterdam, The Netherlands Ivan Drinkovic Marijan Lovrencic Slavko Šimunic Zagreb, Croatia Zagreb, Croatia Zagreb, Croatia Luka Milas Bela Fornet Lojze Šmid Houston, USA Budapest, Hungary Ljubljana, Slovenia Tullio Giraldi Maja Osmak t4ndrea V eronesi Udine, Italy Zagreb, Croatia Gorizia, Italy Publishers Slovenian Medica[ Society -Section of Radio­ logy, Croatian Medica/ Association -Croatian So­ ciety of Radiology In collaboration with Societas Radiologorum Hungarorum Friuli-Venezia Giulia regional groups of S.I.R.M. (Italian Society of Medica[ Radiology.) Correspondence address Radiology and Oncology Institute of Oncology Vrazov trg 4 61000 Ljubljana Slovenia Phone: + 386 61 1320 068 Fax: +386 61 1314 180 Reader for English language Olga Shrestha Design Monika Fink-Serša Key words und UDC Eva Klemencic Secretaries Milica Harisch Betka Savski Printed by Tiskarna Tone Tomšic, Ljubljana, Slovenia Published quarterly Bank account number 50101 678 48454 Foreign currency account number 50100-620-133-27620-5130/6 LB -Ljubljanska banka d.d. -Ljubljana Subscription fee for institutions 100 USD, individuals 50 USD. Single issue far institutions 30 USD, individuals 20 USD. According to the opinion of the Slovenian Office of Information, the journal RADIOLOGY AND ONCOLOGYis a publication of informative value, and as such subject to taxation by 5 % sales tax. Indexed and abstracted by: BIOMEDICINA SLOVENICA CHEMICAL ABSTRACTS EXCERPTA MEDICA/ELECTRONIC PUBLISHING DIV/SION TABLE OF CONTENTS INTERVENTIONAL RADIOLOGY Transcatheter baloon valvuloplasty Pavcnik D, Cijan A, Bricelj V 175 Percutaneous removal and replacement of occluded biliary endoprostheses K6nya A, Vigvdry Z 180 Computed tomography guided percutaneous drainage of hepatic abscessus Maškovic J, Mimica Ž, Cambj-Sapunar L 186 NUCLEAR MEDICINE Radionuclide studies of the reproductive system and their significance in clinical practice Fettich J New prototype 99111Tc extraction system Fajgelj A, Novak J 191 200 CLINICAL ONCOLOGY Glomerulonephritis and cancer Burnet NG, Sadler GM, Griffiths MH, Duchesne GM 205 Neoadjuvant intraarterial chemotherapy of malignant soft tissue tumors. A new treatment protocol K6nya A, Vigvdry Z, Rah6ty P 214 Reconstruction of cutaneopharyngeal fistula of the neck· Car M, Juretic M, Štalekar H, Tomljanovic ž, Rakuljic I, Luštica I, Žgaljardic Z, Cerovic R 223 Leiomyosarcoma of the maxilla: report of a case Sotošek B 228 Breast cancer screening Jancar B 232 REPORT Visit to Centre Hospitalier Lyon Sod, France Kuhelj J 236 NOTICES 238 Transcatheter balloon valvuloplasty Dušan Pavcnik, 1 Andrej Cijan, 2 Vlado Bricelj3 1 Institute of Radiology, Clinical Center, Ljubljana, Slovenia 2Department of Cardiology, Clinical Center, Ljubljana, Slovenia 3 Department of Pediatrics, Clinical Center, Ljubljana, Slovenia Authors describe percutaneous balloon valvuloplasty in congenital and acquired heart disease. In pulmonary, mitra! and aortic stenotic valve, balloon valvuloplasty splits fused commissures and fractures calcific deposits. Since 1987 we performed 13 pulmonary, 26 mitra! and 5 aortic valve dilatation. The method is recently used as a method of choice in isolated pulmonary valve stenosis, and as an alternative to surgical commissurotomy in mitra! valve stenosis. Aortic valvuloplasty is performed in patients who are not candidates far surgery or far whom surgery carries a very high ris k. Key words: heart valve diseases-therapy; angioplasty, balloon; percutaneous balloon valvuloplasty; pulmonary, mitra! and aortic val ve Introduction The idea that catheter could be used to treat pulmonary valve stenosis was first proposed by Limon Larson in 1960, but Dotter and Judkins first described a catheter technique to dilate peripleural arteries stenosis in 1964.1 In 1974, Andreas Gruntzig developed a noncompliant balloon and performed the first coronary angio­plasty in 1977. 2 Kan et al., in 1982 performed fist balloon valvuloplasty for pulmonary steno­sis, 3 Inoue et al. for mitra! valve stenosis4 and Lababidi et al. for aortic valve stenosis. 5 In our institution percutaneous valvuloplasty has been performed since 1987. 6• 7 Correspondence to: Prof. dr. Dušan Pavcnik, Institute of Radiology, Clinical Center, Zaloška 7, 61105 Ljub­ljana, Slovenia. Material and methods Pulmonary valve Between 1987 and April 1993 balloon valvulo­plasty was used to treat congenital pulmonary valve stenosis (PS) in 13 patients. Their age ranged from 10 days to 18 years. Ali patients had isolated pulmonary valve stenosis with no associated cardiac defects. PS diagnosed by sector scan, pulsed, continuous and colour Doppler echocardiography underwent cardiac catheterisation and angiography. A pullback pressure recording across the pulmonary valve was performed. The mean peak systolic pulmo­nary valve gradient was 75 ± 26 mm Hg. The right sided 5-French catheter through the pulmonary valve was then replaced over 0.035 inch guide wire with balloon catheter UDC: 616.126.4-089.819.5 (Meditech, Schnider or Balt). The balloons Pavcnik D et al. Figure l. Two balloons of 8 mm in diameter centred across the stenotic pulmonary valve in an early phase of inflation -lateral view. were from 6 mm to 18 mm in diameter. Balloon diameters were approximately 10 % larger than the diameter of the pulmonary valve annulus. It can be used as single or double balloon technique (Figure 1). Mitra! valve Between 1988 and May 1993, we undertook balloon dilatation of mitra] stenosis in 16 female and 10 male patients. In this group of 26 patients the mean age was 46 years (range: 31-75 years), left atrium diameter (echo) was 4.5 ± 6.7 mm X 7.5 ± 5.5 mm. Patients presen­ted with severe symptoms: 3 patients were in NYHA class II, 19 patients in class III and 4 patients in class IV. Echocardiographic evaluation showed mild regurgitation (n = 6, Sellers grade 1). Fluoroscopy documented evidence of valve calcification in 6 patients. Babic's retrograde transarterial approach with double balloon tech­nique8 was performed in 11 patients and Inoue single balloon technique4 with transeptal ap­proach in 15 patients. The Inoue technique is simple and takes the shortest tiine in comparison with all other tech­niq ues. A curved guide wire of 0.025 inch is inserted in the left atrium after transseptal punc­ture. The septum is then dilated by a 14-French dilator. A single balloon (Toray Medica! Co., Tokyo) which fits well into the annulus is inflated until waist disappears. Haemodynamic parameters (PCW, MG, PT and CO) and echocardiography (MV A) were compared before and after procedure. Statisti­cal analysis was made. The long term observa­tion of clinical state and measurement were performed in 15 patients. Aortic valve Percutaneous balloon valvuloplasty was perfor­med between 1989 and 1993 in 5 patients. One patient, 22 years of age, had congenital aortic valve stenosis and 4 patients (mean age 81 years) had severe calcified aortic stenosis, and were not candidates for surgery. The technique of dilatation is grossly similar to that mentioned above. Left heart catheterisation is done by the femoral route, the aortic orifice is crossed with standard 0.038 inch or 0.035 inch guide wire. Over the wire a standard 7-French catheter is introduced into the left ventricle and left ventri­cular and aortic pressures are recorded to mea­sure the transvalvular systolic gradient. Then 300 m long 0.038 inch guide wire is positioned through the 7-French catheter into the left ventricle, the catheter is removed and the bal­loon catheter (Meditech or Balt) is introduced over the wire across the aortic valve. Balloons are 40 mm long and inflated diameter is up to 30 mm. The balloon is inflated with diluted contrast medium for a few seconds. Results Pulmonary valve Technically successful dilatation was achieved in 12 of the 13 patients. In one case we were not able to cross the critical stenosis with a guide wire , and the patient later underwent surgery. In 12 patients dilatation was successful. Transvalvular gradient was reduced from 75 ± 26 mm Hg to 29 ± 16 mm Hg after valvuloplasty (p < 0.001). We had one complication -perforation of the right ventricle outflow tract with a 0.021 inch e 1982, the efficacy and safety of this pulmonary valve stenosis in an attempt to per­form balloon dilatation of the pulmonary valve. The complication was diagnosed by contrast injection into the pericardium. The infant was followed by sector scan echocardiography and recovered completely from the injury. The dila­tation was successfully repeated a few days Iater with a soft tipped wire to cross the pulmo­nary valve. Mitra! valve Successful dilatation produced a functional im­provement in 24 of the 26 patients. In those cases valvuloplasty resulted in significant de­crease in PCW from 24.3 ± 5.2mm Hg to 14.5 ± 4.2mm Hg (p < 0.0001), mitral gradient from 15.7 ± 4.1mm Hg to 6.9 ± 3.1 mm Hg (p < 0.0001). Mitra) valve area increased from 1.00 ± 0.002 cm2 to 1.8 ± 0.03 cm2 (p < 0.0001). Follow up of 15 patients revealed slight re­duction of MV A in ali patients. In 3 patients replacement of mitra) valve after 12-1 6 months was done. Twelve patients are stili in good haemodynamic and clinical state (3 6 months). At the end of the hospital stay 18 patients were in NYHA clas I or II, 6 in class III and two in class IV. We had some mild complications as menshe­ned bellow, but no death or embolic pheno­mena occurred in this serie. Four patients had insignificant left to right shunt of atrial septum. There were two unsuccessful procedures. Mi­tra! regurgitation developed in one patient with calcified valve and in another pulmonary oe­dema developed. Mitral valve replacement was necessary. Aortic valve In ali five patients dilatation of the valve was successful, transvalvular gradient was reduced. In patients with congenital aortic valve stenosis the transvalvular pressure gradient was reduced from 75 to 30 mm Hg. In this case a mild aortic regurgitatation increased from I to stage II after procedure. One additional aortic regurgi­tation was noted in the remaining 4 patients. No further complications occurred. Redilata­tion was performed in one patient with residual stenosis. Discussion Pulmonary valve Pathology that occurs in congenital pulmonary valve stenosis is characterized by commissural fusion of the cusps. In cases with central or Iateral perforation a tipical dome shaped defor­mity is recognizable. In such valve deformity, balloon dilatation can be successfully applied with low risk.9 Significant reductions of pulmo­nary valve pressure gradients immediately after percutaneous balloon valvuloplasty and at fol­low-up studies indicate that the relief of pulmo­nary stenosis from commissural splitting or cusp teating is almost permanent. 10 When infundibu­lar stenosis accompanied pulmonary stenosis the procedure was less efficient.9 Since 1982, the efficacy and safety of this technique has been fully established and more than 1000 cases have been reported. It is con­sidered a procedure of choice for the pulmony valve stenosis. It is the most common interven­tional procedure performed on cardiac catheri­zations in children.10 Mitra! valve Mitral valve stenosis remains the commonest of ali cardiac valve disorders as a result of rheuma­tic heart disease in the developing countries. Since 1983 three technique have been develo­ped. Antegrade double balloon technique, re­ Pavcnik D et al. Figure 2. Inoue balloon mitra] valvuloplasty. Inflated balloon across the narrow mitra] valve. trograde double balloon technique and Inoue 89 balloon technique.4• • Antegrade double bal­loon technique was in the past the most widely used technique. Besides complications resulting from transseptal puncture, embolic phenomenon or mitra[ regurgitation which can happen in ali the techniques, left to right shunting is frequent in this technique as the atrial septum has to be dilated to insert dilation balloon catheters. Du­ring follow-up period, most of the shunts disap­pear. A well conducted study has shown that about 10 % of ali the patients who underwent this technique, had a significant shunt after a period of 6 months.11• 12 We never used this technique in our hospital. Retrograde double balloon technique. In this technique, the balloons are inserted through the femoral arteries over long guide wires which are inserted in the femoral vein and brough into the aorta through a transseptal puncture and exteriorised through femoral arteries by the help of a snare. This procedure has the advantage of not producing any left to right shunt. This technique is relatively complex. Local complications at the site of femoral arte­ries are more frequent. 13 This technique was used in our hospital in 11 patients. Inoue balloon technique. Inoue who has per­formed more than 1000 cases of mitra! balloon valvuloplasties has named this procedure as percutaneous transluminal mitra! commissuro­tomy (PTMC). This technique has a lot of advantages over other techniques, however, the major drawback has been the short tirne available for the procedure in order to minimize the complications rate14 (Figure 2). Closed mitra! commissurotomy was for a long tirne the only effective treatment for mitra! stenosis.15 Balloon mitra! valvuloplasty is un­doubtedly an attractive alternative to surgical commissurotomy. Initially, it was limited to young patients with pliable non-calcified valves. Mitra) valvuloplasty has now been extended to calcified valves and in patients with moderate involvement of the subvalvular apparatus. The commissural splitting is a mechanism in success­ful balloon technique. A good haemodynamic stability was achieved in 70 % of patients for five years if one commissure was split and in 89 % of patients if both commissures were split. In patients with calcified valves dilatation im­proves valve area through commissural opening and fracturing the nodular depozit.16· 17 Its use in recurrent mitra! stenosis is safe and effective. The main contraindications for this technique is presence of fresh thrombus and significant mitra! regurgitation. Aortic valve The clinical improvement of dilated patients is often dramatic, but restenosis note is consider­ably high.18 It is unknown which patients will receive long term benefits from balloon dilata­tions.19 It appears to be a technique of choice in patients who are not surgical candidates or for whom surgery carries a very high risk, i.e. mainly in elderly patients. It is also considered as an effective and less invasive alternative to surgical valvotomy in infants and children par­ticularly in the cases of restenosis after surgical valvotomy where the need for the implantation of a prosthetic val ve is otherwise mandatory. References l. Dotter CT, Judkins MP. Transluminal treatment of arteriosclerotic obstruction. Description of a new technic and a preliminary report of its appli­cation Circulation 1964; 30: 654-70. 2. Gruntzig AR. Transluminal dilatation of coronary artery stenosis. Lancet 1978; 1: 263-4. 3. Kan JS, White Rl, Mitchell SE, Gardner TJ. Percutaneous balloon valvuloplasty: A new met­ hod for treating congenital pulmonary valve steno­ sis. N Engl J Med 1982; 370: 540---3. 4. Inoue K, Owaki T, Nakamura T, Kitamura F, Miyamoto N. Clinical application of transvenous mitra! commissurotomy by a new balloon catheter. J Thorac Cardiovasc Surg 1984; 87: 394-402. 5. Lababidi Z, Wu JR. Walls JT. Percutaneous bal­loon aortic valvuloplasty: Results in 23 patients. Am J Cardiol 1984; 53: 194-7. 6. Robida A, Pavcnik D. Perforation of the heart in a newborn with critical valvar pulmonary stenosis during balloon valvoplasty. lntervent J Cardiol 1990; 26: 111-2. 7. Pavcnik D, Cijan A, Robida A, Šurlan M. Posegi intervencijske radiologije v prsnem košu. 10 Ra­denski dnevi 1989; 46 (abstract). 8. Babic U, Pejcic P, Djurišic Z, Vucinic M, Grujicic S. Percutaneous transarterial balloon valvuloplasty for mitra! valve stenosis. Am J Cardiol 1986; 57: 1101-4. 9. Becker AE, Hoedemaker G. Balloon valvulopla­sty in congenital and acquired heart disease: Mor­phologic consideration. Z Kardiol 1987; 76 (6): 73-9. 10. Stanger P, Cassidy SC, Girod DA. Balloon pulmo­nary valvuloplasty: Results of valvuloplasty and angioplasty of congenital anomalies. Radiology 1990; 177: 599---{500. 11. Vahanian A, Michel PL, Cormier B, et al. Results of percutaneous mitra! commissurotomy in 200 patients. Am J Cardiol 1989; 63: 847-52 . 12. Reid CL, McKay CR, Chandraratna PA, Kawa­nishi DT, Rahimtoola SH. Mechanisms of increase in mitra! valve area and influence of anatomic features in double-balloon, catheter balloon valvu­loplasty in adults with rheumatic mitra! stenosis: A Doppler and two dimensional echocardiograp­hic study. Circulation 1987; 76 (3): 628-36. 13. Babic U, Dorros G, Pejcic P, Djurišic Z, Grujicic S. Percutaneous mitra! valvuloplasty: Retrograde, transarterial double-balloon technique utilising the transseptal approach. Catheterzation and cardio­.vascular diagnosis 1988; 14: 229-37. 14. Nobuyoshi M, Hamasaki N, Kimura T, Chen CH. Indications, complications and short-team clinical outcome of percutaneous transvenous mitra! com­ missurotomy. Circulation 1989; 80: 782-92. 15. Belcher RJ. Closed mitra! valvotomy. In: Jackson WJ ed. Operative surge1y, Fundamental internatio­nal technique cardiothoracic surge,y. London, Bo­ston: Butterworth, 1978: 158---{53. 16. Kaplan JD, Isner JM, Karas RH. In vitro analysis of mechanisms of balloon valvuloplasty of stenotic mitra! valves. Am J Cardiol 1987; 59: 318-23. 17. Block PC, Pallacios IF, Jacobs M, Fallon JJ. The mechanism of successful mitra! valvotomy in hu­mans. Am J Cardiol 1987; 59: 178-82. 18. Cribier AI, Berland J, Savin T, Koning R, Mech­meche R, Letac B. Percutaneous aortic valve dilatation: Indications and results in adult acquired calcified aortic stenosis. Z Kardiol 1987; 76 (6): 90---103. 19. Davidson DJ, Harrison JK, Leithe ME. Failure of balloon aortic valvuloplasty to result in sustai­ned clinical improvement in patients with depres­sed left ventricular function. Am J Cardiol 1990; 65: 72-7. Radio/ Oncol 1993; 27: 180-5. Percutaneous removal and replacement of occluded biliary endoprostheses Andras Konya and Zoltan Vigvary Department of Radiology, Semmelweis University of Medicine, Budapest, Hungary Percutaneous removal and replacement of occluded biliary endoprostheses performed five times in three patients with highly located malignant biliary stenoses are presented. Percutaneous manipulation and its difficulties are discussed in detail. The methods available far treating obstructed endoprostheses are also described with special regard to percutaneous approach. Percutaneous stent exchange was performed once in one patient, and twice in two other patients. With all these interventions a survival of 25, 60 and 89 weeks, respectively, could be achieved. The authors emphasize that percutaneous method is indicated only if endoscopic attempts are excluded or have failed. Despite its invasiveness percutaneous stent replacement in skilled hands is considered a relatively simple and safe procedure which is able to improve greatly the patient's quality of life. Key words: bile ducts, interventional procedure, bile ducts neoplasms-surgery, stents, endopros­thesis -transhepatic removal of stents, biliary endoprosthesis; complications, prosthesis failure Introduction Insertion of a percutaneous transhepatic endo­prosthesis is an accepted option for the pallia­tive treatment of inoperable malignant strictu­res of the bile duet mainly if there is no chance for an endoscopic stent placement. One of the major arguments against percutaneously placed endoprostheses, however, is the difficulty of replacing or exchanging them percutaneously when they become occluded. The fact that hilar, intrahepatic and/or multiple lesions can only be treated by percutaneous precedures Correspondence to: Andras K6nya M.D., Ph. D, Department of Radiology, Semmelweis University of Medicine, H-1082 O116i ut 78/a, Budapest, Hungary UDC: 616.361-006.6-089.28 including insertion of double/multiple endo­prostheses further complicates this problem. For this reason, some interventional radiologists reserve endoprosthesis for patients with a short life expectancy. 1 An increasing number of ra­diologists consider that the drawbacks of exter­nal/internal drainage render the use of this system for the relief of biliary obstruction in appropriate for most patients with malignant disease, including those with a relatively long life expectancy. 2•3•4 In our institution the patients who are not surgical candidates or not suitable for endosco­pic stent placement receive one or more percu­taneously placed endoprostheses instead of one or more external/internal catheters. For this reason, we are ready to treat the occluded endoprostheses either endoscopically or percu­taneously. Percutaneous removal and replacement of occluded biliary endoprosth.eses We present bere our experience gained in 3 patients with exchanging endoprostheses 5 ti­mes by percutaneous approach. Patients and methods Ali three patients had malignant bile duet ob­struction, where tumorous lesions were bighly located and/or intrahepatic (No 1: gallbladder carcinoma, No 2: cholangiocarcinoma, No 3: gastric carcinoma -sbigellocellular type). In two patients a 14 French Carey-Coons endo­prosthesis (Meditecb, U.S.A.), while in one a 12 Frencb Lammer endoprostbesis (COOK) had been inserted using a several-step tecbni­que. In pa_tient No 3 another Allison-Gibson spiral sbaped stent (COOK) had also been placed to create and maintain communication between tbe two !obes of tbe )iver. Ali patients presented witb recurrent jaundice (19, 26 an 48 weeks after the initial stent placement). In patient No 3, two successful endoscopic removals and replacements of tbe occluded prostbeses in tbe common bile duet bad been carried out in a period of 48 weeks. Ultrasound examination confirmed tbe su­spected occlusion. In patient No 3 an attempt for endoscopic excbange failed at tbat tirne. Transbepatic cho!angiography (PTC) was then performed and an appropriate segmenta! bile duet leading directly to tbe proximal end bole of the occluded endoprosthesis was chosen for anotber fine needle puncture. In ali of three cases after a successful puncture of tbe right bile duet a 0.018 inch superglide guide wire witb excellent torque control (RadiofocusR, Te­rumo, Japan) was inserted througb tbe lumen of a Chiba needle. In patients No 2 and No 3 (21 and 22 weeks later) furtber attempts were made to excbange the existing endoprostbesis for anotber one. In these cases, the proximal extension of the tumor did not leave enougb room for sufficient mani­pulation from the right bile duet leading to tbe proximal end of tbe endoprosthesis. Tberefore, the proximal end hole of the stent was directly punctured and cannulated witb a 22 gauge Chiba needle and a 0.018 incb guide wire being inserted into the lumen of tbe obstructed endo­prostbesis. In tbe first three cases where tbe initial step included bile duet manipulation, after baving excbanged tbe 0.018 inch guide wire for a 0.038 incb wire, a 9.3 French Portner-Koolpe biliary biopsy device (COOK) with a double cbannel was inserted over the 0.038 incb guide wire. This wire served later as a "safety" guide, while another 0.032 incb superglide wire with excel­lent steerability was also placed into tbe com­mon bile duet for cannulating tbe clogged lumen of tbe stent. After withdrawal of tbe double channeled device two guide wires were in tbeir place and one of tbem was applied for entering tbe incrusted lumen wbile tbe "safety" guide ensured the already achieved position in tbe bile duet and made tbe manipulation safer. Once tbe guide wire was positioned into tbe stent's lumen a 5 French straight angiograpbic catbeter was advanced into the stent over the 0.032 inch wire (or over tbe 0.018 inch guide wire when the occluded lumen was punctured directly) which was then exchanged for a 0.038 inch superglide or Lunderquist (COOK) wire. Tbe parenchymal tract was dilated for 14 or 16 Frencb using progressively larger dilators. A 6 or 7 French balloon catheter (balloon lengtb 3-4cm, inflation diameter 4-5 mm) (Meditecb) was inserted over tbe guide wire into the proxi­mal part of the stent. Tbe balloon was inflated and tbe catbeter and endoprostbesis as a unit were withdrawn togetber. In three cases the removal and excbanging were performed in the same session (2 Carey-Coons, 1 Miller double musbroom endoprostheses). In two cases tbe new prostbesis was placed after a few days' external/internal drainage (2 Carey-Coons prostheses). In the lumen of the new endopros­tbesis a 5 Frencb straight "safety" catheter was left for one or two days. If the control contrast filling did not reveal any dysfunction of the stent, tbe catbeter was witbdrawn and tbe pa­rencbymal tract was tben embolized with Gel­foam to prevent bile leakage. Figures la, b. Patient No 3 with highly located multiple malignant stenoses originating from recurrent gastric carcinoma (shigellocellular type). A spiral-shaped en­doprosthesis had been placed connecting the two !obes of the !iver. After puncturing the right sided bile duet leading to the proximal end of the occluded 14F Results In our three patients the initial endoprosthesis patency lasted 19, 26 and 48 weeks, respective­ly. In patient No 3 the 48-week patency could be achieved by two additional endoscopic stent removals and changings (Figures la, b). Second stents remained patent for 16, 21 and 22 weeks, respectively. Patient No 1 died of tumor dissemination without recurrence of jaundice. In patients No 2 and 3 a third endo­prosthesis was implanted after successful percu­taneous removal of the occluded stent. These patients survived additional 13 and 19 weeks, respectively (Figures 2a, b, c). Discussion There are severa! possibilities for treating oc­cluded endoprostheses: l. Endoscopic extraction and exchanging. Stent replacement is undoubtedly more easily Carey-Coons endoprosthesis, a double channeled de­vice was used far inserting two guide wires. One of them served as a "safety" guide while the other was manipulated into the stent lumen. After appropriate dilatation of the parenchymal tract over the latter guide wire a balloon catheter was then advanced in the proximal part of the stent. After balloon inflation the stent moved downward first, but mainly because of its size we decided to withdraw it percutaneously instead of pushing it into the duodenum. The intlated balloon catheter together with the stent was pulled out with careful traction. achieved endoscopically than percutaneously, and its complication rate is lower; the patient's discomfort is significantly less serious.5 Anato­mical situation caused by the bulk of the tumor and/or some fonns of previous surgery may prevent the passage of the endoscope or cannu­lation of the papilla. 2. Deocclusion by unclogging the lumen. The deocclusion can be accomplished by using a brush biopsy wire inserted percutaneously through a suitable biliary radicle.4 After tho­rough washing a "safety" catheter is left in the lumen for one or two days to ensure the patency of the stent. The main disadvantage of the method is that it requires tirne consuming per­cutaneous manipulation, and despite successful deocclusion the incrusted endoprosthesis as a source of potential incrustation remains in its place. 3. Pushing the stent into the duodenum with new prosthesis replacement. From a suitable Percutaneous removal and replacement of occluded biliary endoprostheses Figures 2a, b, c. 3 months later a partial occlusion occurred due to incrustation of the stent. This tirne the proximal end hole of the endoprosthesis was punctured directly by a 22 gauge Chiba needle. A 0.018 inch superglide guide wire was inserted into the lumen of the fine needle, which was exchanged for a 0.038 inch Lunderquist wire. Over this guide a balloon catheter was inserted and by means of that the stent was successfully withdrawn. percutaneous approach the occluded prosthesis can be pushed downward into the duodenurn by a pusher (which is a part of the introduction set) or by using a balloon catheter.3• 6 Although this rnethod is less traurnatic to the liver paren­chyrna than the percutaneous rernoval, yet it rnay also need a lengthy rnanipulation for can­ 6 nulation of the clogged lurnen. 4. Percutaneous transhepatic extraction and exchanging. The use of angioplasty balloon catheters for the repositioning or transhepatic rernoval of biliary endoprostheses has been reported previously.3• 4• 7 In these reports access to the endoprosthesis was obtained by cathete­risation of a suitable biliary duet and rnanipula­ tion of either an angiographic catheter or a guide wire into the endoprosthesis. This rnethod bas severa! disadvantages: rnani­pulation of the guide wi.re or catheter into the stent is sornetirnes rather difficult, the catheter, or guide wire rnay fail to enter the stent, and ali this rnay result in repeated punctures of the biliary systern. 2 In three of our five cases the endoprosthesis cannulation was successful, but the tirne needed for this was relatively long (15-25 rninutes fluo­roscopy tirne) even with the use of a guide wire with excellent torque control. Therefore, on the last two occasions we punctured the proxi­mal end-hole of the occluded endoprosthesis with the Chiba needle directly according to Adam.2 We were also forced to choose this kind of procedure by the fact that the proximal progression of the tumor made the anatomical situation more complicated, thus further limi­ting the possibilities of necessary manipulation. In these instances the absence of contrast me­dium in the bile ducts ensures excellent visuali­sation of the endoprosthesis, which facilitates its cannulation.2 The way of removal of occluded endoprosthe­ses is greatly influenced by their shape and configuration. Percutaneous distraction of an endoprosthesis with special end-configuration (Miller double mushroom tipped or Allison­Gibson spiral shaped endoprostheses) may be much more difficult than that of a larger bore stent (Carey-Coons type) and may need special • 3• 6• 8• devices29 After successful cannulation of the endoprostheses' lumen we had no further technical difficulties with their distraction (4 Carey-Coons type and 1 Lammer endoprosthe­ses). Using percutaneous removal, one must take into account that this procedure requires a significant parenchymal tract to be create, but this disadvantage can be turned into advantage as another endoprosthesis should otherwise be implanted. Gelfoam embolization of the tract prevented bile leakage in ali of our cases. Indwelling biliary stents should be used rou­tinely in patients requiring non-surgical relief of obstructive jaundice. The availability of a simple and effective method of replacement of malfunctioning endoprostheses will encourage their application in preference to external/inter­nal drainage, thereby improving the quality of the patient's remaining life. More recently the presence and availability of self-expandable metallic stents have given a breakthrough also in the treatment of malignant biliary stenoses. These devices can be inserted by a 5-10 French balloon catheter, therefore, it does not require significant dilatation of the parenchymal tract. Furthermore, there is no need for a multi-stage procedure and since the metallic stents provide a lumen of 1 cm the potential for occlusion is much less than that or other plastic endoprostheses.10, 11, 12 The use of self-expandable metallic stents increases the cost of biliary drainage due to their high price. But they can be placed in one session, thereby markedly reducing the cost of hospitalisation, and eliminating the need for a multi-stage procedure with ali the cost as well as discomfort to the patient involved. The cost of treatment can be further diminished by de­creasing the rate of potential complications. The total fluoroscopy tirne needed for lengthy manipulation can also be decreased significant­ly, thus keeping the amount of scattered radia­tion, which both the personnel and the patient are exposed to, within acceptable limits. 13 Ali these advantages can effectively counter­balance the high price, the only major drawback of self-expandable metallic stents. We are strongly confirmed that the use of these promi­sing devices will be the tratment of choice in selected patients with highly located, intrahepa­tic and/or multiple malignant stenoses who can expect a relative long survival. References l. Ferrucci JT. Biliary endoprosthesis. In Jnterventio­nal Radiology of the Abdomen. ed Ferrucci JT, Wittemberg J, Mueller PR and Simeone JF. Wil­liams and Wilkins, Baltimore. 1990; 267-81. 2. Adam A. Use of the modified Cope introduction set for transhepatic removal of obstructed Carey­Coons biliary endoprosthesis. Ciin Radio/ 1987; 39: 171-4. 3. Jackson JE, Roddie ME, Yeung EYC, Benjamin IS, Adam A. Biliary endoprosthesis dysfunction in patients with malignant hilar tumors: successful treatment by percutaneous replacement of the stent. Am J Roentgenol 1990; 155: 391-5. 4. Teplick SK, Haskin PH, Pavlides CA, Goldstein RC. Management of obstructed biliary endoprost­heses. Cardiovasc Intervent Radiol 1985; 8: 164-7. S. Gibson RN. Transhepatic biliary endoprostheses. Jntervent Radio/ 1989; 4: 7-12. 6. Yeung EYC, O'Donnel C, Carvalho P, Adam A. A new technique for removing occluded double mushroom-tipped biliary endoprostheses. Am J Roentgenol 1989; 152: 527-8. 7. Honickman SP, Mueller PR, Ferrucci JT, vanSon­nenberg E, Kopans DB. Malpositioned biliary endoprosthesis: Radiology 1982; 144: 423-5. Percutaneous removal and replacement of occ/uded biliary endoprostheses 8. Brown AS, Mueller PR, Ferrucci JT. Transhepatic removal of obstructed Carey-Coons biliary endo­prosthesis. Radiology 1986; 159: 555-6. 9. Yeung EYC, Adam A, Gibson RN, Benjamin IS, Allison DJ. Spiralshaped biliary endoprosthesis: initial study. Radiology 1988; 168: 365-9. 10. LaBerge JM, Doherty M, Gordon RL, Ring EJ. Hilar malignancy: treatment with an expandable metallic transhepatic biliary stent. Radiology 1990; 177: 793-7. 11. Lameris JS, Stoker J, Nijs HGT, Zonderlan HM, Terpstra OT, Blankenstein M, Schi.itte HE. Malig­nant biliary obstruction: percutaneous use of self­expandable stents. Radiology 1991; 179: 703-7. 12. Lammer J, Klein GE, Lkeinert R, Hausegger K, Einspieler R. Obstructive jaundice: use of expand­able metal endoprosthesis for biliary drainage. Radiology 1990; 177: 789-92. 13. K6nya A, Vigvary Z. Personnel exposure to radia­tion at biliary interventional radiological procedu­res with an overhead tube. Radio/ Oncol 1992; 26: 150-6. Radio/ Oncol 1993; 27: 186-90. Computed tomography guided percutaneous drainage of hepatic abscessus Josip Maškovic,1 Željko Mimica,2 Liana Cambj-Sapunar1 1 KBC Split Zavod za radiologiju, 2 KBC Split Klinika za kirurške bolesti, Croatia The indications, technique and results of percutaneous drainage in 60 patients with 64 hepatic abscesses are presented. The procedure was guided by computed tomography (CT). Evacuation of the cavity was achieved in 59 (92.2 %) abscesses and surgery treatment was not necessary in 55 (91.7%)patients. Percutaneous abscess drainage failed in 5 (8.3%) patients. There were no major complications. The drainage catheter was removed on average 14 days after insertion. We believe that percutaneous drainage of hepatic abscesses guided by computer tomography is an effective alternative to operative drainage, espacially in severely ill patients. Key words: liver abscess; drainage; tomography, x-ray computed Introduction The incidence of hepatic abscesses has not changed significantly in the past 50-60 years, but with respect to age, incidence has shifted to older population.1 At the turn of this century, the majority of hepatic abscesses were amoebic in origin.2 Non­ameobic abscesses, in particular pyogenic ab­scesses, may spread by direct propagation of contagious disease processes such as cholecysti­tis or peptic ulcer. However, other modes of transportation of bacteria's virulent strains are drainage via the portal system, such as necrotic Correspondence to: Doc. dr. Josip Maškovic, KBC Split, Zavod za radiologiju, Spinciceva 1, 58000 Split, Croatia. inflammed bowel, trauma with penetrating inju­ries and direct introduction into the liver, and blood-borne infections via hepatic arteries.2-4 Traditionally, abscesses have been treated surgically, with wide incision and drainage as the main principles of surgical treatment. The incorporation of computed tomography and ul­trasound (US) allows detection and accurate location of fluid collections in hitherto inacces­sible sites. This has led to the development of diagnostic aspiration which is now being exten­ded to terapeutic aspiration of the located 6 fluid.5• In 1978, Gerzof7 advocated percuta­neous catheter as routine treatment for absces­ses. In 1983, Walch8 concluded that percuta­neous catheter drainage had then been acknow­ledged as "one of great advances in abdominal surgery in the past few years". We present our UDC: 616.36-002.3-089.48 favorable experience. Materials and methods Percutaneous hepatic abscess drainage was per­formed on 64 occasions in 60 patients during a five-year period (1988-1992) Amoebic abscess was present only in one patient. Piogenic ab­scesses were present in 39 (65 % ) patients. Four of them had two abscesses in the liver. Hema­toma and tissue devitalization after blunt trauma to the !iver were the sites of abscesses in 20 patients. CT was performed with Somatom DR Sie­mens scanner. Examinations were performed in the supine position, and prone and lateral decu­bitus position were added when reguired. Ali patients received 60 to 90 ml of ioheksol con­trast medium (Omnipague -Nycomed). When optimal visualization of the lesion was achieved and the puncture site selected, an opague mar­ker was placed on it and scan repeated to confirm the position. The site adjacent to the lesion was selected and the intended depth and angle of puncture were calculated from measu­rements made on the console monitor. Rescan­ning with a needle in site was performed after aspiration to abscess the amount of residual fluid. After localization, the site selected for puncture was sterilized and local anesthetic injected. Initially, an 18 gauge teflon-coated catheter-needle combination was used. When thick pus was encountered, this catheter was exchanged for a larger one by using a guide-wi­re. As much fluid as possible was withdrawn and bacteriologically and chemically analysed (Figures la, b, c). The total volume of the aspirate was measured and compared to that estimated from scan. If a gross discrepancy was found, a repeat scan is indicated to rule out an undrained loculus, since this may reguire a second drai­nage catheter. The catheter was left in situ and removed when the temperature and white blood celi count returned to norm.I and when daily drainage volumes decrease to less then 5-10 ml per day. A fina! CT scan is generally unneces­sary, but may be used to confirm the res uit. Results Percutaneous aspiration was performed in five patients with hepatic abscesses. One of them was a patient with amoebic abscess. A single drainage procedure was sufficient in 51 (92.7 % ) out of 55 patients. Multiple punctures were required in the remaining four patients who had a second catheter placed. Drainage of the cavity was accomplished in 59 (92.2 % ) out of 64 abscesses. Percutaneous abscess drainage failed in five cases. In ali these patients, drainage of nondis­ Maškovic J et al. crete areas of necrotic material was attempted. In two patients surgical exploration was perfor­med on the fourth day and in three patients six days after the percutaneous drainage. Three collections recurred. One was an in­adequate drainage of infected hematoma after a blunt !iver trauma, and two were inadequate drainages of pyogenic abscesses. This recurrent collections were retreated percutaneously. The smallest fluid collection aspirated was 10 ml in a patient with amoebic abscess, whe­reas the largest was 500 ml in a patient with infected hematoma. Mean fluid collection by drainage was 124 + 74 cc. Fifty-three (96.4 % ) out of 55 patients with sufficient percutaneous abscess drainage became afebrile within 48 hours and leukocytosis returned to normal wit­hin 7-10 days. When the temperature and white blood cell count returned to normal, and daily drainage volumes decreased to less than 5­10 ccm, the catheter was removed. The duration of the percutaneous drainage was 6-21 days, mean 14 + 5. Discussion The percutaneous drainage of abscesses is beco­ming an accepted treatment method in many hospitals. The optimal use of this procedure reguires close cooperation between a radiologist and the referring physician. The radiologist performing the procedure should decide on the optimal modality to image the abscess and if one method is found to be inadequate, there should be no hesitation trying another one. The following criteria should be considered in attempting percutaneous abscess drainage: 1) a well-defined abscess cavity, 2) a safe per­cutaneous access route, 3) concurring surgical opinion and 4) capability for immediate opera­tive intervention in case of failure or complicat­ion. 9• 10 Contraindications include the absence of any of the above criteria or the presence of a bleeding diathesis.9 Some authors11-13 consi­der multiple and complex abscesses as a con­traindication to percutaneous drainage. Donde­linger14 in his study showed scarcely any diffe­rence in the clinical outcome when comparing multiple or complex abscesses with unilocular abscesses. Interna! loculation of abscess is not a contra-indication for percutaneous drainage10 since it occurs rarely and since loculated hepatic abscess may be treated by placement of sepa­rate catheter to drain each loculus. Because the radiologist's role of treating ab­dominal abscesses and fluid collection has ex­tended beyond the traditional task of detection, to the additional steps of percutaneous needle aspiration and percutaneous drainage, an inte­grated and complementary use of CT and US is essential. However, even if the fluid co!Jec­tion is delineated by initial US study, CT may be required to further define the extent of the process and to relate the collection to the surrounding structures. Once the presence and location of the fluid collection have been deter­mined, percutaneous needle aspiration and sub­seguent percutaneous catheter drainage can be performed. A small abscess may be treated by complete aspiration of pus via a needle without using catheter.16 We used this technique in five pa­tients. For percutaneous drainage of hepatic abscess the shortest possible route to abscess cavity that avoids vital structures should be selected. Drai­nage route must avoid the gallbladder, portal vein and hepatic flexure of the col on .17 Many abscesses have significant interna! pres­sure, and rapid free flow of pus is quickly accomplished by aspiration with a large syringe. At the initial procedure, the abscess cavity is completely evacuated. The catheter and the cavity are irrigated with sterile saline. This step prevents early blocking of the catheter by ne­crotic tissue and debris. The utility of daily irrigation of catheter is debatable. Some authors do not irrigate the catheter as long as the purulent material is draining. 18 In Gerzof19 experience, lavage cau­sed rupture of abscesses precipitating emer­gency surgery treatment and at least one conse­guent death. Haaga6 preferes irrigation with a smaller amount of normal saline, 1-25 ml/hr, depending on the size of the cavity, with gradual decrease in the amount of irrigant until it is injected only once every day. We prefer such irrigation in our patients. Van Weas suggests instillation of a proteolyitic agent ( acetylcy­ 20 stein), antibiotics, or both.Miller21 irrigates the abscess cavity with normal saline immedia­tely after catheter placement and follows this with constant gravity drip of saline at 50-lOOml/hr. Surgeons have traditionally racommended dependent or gravitational positioning of their drainage catheter. Gerzof19 finds such positio­ning to be unnecessary. Most percutaneous catheters are placed anteriorly and experience has shown they drain well against gravity. Be­cause both the abscess wall and the abdominal wall remain intact, they transmit increasing intra-abdominal pressure on any residual ab­scess fluid, forcing it out of the drainage cathe­ter. Even when the initial pus in very thick, as soon as the pressure in the abscess is relieved the ensuing exudate produced by the abscess wall is very thin and drains easily. The first 72 hours of drainage are the most important for determining its adequacy. During that tirne most patients become afebrile. If clinical sings of sepsis persist beyond 24-48 hours, percutaneous drainage may not have been sufficient. A repeated CT is then required to rule out undrained residuum. 19 Success of catheter drainage can be assessed by clinical, catheter, and radiological follow-up. The clini­cal features of drainage success are decrease in fever, pain, local tenderness, and leukocytosis. Catheter features are decreased drainage, and return of clear irrigated saline. Radiographic feature is the diminishing of the cavity on CT scan, US or contrast abscessogram. Clinical response of patients with pyogenic hepatic abscess can be dramatic. Prompt defer­vescence and fall in leukocytosis, along with an improved sense of well-being, are customary. However, one or more temperature pikes may occur in the hours following catheter insertion. Reduction of the local pain and tenderness are additional positive signs. Gradual decreases in the amount and visco­sity of the catheter drainage are characteristic in successful cases. However, a sudden cassa­tion of catheter output may indicate catheter plugging by necrotic debris. This can usually be corrected by performing irrigation on ward. Follow-up by subsequent US, CT or absces­sogram is valuable to assess the cavity size and to disclose walled-off or undrained loculi. A direct contrast abscessography is more powerful than either US or CT. The average hospitalization time22 • 23 is 40 days, compared to 21 days with percutaneous treatment. In our patients mean hospitalization tirne was 23 days. Percutaneous treatment of !iver abscesses should be considered the treatment of choice when a suspicious image is evidenced on US or CT. Complete cure is achieved in 76 %24 to 25•26 100 % of cases. In our research percuta­neous treatment of !iver abscesses was much more effective than surgical drainage in 55 (91. 7 % ) patients. When an underlying causal disease persists, this should be treated seconda­rily in order to prevent abscess recurrence. Septic complications and bleeding at the tirne of drainage are exceptional. The mortality asso­ciated with percutaneous treatment of !iver abscess is less than 5 % , which compares very favorably with the surgical literature. 12 Conclusion Percutaneous drainage of !iver abscesses is a rapid, safe and highly effective method of treat­ment which, in most cases replaces major surge­ry. CT should be used in ali instances for diagnosis, location of the Jesion and for plan­ning optimal approach to drainage, as well as to access the degree of aspiration achieved during and after the procedure. Surgical expo­sure provides decompression and evacuation, and surgically placed drainage provides conti­nuous drainage. These mechanical processes are performed by percutaneously placed drai­nage without a damage of surrounding tissues. The most essential advantage of percutaneous drainage over operative drainage is the total avoidance of general anesthesia. Percutaneous drainage in hepatic abscesses, compared with surgical treatment, yields a higher success rate Maškovic J et al. with a shorter hospitalization and correspondin­gly lower rates of cornplications, recurrence and rnortality. References l. Pitt HA, Zuidema GD. Factors influencing morta­lity in the treatment of pyogenic hepatic abscess. Surg Ginecol Obstet 1975; 140: 228-31. 2. Landay MJ, Setiawen H, HHirsh G, Christansen EE, Conrad MR. Hepatic and thoracic amoebia­sis. AJR 1980; 135: 449-54. 3. Balasegarm M. Management of hepatic abscess. Cur Prob Surg 1981; 18: 3-7. 4. Hiatt JR, Williams RA, Wilson SE. Intraabdomi­nal abscess; etiology and pathogenesis. Semin Ul­trasound 1983; 4: 27-31. 5. Callen PW. Computed tomography evaluation of abdominal and pelvic abscesses. Radiology 1979; 131: 171-5. 6. Haaga JR, Weinstain AJ. CT-guided percutaneous aspiration and drainage of abscesses. AJR 1980; 135: 1187-94. 7. Gerzof SG, Robbins AH, Birkett DH. Computed tomography in the diagnosis and management of abdominal abscesses. Gastrointest Radio/ 1978; 3: 387-49. 8. Walch CE, Malt RA, Abdominal surgery. N Engl J med 1983; 308: 753-60. 9. Gerzof SG, Robbins AH, Birkett DH, Johnson Wc, Pugatch RD, Vincent ME. Percutaneous ca­theter drainage of abdominal abscesses guided by ultrasound and computed tomography. AJR 1979; 133: 1-8. 10. Gerzof SG, Spira R, Robbins AH. Percutaneous abscess drainage. Semin Roentgenol 1981; 16: 62­71. 11. Deveney CW, Lurie K, Deveney KE. Treatment of intraabdominal abscesses: a result of improved localization, drainage, and patient care, not tehni­que. Arch Surg 1988; 123: 1126-30. 12. Johhnson WC, Gerzof SG, Robbins AHH, Neb­seth DC. Treatment of abdominal abscesses: com­parative evaluation of operative drainage versus percutaneous catheter drainage guided by compu­ted tomography or ultrasound. Ann Surg 1981; 194: 510-20. 13. Lurie K, Pizak L, Deveney CW. Intraabdominal abscess in the 1980s. Surg Ciin Northh Am 1987; 67: 621-35. 14. Dondelinger RF, De Beats P, Kurdziel JC. Percu­taneous aspiration and drainage of abdominal abscesses under X-ray computed tomography con­trol: perspective study appropos of 63 cases. Ann Radio/ 1987; 30: 373-8. 15. Muller PR, Simone JF. Intra-abdominal abscesses: diagnosis by sonography and tomography. Radio/ Ciin North Am 1983; 21: 425-43. 16. Gronvall S. Drainage of abdominal abscesses gui­ded by sonography. AJR 1982; 138: 527-31. 17. Haage JR. CT guided procedures. In Haage JR, Alfidi RJ (eds): Computed tomography of the whole body. St Louis, C. V. Mosby, 1983; 867­933. 18. Gerzof SG. Percutaneous cateter drainage of ab­dominal abscesses: 5-year experience. N Engl J Med 1981; 305: 653-7. 19. Gerzof Sg. Results and clinical correlations of percutaneous abscess drainage. In: RSNA, ed. Categorical course on interventional radiology. 1985; 173-8. 20. Van Waes PFGM. Management of loculated ab­scesses that are difficult to drain: an new ap­proach. Radiology 1983; 147: 37-41. 21. Miller MH, Frederick PR, Tocine I, Bahr AL. Percutaneous catheter drainage of intraabdominal fluid collections including infected billiary ducts and gallbladers. Am J Surg 1982; 144: 660-7. 22. Aeder Ml, Wallman JL, Haage JR, HHau T. Role of surgical and percutaneous drainage in the treatment of abdominal abscesses. Arch Surg 1983; 118: 273-80. 23. Brolin RE, Nushher JL, Leiman S, Lee WS, Greco RS. Percutaneous catheter versus open surgical drainage in the treatment of abdominal abscesses. Am Surg 1984; 50: 102-7. 24. Johnson RD, Mueller PR, Ferrucci JT Jr, Dawson SL, Butsc RJ, Papanicolau N. Percutaneous drai­nage of piogenic !iver abscess. AJR 1985; 144: 463-8. 25. Barger LA, Osborne DR. Treatment of pyogenic abscesses by percutaneous needle aspiration. Lan­cent 1982; 1: 132--{i. 26. van Sonnenberg E, Ferrucci JT Jr, Mueller PR, Wittenberg J, Simeone JF, Malt RA. Percuta­neous radiographically guided catheter drainage of abdominal abscesses. JAMA 1982; 247: 190-2. Radionuclide studies of the reproductive system and their significance in clinical practice Jure Fettich Department of Nuclear Medicine, Clinical Centre Ljubljana, Slovenia Radionuclide studies of the reproductive system are used to evaluate perfusion and function of this system and are frequently complementary to morphological studies such as ultrasound and X-ray investigations. Radionuclide scrotal imaging remains the most effective method in the evaluation of acute hemiscro­tum to assess testicular perfusion and differentiate between testicular torsion and epididymitis. Labelled red blod cells are used successfully for varicocele detection in infertile men. Only radionuclide techniques (radionuclide hysterosalpingography) allow for noninvasive assessment of functional tuba! patency in infertile women. Assessment of penile blood flow, using labelled red blood cells and/or xenox clearance, can play an important and useful role in evaluating men with impotence due to vascular causes. Key words: urogenital diseases-radionuclide imaging Introduction In contrast to morphological investigations, such as ultrasound, MRI and X-ray, radionuc­lide studies of the reproductive system are used to evaluate its function and perfusion, and can be frequently considered as complementary rat­her than competitive to studies mentioned abo­ve. Clinical conditions, where radionuclide inve- This paper was presented in shortened version as a Keynote address at the IAEA/WHO Seminar on Ga­strointestinal and Urogenital Functional Studies with Radionuclides, Vienna, 16-19 August 1993. Correspondence to: Doc. dr. J.Fettich. Clinical Centre Ljubljana, Department of Nuclear Medicine, Zaloška 7, 61000 Ljubljana, Slovenia stigation have proven useful, are acute hemis­crotum, where perfusion and blood pool ima­ging of the testes with pertechnetate is used to evaluate its perfusion, male and female sterility, i. e. detection of varicocele in males and deter­mination of functional fallopian tube patency in females, and studies of penile blood flow in the assessment of male impotence (Table 1). Table 1: Radionuclide studies of the reproductive system. l. Evaluation of acute hemiscrotum Differentiation between torsion and inflammation 2. Evaluation of sterility Male: detection of varicocele Female: functional patency of the fallopian tubes 3. Evaluation of male impotence Diagnosis of arterial insuficiency Diagnosis of venous leakage UDC: 616.6-07;539.163 Acute hemiscrotum The most eommon and most widely aeeepted radionuclide teehnique of investigation of the reproduetive system is perfusion and blood pool imaging of the testes with perteehnetate to evaluate testieular perfusion in the evaluation of aeute hemiserotum, to differentiate between torsion of the testes, whieh is an absolute surgieal emergeney, from other less urgent eon­ditions, sueh as epididymitis, whieh are usually treated eonservatively. Pathophysiology The testis, tightly attaehed to the epididymis, is suspended in the serotal sae by the spermatie eord eomposed of the entering arteries and draining veins, the spermatie duet, lymphatics and nerves. Normally, the testes is eovered partially by tuniea vaginalis and firmly fixed posterolaterally to the epididymis and serotum, what makes torsion of the eord impossible. Two malformations, eomplete investment of the testes with tuniea vaginalis, whieh precludes tight fixation of the testis and epididymis to serotal wall, and elongated mesorehium, i. e. attaehment of the testis to epididymis, permit 2 spermatie eord to twist. 1' Complete torsion will result in infaretion of the testis due to impaired arterial blood supply. Spermatogenie eapability is !ost within a few hours. Ineomplete torsion, where arterial blood supply is still suffieient, will also eause isehemia and infaret due to venous obstruetion, but more slowly. Three or more eomplete turns of the sperma­tie eord are suffieient for irreversible isehemia to oeeur within a few hours, whereas up to 24 hours are neeessary with a single 360 degree twist. Salvage of the testes has been reported after 30 hours with ineomplete twist. 1 Clinical presentation Most eommon eauses of aeute hemiserotum include inflammation ( epididymitis or rarely orehitis), torsion of the testes and trauma, whieh ean also lead to infeetion, as well as torsion. Peak ineidenee of torsion is in adoleseents, around puberty. Patient presents with abrupt, exerueiating pain. Fever and urinary traet symp­toms are absent and urinalysis is normal. He­miserotum is swollen and edematous. Elevation of the serotum will intensify the pain. Typieal patient with epididymitis is young adult, with more gradual onset of pain, whieh is also less intense than in torsion. Dysuria and burning with urination are eommon, as well as moderate to high fever. Elavated number of white eells is present in urine. On physieal examination, hemiserotum is swollen and eryt­hematous. Elevation of the serotum eauses re­lief of pain. 1 Unfortunately, clinieal presentation of these eonditions is often, in up to 50 % of the eases, not suffieiently typieal for torsion to be eonfir­med or excluded with eonfidenee. Without a reliable way to exclude the possibility of testieu­lar torsion in a patient with aeute hemiserotum, immediate surgieal exploration has been reeom­mended to maximize testieular salvage, but at the eost of many unneeessary operation. Tor­sion is easily remedied by detorsion and orehio­pexy, but surgieal exploration in inflammatory eonditions earries a small, but real risk of eomplieations. 3 Therefore, the diffieulty lies not in treatment, but in diagnosis. Diagnostic methods Serotal perfusion seintigraphy, grey seale ultra­sonography, eontinuous-wave and eonventional duplex Doppler imaging and eolour Doppler have been used for this purpose. l. Serotal perfusion seintigraphy. Usually the study is performed with patient positioned supine, with serotum elevated, as close to the gamma eamera as possible. Penis is taped eephalad, and positioning of the sero­tum should be sueh that raphe are in midline, a neeessary eondition for eomparison of perfu­sion of both testicles. 600-800 MBq ( adult dose) of 99 m Te perteeh­ Radionuclide studies of the reproductive system netate are injected as bolus and dynamic study of the first pass is acquired using magnification. Usually two high quality static images, one with and one without markers, using magnification or pinhole collimator, are acquired immediately after the dynamic phase. In some laboratories a Iead shielding interposed between the thighs and scrotum is used for static images.2 Additional processing of the data, such as tirne activity curves over the testicles does not seem to provide any essential additional infor­mation .4 In cases of normal perfusion of the testes dynamic images show only the iliac and femoral vessels with a low grade blush visible over the scrotum. Scintigraphic findings in torsion corespond to viability of testes because the extent of ischemia will vary depending on the duration and degree of twist of the spermatic cord. In the early phase increased activity in the testicular artery up to the site of twist ("nubbin sing")5 is occasionally seen. In the late dynamic and static images there is absent activity in the region of affected testis. Such images are very specific for acute torsion. This phase is seen up to 4---6 hours following complete twist, and can be seen after a longer interval if the twist is incomplete. In this phase surgical salvage is virtually 100 % . Later the testis progresses to infarction and scrotal wall becomes hyperemic, appearing with tirne as increasingly hot rim around the photon deficient testis. Likelihood of testicular salvage approached zero in one study5 when scrotal activity exceeded that of the femoral arteries. Images of so called "mis­sed torsion" are not specific, and have been observed also in a number of other conditions.6• 7, 8 If spontaneous detorsion occurs the study may be normal or show only reactive hyper­emia. 9 In such case immediate emergency sur­gery is not necessary but there is still deffinite indication for elective opera ti on ( orhiopexy) as soon as feasible. In epididymitis epididymal perfusion is in­creased, while in epididymo-orchitis both epidi­dymis and testis show an increased tissue phase activity. Findings in torsion of a testicular appendage are usually normal, or show slight hyperemia as is seen also in mild inflammation or sponta­neous detorsion. In testicular trauma diffuse hyperemia may be seen. In postraumatic torsion photodefficient area of the size of the tesis is seen, and is usually larger than photodeficiency due to testi­cular hematoma or contusion, if presenL 1 2. Conventional grey scale ultrasonography. Grey scale sonography provides high resolu­tion images of the testis and epididymis. How­ever, morphological abnormalities on torsion and epididymo-orchitis are identical, and ultra­sonography provides no information about blood flow .10, 11 Therefore it is not adequate for the evaluation of acute scrotal disorder, but is frequently useful in further diagnostic work­up, after normal blood flow to the testes bas been documented. 3. Doppler ultrasonography. Doppler sonography allows also for evalua­tion of blood flow. Continuous wave Doppler was found to pro­duce false negative results in up to two thirds of cases, because it is difficult to distinguish arterial flow in peritesticular vessels from intra­testicular blood flow with nonimaging equip­ment.12 Duplex Doppler instruments combine pulsed Doppler and gray scale ultrasonography. It was found that this procedure is time consu­ming, operator dependant and not very reliable in acute hemiscrotum, since testicular arteries are difficult to evaluate because of their small 12 size and low flow even in normal testes.9• Colour Doppler ultasound allows simulta­neous real-tiine display of tissue morphology in gray scale and blood flow in colour. Some recent reports have shown the sensitivity in diagnosing testicular torsion to be between 86 and 100 % , and specificity 100 % in small series of patients.9• 12 It is important, that for the assessment of acute hemiscrotum a high fre­quency transducer, 7.5MHz or more is used for high quality results.13 Diagnostic value To be useful, a diagnostic test for acute hemi­scrotum must be immediately available, quick to perform and very sensitive. It should be also noninvasive and simple to perform. The study should be completed as soon as possible after presentation of the patient be­cause viability of the testes can be preserved only if surgery is performed before 4-6 hours after the onset of symptoms. It should not be used in patients with typical symptoms which have to be explored surgically immediately, but to exclude torsion in cases, where conditions other than torsion are suspected. Scrotal perfusion scintigraphy should be re­garded as an evalu_ation of current testicular perfusion rather thart a method to diagnose a specific condition. Properly performed, it has sensitivity and specificity of over 90 % for the presence or absence of intact blood supply to the testes. Normal or moderately increased testicular perfusion does not rule out the possi­bility if incomplete or spontaneously resolved torsion, in which case surgery is also indicated, but usually not as an imminent emergency. Scintigraphy has no use for the evaluation of neonatal torsion, which is virtually the only condition of scrotal swelling at this age. Perfusion scintigraphy is more accurate than grey-scale ultrasonography, continuous wave and conventional duplex Doppler imaging, while some reports indicate that colour Doppler might be as accurte as perfusion scintigraphy. Advantages of colour Doppler include its ability to provide also high resolution images of the testis and epididymis, and absence of any radia­tion to the patient. It is operator dependent and can be difficult to perform properly in a young patient with very painful hemiscrotum. Nevertheless, as with any new technique, a prospective investigation using larger numbers of patients is needed to determine ultimate value and indications for this technique. 12 At present tirne, perfusion scintigraphy is stili the procedure of choice to evaluate testicu­lar perfusion in acute hemiscrotum, as it has the advantage of many years of clinical valida­tion. 14• 15 It has to be available on an emergency basis. Male infertility Varicocele due to varicosities of the pampm1­form plexus is an important cause of surgically treatable male infertility. Left testicular vein is draining into the left renal vein at right angle, with resultant increa­sed pressure, permitting retrograde flow in the left testicular vein. Therefore, varicocele is usually left sided and unilateral. Right sided varicocele may form by extension from the left, further affecting semen quality .16 Diagnostic methods Apart from simple palpation, which may miss small varicoceles, and invasive direct venogra­phy, which can be combined with therapeutic occlusion, a number of noninvasive techniques, including contact scrotal thermography, ultraso­nography and Doppler imaging, and perfusion and blood pool scintigraphy with labelled red blood cells were developed to detect inpalpable, bilateral, or to confirm clinically suspected va­ricoceles. Difference more than 0.3 degrees centigrade on contact scrotal thermography, diameter of pampinifirm plexus veins larger than 3 mm on ultrasonography, evidence of retrograde flow on Doppler ultrasonography, and pooling of labelled red blood cells on scintigraphy, is considered as positive for vari­cocele.17 Scrotal perfusion scintigraphy Red blood cells are labelled in vivo, with patient standing in front of the gamma camera. 99 m Te pertechnetate is injected as bolus and the first pass is recorded as dynamic study. Static blood pool images are acquired usually with and without Valsalva manouver. Increased ra­dioactivity, due to venous blood pooling is seen in the case of varicocele. Diagnostic value Sensitivity of radionuclide investigation has been reported to be 92 % . 18 Similar results are Radionuclide studies of the reproductive system available also for grey scale and Doppler ultra­sonography .17 Scintigraphy using both dynamic first pass and static blood pool images provides useful data on Iocal hemodynamic in the scro­tum. According to some investigators no impro­vement in the semen quality was seen after surgery in patients who showed no scrotal accu­mulation of radioactivity in the dynamic images, while semen quality did improve after surgery in those with patchy tracer uptake allready in the dynamic phase of the study, 19 so scinti­graphy could have also prognostic importance. Female infertility Functional or organic obstruction of fallopian tu bes is one of the causes of female infertility, especially in regions where tuberculosis of the adnexes is not uncommon. Diagnostic methods To assess fallopian tube patency usually contrast histerosalpingography is performed, where con­trast medium is introduced under positive pres­sure intracervically. Alternatively, observation of spilling of me­thylene blue dye from the fimbirated ends of the fallopian tubes into the peritoneum at lapa­roscopy or surgery, after the dye had been introduced into the cervix under low pressure, can be used. Also, a radionuclide method to assess fallo­pian tube patency has been developed.20 Radionuclide hysterosalpingography 99 m Te labelled human serum albumin micros­pheres (37 Mbq suspended in 1 ml of saline) are applied intravaginally directly onto the cervical mucosa. A nonabsorbable tampon is inserted into the vagina after speculum is removed and the patient is imaged 30min, 1 hour and, if necessary, delayed images up to 4 hours la ter can be obtained, using a pinhole collimator with approximately 30 degrees caudal angula­tion to visualise the region posterior to the uterus, which helps to separate ovaries from 21 the uterus.Labelled microspheres normally migrate spontaneously from the vagina to the ovaries, while in the presence of fallopian tube obstruction, fibrosis or Jack of motility, migra­tion does not take place. A fallopian tube is considered patent, if a focus of activity is seen in the area of the adjacent adnex, including outlining of the ovary. Technical and radiopharmacological improve­ments of the method described are possible. Using SPECT technique spatial resolution of the investigation can be improved, and the dose of the radiation to the ovaries decreased. Re­cently a technique of labelling the sperm cells using 99m Te HMPAO for radionuclide hyste­rosalpingography has been described in ani­mals. 22 The radionuclide technique is less invasive than contrast hysterosalpingography or laparo­scopy with methylene blue applied intracervi­cally under pressure. Radiation burden to the ovaries during radio­nuclide hysterosalpingography is lover than in contrast hysterosalpingography. 21 Diagnostic value Both the sensitivity and specificity of radionuc­lide test for the detection of tuba! patency and tu bal obstruction was reported to be over 90 % when correlated with anatomic criteria.23• 24 Due to its high sensitivity and specificity the test is useful not only for the evaluation of female infertility, but also in assessment of surgical reanastomosis or tuboplasty for pre­viously ligated or fibrosed tubes, and, alternati­vely, to confirm the success of tuba! ligation. It seems that radionuclide histerosalpingo­graphy has an important advantage over mor­phological techniques, because it can detect not only organical, but also functional tuba! ob­struction, what can be otherwise achieved only by direct observation of tuba! motility at surge­ry. The radionuclide study may facilitate detec­tion of diseased but patent tubes and provide useful information about tuba! function.23• 24 Because of this, it has been even proposed that failure of migration of the tracer to the ovaries may render patients with anatomically patent tubes eligible for in vitro fertilisation. 25 Male impotence Normal penile erection requires the co-ordina­ted function of neurological, arterial and venous systems and, even if these systems are intact the process may be inhibited by psychological factors. Patophysiology Erection results following penile smooth muscle relaxation. Dilatation of the cavernosal and helicine arteries increases blood flow into lacu­nar spaces within lacunar bodies. Relaxation of the trabecular smooth muscle enables dilatation of lacunar spaces. The systemic arterial blood pressure transmitted through dilated arteries expands the relaxed trabecular walls against the nonyielding, rigid tunica albuginea and com­presses the subtunical venules causing reduction of venous outflow by so called corporeal veno­occlusive mechanism. This reduces venous out­flow and elevates intracavernosal pressure cau­sing erection. The intracavernosal pressure du­ring erection is a result of equilibrium between the perfusion pressure in the cavernosal artery and the resistance to blood outflow through the compressed subtunical venules. In normal erec­tion it approximates systemic arterial blood 26 pressure. The intracavernous mJection of potent smooth muscle vasodilators such as papaverine or phentolamine results in erection in normal subjects, as well as those suffering from neuro­logical and psychological causes. 27 Causes of male impotence are psychological, neurological or organic. It is estimated, that approximately 50 % of men with impotence have an organic etiology and presumably the majority are vascuiar in origin. Normal rigid erection following intracorpo­real injection of papaverine can rule out vascu­Iogenic impotence. Vascular causes of impotence can be either due to reduced arterial supply (arterial insuffi­ciency) of excessive venous outf!ow (venous Ieak). Arterial insufficiency is consequent to atherosclerotic occlusive disease of the hypo­gastric-cavernous arterial bed, while venous leak is due to smooth muscle myopathy or poor compliance of the erectile tissue caused by structural alterations in the fibroelastic compo­nents secondary to vascular risk factors such as hypercholesterolemia, diabetes mellitus, age­ing, previous priapism, surgery or trauma of the penis. The differentation bettween the two is impor­tant for patient management, eg. restorative surgery or implants for arterial insufficiency and ligation of penile vein for venous Jeaking. 28 Diagnostic methods At this tirne there is as yet no universally agreed upon diagnostic algorithm for vascular evaluation of impotent patients. Vascular impo­tence has been evaluated by different radionuc­lide and nonradionuclide techniques, both usually in association with vasodilatation. For the evaluation of arterial insufficiency: l. Selective angiography, which is an invasive technique. Significance of arterial lesions are difficult to assess because no functional infor­ 30 mation is provided.29• 2. Penile/brachial systolic pressure index, which is not always accurate, since Doppler signal is often obtained from dorsal penile ar­tery and therefore does not correlate with erec­ 32 tile function. 31• 3. Duplex scanning, which can measure changes in individual arterial diameters and peak flow velocities. This method is still under evaluation and is operator dependent.31 4. Radionuclide penile pletismography and radionuclide penile blood flow study using 99m Te labelled red blood cells in combination with papaverine test were used to determine peak corporeal flow and volume changes. Peak cor­poreal flow correlates well with arteriaI disease, but does not correlate with venous leaking. 33 For evaiuation of venous Ieakage: 1. Dynamic infusion cavernosometry34 and cavernosography35 during vasodilataion, which are invasive techniques with infusion of large amounts of saline and contrast media intracor­poreally, to determine the flow rate required to mantain erection and to demonstrate opaci­fication of pelvic veins.32 Radionuclide studies of the reproductive system 2. Severa! 133-Xe clearance studies have pro­posed to measure penile blood flow. Subcuta­neous injection was not appropriate since circu­lation of the erectile tissues is largely indepen­dent of penile skin circulation. Cavernosal injection in flaccid penile state was not able to differentiate between normal and impotent men. Also with cavernosal injection and vasodilata­tion it was reported that xenon outflow alone could not be used to predict competence of 36-38 , veno-occlusive mechanism. 29 3. Injection of 99mTc labelled red blood cells intracavernosally, instead of xenon, has been proposed to evaluate venous leakage. In this case 18.5 MBq of in vitro labelled red blood cells are injected intracorporeally twice, in flac­cid state and after vasodilator injection. Data are acquired for 20 min for each study at the rate of 1 frame/15 seconds. Time activity curves are generated and tirne to half outflow (T50 % ) is calculated. It is significantly shortened in case of venous Jeakage as compared with normal venous out­flow if the study is performed after appropriate papaverine vasodilatation. High sens1ttv1ty (100 % ) and specificity (92 % ) for differentia­ting venous leak from normal venous outflow were reported for study after vasodilation in a small group of patients.39 To determine, if arterial insufficiency is also present, this technique has to be combined with another radionuclide or duplex sonography study. 3. To assess both arterial and venous prob. Iems during the same pharmacological interven­tion, penile pletismography and xenon washout have been successfully used simultaneously by 41 , severa! groups.40 Among severa! similar protocols the techni­que used by the Cleveland group is decribed, as an example.41 Red blood cells are labelled in vivo with 200Mbq of 99m Te. The lower abdomen is covered with lead shields leaving only genitalia exposed. A venous blood sample is obtained to convert count rate into mililiters. A butterfly needle is inserted into cavernosal body at the base of the penis. A gamma camera with an all-purpose collimator, set for dual isotope acquisition, is centered over the penis. 200-400 MBq of 133 Xe in 1 ml saline are injected through the buttelily needle and flush­ed with 5 ml of saline. Data acquisition is started after 2-3 min, at the rate of 1 frame/ 10 seconds for 15 minutes. Vasodilators (papaverine 45 mg and phentolamine 1 mg) are injected slowly and flushed with 1 ml saline through the indweling butterfly needle. Data acquisition is continued at the same rate for 20 min. Results are presented as tirne activity curves after correction for technetium downscatter into xenon window. Arterial and venous flows are calculated according to one compartment model from the flaccid through the tumenescent and erect states. (Erection is reduced prior to dis­charge of the patient). Combining the results of arterial and venous blood flow during pharmacologic intervention, this technique was able to differentiate between normals, having normal both peak arterial and peak venous blood flow, arterial insufficiency patients, having decreased peak arterial flow and normal peak venous flow, and patents with venous leak, having essentially normal peak arterial flow and markedly increased peak ve­nous flow in a small group of selected patients.42 Diagnostic value Although not yet properly evaluated, these methods may be useful as a screening test in patients with erectile dysfunction to differen­tiate patients with arterial insufficiency from those with venous leak. The disadvantage of radionuclide methods is their inability to pro­vide anatomic information in comparison with angiographic and duplex sonography, their advantage however is that they are simple to perform, are operator independent, and should be well reproducible, what is useful, especially to follow up the success of surgical treatment of such patients. Conclusion Despite the emergence of new, especially cross sectional imaging approaches, radionuclide Fettich J techniques maintain a significant role in genital imaging, since they can demonstrate the physio­logic status of organs. Radionuclide scrotal imaging remains the most effective method for differentiating bet­ween testicular torsion and epididymitis.9 La­belled red blood cells are used successfully for varicocele detection in infertile men. Only ra­dionuclide techniques allow for noninvasive as­sessment of functional tuba! patency in infertile women. They can play an important and useful 15 role in evaluating men with impotence. References l. Lutzker LG and LS Zukier. Testicular scanning and other applications of radionuclide imaging of the genutal tract. Semin Nucl Med 1990; 20: 159-88. 2. Chen DPC, Holder LE, Melloul M. Radionuclide scrotal imaging: further experience with 210 new patients. I. Anatomy, patophysiology, and met­hods. J Nucl Med 1983; 24: 735-42. 3. Fenner MN, Roszhart DA, Texter JH Jr. Testicu­lar scanning: evaluating the acute scrotum in the clinical setting. Urology 1991; 38: 237-41. 4. Van Ahlen H, Bockisch A, von Staufenberg A, et al. Static and dynamic radionuclide imaging in the diagnosis of the acute scrotum. Uro/ lnt 1991; 47: 20-4. 5. Chen DPC, Holder LE, Melloul M. Radionuclide scrotal imaging: further experience with 210 new patients. II. Results. J Nucl Med 1983; 24 841-53. 6. Vieco PT, Stern J, Herschon S. Patent processus vaginalis. Ciin Nuc/ Med 1990; 15: 263. 7. Dunn EK, Chen PC, Lipschitz SS, et al. Testicular scanning two to three months after torsion: corre­lation with sonography and histopatology. Ciin Nucl Med 1991; 16: 37-9. 8. Melton JW, Chung CJ, Gordon L. Pseudobullseye of the testicle: a window to the peritoneum. Ciin Nuc/ Med 1991; 16: 604--5. 9. Middleton WD, BA Siegel, GL Melson, et al. Acute scrotal disorders: prospective comparison of color Doppler US and testicular scintigraphy. Radiology 1990; 177: 177-81. 10. Chen DPC, Holder LE, Kaplan GN. Correlation of radionuclide imaging and diagnostic ultrasound in scrotal diseases. J Nucl Med 1986; 27: 1774--81. 11. Middleton WD and GL Melson. Testicular ische­mia: color Doppler findings in five patients. AJR 1989; 152 1237-41. 12. Krieger JN, Wang K, Mack L. Preliminary evalua­tion of color Doppler imaging for investigation of intrascrotal pathology. J Urol 1990; 144: 904-7 13. Taylor KYW, S Holland. Doppler ultrasound l. Basic principles, instrumentation and pitfalls. Ra­diology 1990: 174; 297-307. 14. Lerner RM, RA Mevorach, WC Hulbert, et al. Testicular imaging. Curr Opin Radio/ 1991; 3: 694--9. 15. McCartney WH. Radionuclide genital imaging. Uro[ Radio/ 1992; 14: 96-106. 16. Wheatley JK, Fajman W A, Witten FR, et al. Clinical experience with radioisotope varicocele scan as a screening method for detection of subcli­nical varicoceles. J Uro/ 1982; 129: 57-9 17. Kupeli S, Ari kan N, Aydos K, et al. Multiparametric evaluation of testicular atrophy due to varicocele. Urol-lnt; 1991; 46: 189-92. 18. Suga K, Ariyosi I, Nokanishi T, et al. Clinical study of varicocele by sequential scrotal scintigrap­hy. Andrologia 1990; 22: 525-9 19. Hirokawa M, Iwamoto T, Asakura, et al. Clinical study of infertile males with varicocele showing no typical radionuclide blood pooling on dynamic image of scrotal scitigraphy. Urology 1991; 37: 562-7. 20. lturralde M and Venter PF. Hysterosalpingo-ra­dionuclide scintigraphy (HERS). Semin Nucl Med 1981; 11: 301-14. 21. McCalley MG, P Braunstein, S Stone, et al. Radionuclide hysterosalpingography for evalua­tion of fallopian tube patency. J Nuc/ Med 1985; 26: 868-74. 22. Bockisch A. Sperm celi dynamics in female rabbit genital tract after insemination monitored by ra­diolabeled spermatozoa. J Nucl Med 1993; 34: 1134--9. 23. McQueen D, JH McKillop, HW Gray, et al. Investigation of tuba! infertility by radionuclide migration. Hum Reprod 1991; 6: 529-32. 24. Gurgan T, HA Kisnisci, H Yarali, et al. Evalua­tion of the functional status of the fallopian tubes in unexplained infertility with radionuclide hyste­rosalpingography. Gynecol Obstet lnvest 1991; 32: 224--6. 25. Steck T, Wurfel W, Backer W, et al. Serial scintigraphic imaging for visualisation of passive transport process in the human fallopian tube. Hum Reprod 1991; 6: 1186-91. 26. Krane RJ, Goldstein I, Saenz de Tajeda I. Impo­tence. N Engl J Med 1989; 321: 1648-59. 27. Virag R. Intracavernous injection of papaverine for erectile failure. Lancet 1982; II: 938. 28. Sharlip ID. The role of vascular surgery in arterio­genic and combined arteriogenic and venogenic impotence. Semin Uro/ 1990; 8: 129-37. Radionuclide studies of the reproductive system 29. Seftel AD and I Goldstein. Vascular testing for impotence. J Nucl Med 1992; 33: 46-8. 30. Bookstein JJ, Valji K, Parsons L, et al. Penile magnification: pharmacologic angiography details of intracorporeal anatomy. AJR 1987; 148: 883-8. 31. Schwartz AN, Wang KY, Mack LA, et al. Evalua­tion of normal erectile function with color flow Doppler sonography. AJR 1989; 153: 1155-60. 32. Grech P, Dave S, Cunningham DA, et al. Com­bined papaverine test and radionuclide penile blood flow in impotence: method and preliminary results. Br J Uro! 1992; 69: 408-17. 33. Schwartz AN, MM Graham, GF Ferency, et al. Radioisotope penile pletismography: a technique for evaluating corpora cavernosal blood flow du­ring early tumescence. J Nucl Med 1989; 30: 466-73. 34. Friedenberg DH, Berger RE, Chwe DE, et al. Quantitation of corporeal venous outflow resi­stance in man by corporeal pressure flow evalua­tion. J Ural 1987; 138: 533-8. 35. Lue TF, Hricak H, Schmidt RA, et al. Functional evaluation of penile veins by cavernosography in papaverine induced erection. J Uro! 1986; 135: 479-82. 36. Ishigooka M, Irisava C, Watanabe H, et al. Intra­cavernous injection of prostaglandin El: the appli­ cation to cavernosography and penile blood flo. measurement for the diagnosis of venogenic impo­ tence. Ural lnt 1991; 46: 193-6 37. Nseyo UO, HJ Wilbur, SC Klang, et al. Penile xenon-133 washout: a rapid method for screening for vascular impotence. Urology 1984; 23: 31-4. 38. Haden HT, PG Katz, T Mulligan, et al. Penile blood flow by xenon-133 washout. J Nucl Med 1989; 30: 1032-5. 39. Groshar D, S Lidgi, A Frenkel, et al: Radionuc­lide assessment of penile corporal venous leak using technetium-99m labeled red blood cells. J Nucl Med 1992; 33: 49-51. 40. Schwartz AN and MM Graham. Combined tech­netium radioisotope penile pletysmography and xenon washout: a technique for evaluating corpora cavernosal inflow and outflow during early tumes­cence. J Nucl Med 1991; 32: 404-10. 41. Miraldi F, AD Nelson, WT Jones, et al. A dual radioisotope technique for the evaluation of penile blood flow during tumescence. J Nucl Med 1992; 33: 41-6. 42. Kursh ED, WT Jones, S Thompson, et al. A dynamic dual isotope radionuclear method of quantifying penile blood flow. J Ural 1992; 147: 1524-9. Radio! Oncol 1993; 27: 200-4. New prototype 99mTc extraction system Aleš Fajgelj and Jože Novak "Jožef Stefan" Institute, Ljubljana, Slovenia 99 A new semi-automatic extraction system for production of mTc from low specific activity 99 Mo, developed in the Department of Nuclear Chemistry of the "J. Stefan" Institute in Ljubljana is 99 described. mTc is separated from reactor irradiated natura! molybdenum(Vl) oxide by continuous solvent extraction with methyl-ethyl ketone (MEK). The important characteristics and problems related to this new extraction system are discussed. Key words: technetium-99m, technetium-isolation and purification, isotope production, solvent extraction; molybdenum, molybdenum oxide; nuclear reactors Introduction Three decades after the introduction of 99mTc as an "ideal" radiotracer, it still remains the most widely used isotope for nuclear medicine studies. The pre-eminence of 99mTc as a medi­cally useful radionuclide is directly attributable to its excellent physical and chemical proper­ 2 ties.1· Technetium, in the from of 99mTc, is obtained from 99Mo, which is produced by neutron irradiation of molybdenum ( of natura! isotopic composition or enriched in 98Mo); al­ternatively, 99Mo may be obtained as a fission product of uranium. Chromatography, sublima­tion and solvent extraction are three commonest methods used to separate 99mTc from 99Mo.3 •4•5 Correspondence to: Aleš Fajgelj Ph. D., "Jožef Ste­fan" Institute, Nuclear Chemistry Department, Ja­mova 39, P.O. Box 100, 61111 Ljubljana, Slovenia. UDC: 546.718;539.17;546.77 In a nuclear reactor with a relatively low neutron flux (below say 5 X 1013 n.cm-2s-1) the specific activity of 99Mo is such that only thesolvent extraction and sublimation methods can be used, as the mass of molybdenum required to be loaded on to a generator column would be too great. 98Mo (n 'y) . 99Mo 67 h 99mTc . mu. .-y (n, f) (140 keV) x 10s y Ru --. 99Tc 2 Figure l. Production of 99"'Tc by nuclear reactions. 99mTc has been successfully produced at theTRIGA Mark low power research reactor (250 kW) of the "J. Stefan" Institute in Ljubljana for almost ten years. By the use of manually operated solvent extraction system where sepa­ration of 99mTc from 99Mo was based on bubbler extraction, the needs of the two Slovenian largest nuclear medicine institutions, the Uni­versity Clinic for Nuclear. Medicine and the Institute of Oncology has been covered. 6 For the same purposes a new semi-automatic, cen­tralized, computer controlled continuous ex­traction system has been developed and is de­scribed in the present work. Materials and methods Apparatus The purpose of the apparatus is to separate 99mTc from low specific activity 99Mo by solvent extraction, where some steps of the procedure are automated, and finally to obtain a sterile, chemically pure product in a small volume (cca. 10 ml) of physiological solution for direct medi­ 7 ca! use. Block and schematic diagrams are shown in Figures 2 and 3. The main parts of the prototype apparatus are as follows: j§ hot .ir (T„70-Cl !\ 1 1 -------=.vacuum Figure 2. Schematic presentation of the newly develo­ped extraction system. 14 l. Extraction vessel (V = 200 ml) 2. Evaporator 3. Condenser 4. Diaphragm pump 5. Reservoir for MEK 6. Peristaltic pump 7. Three-way tap 8. Self-filling syringe 9. Membrane filter (0.22µm) 10. Sterile vial 11. Valve for 0.9 % NaCl 12. 0.9 % NaCl solution 13. Membrane filter (0.45 µm) 14. Control unit 15. Motor 16. Silica-Gel column 17. Reservoir for TcO4 Figure 3. Block scheme of the extraction system. -extraction vessel (see also Figure 4); -funnel for methyl-ethyl ketone (MEK); the volume of the glass vessel is 50 ml; -reservoir for NaCl (0.9 % NaCl); physiolo­gical solution is kept in commercial sterile flasks; -evaporator; -condenser for MEK: a glass apparatus of 400 ml capacity, connected to the reservoir for MEK; -silica-gel column, maintained in position inside a glass holder; -automatic self-filling and dispersing syrin­ge; -0.22 µm membrane filters (Millipore or Sartori us); -sterile vials (commercial); -Masterflex peristaltic pump with flow rates of 1 -20 ml per minute; MEK Solution of MoO3 in 6N NaOH 1. ::=:==---c- Org. phase s (1) CfJ ro .C:: Special stirrer o.. ro 4-< o (1) v. Drain flJ Figure 4. Schematic drawing of new prototype extrac­tion vessel. -vacuum line maintained by KOMV AK 1 diaphragm pump; -control unit: this mainatains control over the peristaltic and diaphragm pumps, the mixing of MEK in the generator vessel, the tempera­ture in the evaporator, the dispensing of NaCl (0.9 % ), and determines the sequence and ti­ming of the operations. The tirne sequence of operations is shown in Figure 5. The whole procedure lasts 45 minutes. At the start the mixer and heater of the evapo­rator spiral are switched on. After 5 minutes the peristaltic pump for MEK and the mem­brane pump, which carries over the MEK va­pour into the condenser, are switched on. After 35 minutes the extraction is complete and the control unit switches off the heater of the evaporator spiral, the mixer for dispersion of MEK in the extraction vessel and the peristaltic pump. The membrane pump continues to ope­rate another 3 minutes in order to remove ali the remaining MEK vapour from the system. In the 3sth minue the valve opens to allow 6 ml 0.9 % NaCl to enter the system, thus was­hing down sodium pertechnetate from the eva­ 39th porator spiral to the reservoir. In the minute the membrane pump is switched on for one minute to suck ali remaining NaCl solution into the reservoir. In the 40th minute the dispen­ser is switched on to transfer the sodium pertechnetate solution in physiological saline into a sterile syringe. In the 41st minute the dispenser forces the solution from the syringe via a three-way tap and 0.22 µm membrane filter into a sterile penicilin bottle. Ali parts of the generator system are directly connected via glass, teflon or PE tubes. The characteristic of the apparatus developed are as follows: -dimension of the apparatus: 80 x 80 x 80 cm; -weight: 35 kg; -voltage: 220 V; -cooling system: water cooled; -preparation tirne: 45 minutes. Materials -MoO3, p.a., Merek, is used for irradiation; -NaOH, p.a., Kemika, Zagreb, 6 M solu­ tion; -MEK (methyl-ethy ketone), p.a., Merek or Kemika, Zagreb, is used for solvent extrac­tion separation of 99 mTc from 99Mo in an alka­line solution of Mo(VI) and Tc(VII), distillation before use is required; TIME (mini -H2O2 , p.a., Belinka, Ljubljana, 3 % solu­tion; -silica-gel,' Kemika, Zagreb, for chromato­graphy, particle size 0.2 -0.5mm, washed three times with MEK and then ten times with redestilled water to remove fines, and dried at 105 °C, is used for removal of small amounts of water phase which contains dissolved 99Mo; -physiological solution (0.9 % NaCl) -steri­le. lrradiation 120 g of MoO3 are irradiated in an aluminium container in our TRIGA Mark II research reactor. The optimal irradiation tiine of MoO3 for continuous production on a three day cycle with our previous generator system,6 determi­ned by experirnent, is three weeks in the F-ring, 1012 21 at a neutron flux of 4 X n.cm-s-, and finally 48 hours in the central channel at a 1013 21 neutron flux of 1.1 x n.cm-s-. This activity is equivalent to the activity of five 800mCi commercial generators. By (n, y) reac­tions, 99Mo (t112 = 66 h) and 101Mo (tl/2 = 14.6min) are formed. Beta decay of these two isotopes gives 99mTc (tl/2 = 6.02 h) and 101Tc (t112 = 14.2 min). 101Tc in the fina) product could interfere with radiological investigations, so the irradiated MoO3 is cooled for some hours to allow it to decay away. 2 3 4 5 6 7 Extraction In the hot celi, 25 g of irradiated MoO3 are dissolved in 200 ml of 6 M NaOH, and 1 ml of 3 % H2O2 is added to oxidize Mo reduced during irradiation. The solution is transfered to the extraction vessel after 10 min vigorous stiring. The new smaller scale prototype extrac­tion vessel allows 200 ml of MEK as organic phase to disperse through the alkaline solution in a single continuous pass. The MEK is disper­sed by means of a special stirrer and is siphoned off above the leve) of the aqueous phase. For the experiments an extraction vessel made of pyrex glass was used. With this new system the yield of the extrac­tion is around 85 % (see Table 1), which is very close to the yield of the previous, twice repeated bubbler extraction. Extraction without disper­sal, or by bubling as in the previous system did not give good results. Table l. Yield of extraction. Extraction wi th Without Twice repeated dispersal, single pass dispersal bubbler extraction, previous system6 86.2% 5.4% 87.1 % 85.1 % 7.1 % 82.8% 83.7% 8.3% 86.3% The extraction yield is very dependent on the chemical form of technetium, which has to be in the pertechnetate form. We compared two methods of oxidation, the first being addition of 3 % H2O2, and the second addition of a small amount of K2Cr2O7 . The oxidation efficiency was the same in both cases. For our new generator system oxidation with 3 % H2O2 was chosen, as the possibility of contamination of the fina! product with chromium was elimina­ted. Since in the new extraction system continuous extraction is used, we constructed a new conti­nuous evaporator, which is made from a pyrex glass spiral and heated with an electrical heating tape or hot air to 80 °C. The bottom part of this evaporator is directly connected with the reservoir for TcO4 and the condenser for MEK, where the vacuum is maintained by a KOM­V AK 1 diaphragm pump. Results and discussion The fina! product, solvent extraction produced 99mTcO4 has the following chemical characteri­stics: -pH = 6 -7; -99mTc: 99 % as pertechnetate; -99Mo content: < 1 x 10-5 % ; -specific activity: 4810 MBq .mi-1 (130 mCi. ml-1); -volume: 5 -10 ml; -trace impurities: 95Nb, 188Re, 198Au. In quality tests of the fina! technetium solu­tion we found a measurable activity of 186Re when MoO3 from Kemika Zagreb was used for preparation of technetium. At the same tirne, we also found that the chemical purity of MoO3 produced by Merek is much better, and nor­mally this oxide is used for irradiation. How­ever, if for technical reasons we are obliged to use MoO3 from Kemika Zagreb, there is possi­ 186Re 188Re bility to avoid and activities by precrystalization of MoO3 . Tests carried out at the Microbiological Insti­tute of the Medica! Faculty in Ljubljana and LEK -Ljubljana showed the product to be sterile and apyrogenic, and non-toxic, respecti­vely. The apparatus described is for objective rea­sons not yet used for daily routine production of technetium, but ali the tests performed sho­wed that the fina! product has the same charac­teristics as TcO4 produced by our previously d_eveloped system.8 The main part of the new semi-automatic extraction system -the extrac­tion vessel -will in future be changed from glass to a stainless steel extractor to avoid the problems which can arise from Ieaching of silica from the walls under the combined influence of intense gamma radiation and strong alkali, leading blockage of the membrane filters. The incorporation of additional electronic controls and connection to the micro-computer should make this system more suitable for simple daily routine work. Acknowledgement The authors wish to thank the International Atomic Energy agency for financial support of this work, which is a part of the IAEA Coordi­nated Programme: "Development of 99mTc ge­nerators using low power research reactors". References l. Richards P, Tucker WD, Srivastava BC. Techne­tium-99m: Historical perspective. lnt J Appl Radiat Isotop 1982; 33: 793-9. 2. Anders E. The radiochemistry of technetium. Chica­go: U. S. Atomic Energy Commission, 1960. 3. Boyd RE. Technetium-99m generators -The avai­lable options. Int J Appl Radiat lsotop 1982; 33: 801-9. 4. Zsinka L. 99mTc sublimation generators. Internatio­nal Atomic Energy Agency: Seminar on radionuc­lide generator technology, Vienna: IAEA-SR-131/ 20, 1986. 5. Svoboda K. Solvent extraction 99mTc generator. International Atomic Energy Agency: Seminar on radionuclide generator technology. Vienna: IAEA­SR-131/21, 1986. 6. Novak J, Fajgelj A. 99"'Tc extraction generator. Vestn Slov Kem Drus 1983; 30: 1-7. 7. Fajgelj A, Novak J, Stegnar P, Konda D. New prototype Tc-99m extraction system. Report on IAEA Research Coordination Meeting: Develop­ment of Tc-99m generators using low power research reactors. Bandung, lndonesia, 1987. 8. Pauwels EKJ, Feitsma RIJ. Radiochemical quality control of 99"'Tc-labeled radiopharmaceuticals. Eur J Nucl Med 1977; 2: 97-103. Glomerulonephritis and cancer Neil G. Burnet,1 Giel M. Sadler,1 M. H. Griffiths2 and Gilliah M. Duchesne1 1 Meyerstein Institute of Clinical Oncology, Middlesex Hospital, Mortimer Street, London, WIN 8AA; and 2 Histopathology Department, University College Hospital Medica! School, University Street, London, WCIE 611, United Kingdom Glomerular disease occurring in association with cancer is a rare phenomenon, but it is important because it may precede the appearance, or recurrence, of cancer. The true significance, and further management of a patient, depend on just how rare the association is, and what can be done to treat the tumour. In the earliest report of the association, the incidence of cancer occurring in patients with glomerulonephritis was 11 % . This strikingly high figure had a great influence on clinical thinking, but subsequent papers have suggested that the incidence is much lower, perhaps less than 1%. Three clinical cases are presented which demonstrate the practical management problem of oncology patients who develop glomerulonephritis. Together with a review of the literature, these cases illustrate the relative lack of value of extensive re-investigation of oncology patients who develop glomerulo­nephritis. Key words: glomerulonephritis; neoplasms, cancer Introduction Glomerular disease occurring in association with cancer is a rare phenomenon in oncology practice, though is more common, and better described, in nephrology patients. However, its importance lies in the fact that nephrotic syn­drome may precede or presage the presentation of a new cancer, or relapse of a previous malignancy. In addition, nephrotic syndrome Correspondence to Dr. Neil G. Burnet, MD, FRCS, FRCR, Meyerstein Institute of Clinical Oncology, Middlesex Hospital, Mortimer Street, London, WlN 8AA, United Kingdom. UDC: 616.611-002:616.61-006.6 may be significantly morbid because of the substantial loss of protein which can occur. Just how important the development of nephrotic syndrome is, and the further management which is required, depend on the true incidence of the association between glomerular disease and cancer, and on what can be done to treat both the renal and turno ur pathology. The commo­nest clinical presentation of tumour-associated glomerular disease is nephrotic syndrome. This can be regarded as severe proteinuria, and is defined specifically as severe proteinuria (>5 g/ 24hrs), with hypoalbuminaemia (<20g/l) and peripheral oedema. The first report of malignancy and glomerular disease dates to 1922 when Galloway described a patient with Hodgkin's Disease who develop­ Burnet NG et al. ed nephrotic syndrome.1 However, the associa­tion was probably unrecognised until the semi­nal paper by Lee et al. in 1966.2 These authors reported 11 cases of cancer in 101 patients with nephrotic syndrome. This series of patients very reasonably had a great influence on clinical thinking, chiefly because of the very high inci­dence -11 % -of cancer occuring in patients with glomerulonephritis (GN). Subsequent pa­pers however, have suggested that the relation­ship is less strong, cancer occurring in a smaller proportion of nephrotic syndrome patients (see below). Glomerular disease is known to occur in both lymphomas and solid tumours (Table 1). It has been described occurring as a prodro- Table l. Tumours most commonly associated with paraneoplastic glomerulonephritis. Carcinoma of the bronchus -notable Carcinoma of the colon Carcinoma of the stomach Rena! celi carcinoma Hodgkin's Disease -uncommon Non-Hodgkin's Lymphoma -very rare mal illness preceding the appearance of tumour; it may occur in patients with established tu­mours, though usually early on; and it may presage recurrence. Glomerulonephritis occurr­ing in patients known to have tumours can " j Figure la. Rena! biopsy from Case 1 showing membra­nous glomerulonephritis. The glomerular basement membrane appears as a chain-like structure, the silver­positive (black) membrane enclosing the lucent com­plexes shown in b) (silver methenamine stain x 400 original magnification). therefore present an anxious and difficult mana­gement problem for the oncologist. Patients with nephrotic syndrome and no previous history of malignancy normally present to renal physicians, and this scenario is there­fore less relevant in oncology. What, then, should the oncologist do when faced with a patient with nephrotic syndrome? Three clinical cases are presented to illustrate the association between GN and cancer and to demonstrate some of the clinical problems which arise from it. With reference to the literature, we hope to provide some guide to the management of these patients. Case reports Case 1 A 44 year old painter and decorator, a life-long smoker, presented in 1983 with asthma. At that tirne he was found to be hypertensive and had significant proteinuria, with a concentration in the urine of 4,3 grams/litre. His plasma albumin was 32 g/l and creatinine clearance 125 ml/min. He was treated for hypertension with nifedipine and diuretics, but was a rather erratic hospital attender. In 1990, his blood pressure was found to be 200/140, his albumin had fallen to 30 g/1, creatinine clearance had worsened to 94 ml/min, and his proteinuria was 6,3 g/24 hours. His ECG showed changes consistent with hyperten- .. _ .. .- . --,' i. 'f ·. ,y ..;. ,, • -,, { 1 • .... t t • ........ .... :.1. \,.· •1 . 1\ JI" \ •·"' .'-· II ,I' , . 4 ••"' "' I 1 • ., , . • Figure lb. Rena! biopsy form Case 1 showing membra­nous GN. An immunoperoxidase preparation for IgM, demonstrating granul ar deposits (black) outlining the capillary loops (immunoperoxidase IgM x 250 original magnification). sion. The following year he agreed to renal biopsy and was found to have membranous GN (Figure 1). Treatment for his hypertension was continued. In September 1992, he presented with a 2 week history of shortness of breath and haemo­ptysis and was found on chest X-ray to have right upper lobe collapse. Bronchoscopy sho­wed a tumour occluding the orifice of the right upper lobe and extending into the right main bronchus. Biopsy revealed poorly differentiated, nonsmall celi carcinoma of the bronchus which was inoperable due to its proximity to the carina. A course of palliative radiotherapy was carried out, giving 30 Gy in 10 fractions over 2 weeks to the right upper lobe mass and mediastinum. This was tolerated well, with re­solution of the haemoptysis and shortness of breath. One month after the radiotherapy, chest X-ray demonstrated that the right upper lobe had re-expanded, he remained mildly hyperten­sive on treatment, and his renal function was unchanged. Seven months later he remained clinically well, with the same degree of renal impairment, and continuing proteinuria of 6 g/ 24 hours. Case 2 A 50 year old nursing sister presented with a lump in the left breast in 1989. At that tirne, she was pre-menopausal. Wide local excision was carried out and histology revealed a patho­logical T2 grade 2 invasive dueta! carcinoma of the breast though some tumour remained at one excision margin. She was clinically node negative. Routine staging was otherwise normal and she was started on Tamoxifen. Radical radiotherapy was given to the whole breast using 50 Gy in 25 fractions in 5 weeks with tangential opposed fields, followed by a boost to the tumour bed of 16 Gy in 8 fractions over 10 days. Three years later she was found by chance to have a plasma albumin of 27 g/1, but was not investigated further, until a few months later when she complained of tiredness and malaise and had pitting oedema to the mid thigh level and over the sacrum. There was, however, no clinical evidence of recurrent car­cinoma of the breast. At the tirne her albumin was 24 g/1 and urinary protein 1.26 grams per litre. The serum cholesterol was slightly eleva­ted at 10.7mmol/1 (normal range 2.3 -6.9). Rena! biopsy demonstrated membranous GN. Shortly after, she became mildly hypertensive and was started on enalapril and frusemide. This treatment improved her blood pressure and her oedema resolved. Though her plasma albumin remained low, the protein loss from the kidneys was reduced to below half of its former leve!. She currently remains free of clinical evidence of carcinoma of the breast. She has not been formally re-staged. Case 3 In 1983, a 50 year old university lecturer presen­ted with night sweats, weight loss of 3 kg and lymphadenopathy in the neck and axillae. Biopsy demonstrated high grade 11011-Hodgkin's lymphoma (NHL) and further investigations revealed Stage 2B disease. He was treated with 6 cycles of chemotherapy, using cyclophosph­amide, doxorubicin, vincristine and predniso­lone (CHOP). After 2 cycles he was found to be in complete remission clinically. For the ensuing 10 years he remained well but in February 1993 he presented with a 10 day history of feeling unwell, and a 24 hour history of periorbital oedema, and swelling of the legs up to the level of the knees. There was no clinical evidence of recurrence of the lym­phoma. He had normal renal function but his plasma albumin was slightly low, at 33 g/1. Over the next few days his albumin fell to 22 g/1, and urine collection demonstrated proteinuria of 10.4 g/24 hours. Serum cholesterol was elevated at 13.4 mmol/1, in physiological response to the low albumin. Other investigations including chest X-ray and ultrasound of the abdomen, pelvis and renal tract were normal. A clinical diagnosis of glomerulonephritis was made and renal biopsy showed the glomerular tip lesion, a variant of focal segmenta! glomerulosclerosis (FSGS) (See Figure 2). This variant of FSGS Burnet NG et al. described by Beaman et al. in 1987 may be steroid responsive,3 although 2 patients in this series who did not respond to steroids had malignant disease, renal celi carcinoma and Hodgkin's disease respectively. Our patient was treated initially with frusemide, and following the histology result cyclophosphamide (150 mg = 2.5 mg/kg) and prednisolone ( 60 mg/ day) were added. He developed occasional fe­vers and night sweats, though no clinical evi­dence of recurrence. Re-staging was carried out with CT of thorax, abdomen and pelvis, plus bone marrow trephine and aspirate, ali of which were normal. Over the next 3 months his general condition improved and the peripheral oedema was con­fined to the ankles. The glomerulonephritis also improved, so that by June 1993 urinary protein loss had fallen to 3 g/24 hours. Plasma albumin, thought not to normal, had risen to 29 g/1. Serum cholesterol had dropped somew­hat, to 9. 7 mmol/1. During this tirne the steroid dose had been reduced to 20 mg/day, though he remained on cyclophosphamide. Discussion Incidence of cancer in patients with glomerulo­nephritis The classic paper by Lee et al. in 1966 reported a relatively high incidence of 11 % (11 of 101) of patients with nephrotic syndrome who deve­loped carcinomas.2 In 7 of these 11 patients the glomerulonephritis preceded discovery of can­cer, and ali were over 40. Other series have reported rates between 6 % and 11 % .4• 5 The two series with these high rates of associated malignancy, in particular, have been invoked to justify the search for malignancy in patients with nephrotic syndrome, particularly with membranous GN histology (see below).6 How­ever, Kaplan et al.7 reviewed 14 published series including 1643 patients with nephrotic syndrome, though not those of Lee et al.2 and found that only 6 had co-existing tumours, an incidence of only 0.37 % . In another individual series, Heneghan et al. reported a 1.1 % inci- Figure 2. Rena! biopsy from Case 3 showing the glomerular tip lesion. There is a segmenta! adhesion to the origin of the proximal convoluted tubule asso­ciated with a collection of foam cells in the affected segment (silver methenamine stain x 250 original mag­nification). dence of cancer with GN.8 Carcinomas are the tumours most commonly associated with GN (Table 1). Hodgkin's Disease is uncommon, and non-Hodgkin's lymphoma extremely rare.6• 9-n Benign tumours have occasionally been reported in association with GN, but so rarely that this is probably due to chance alo­ 7, 11 ne_ The incidence of the association is also de­pendent on age. Cancer associated with glome­rulonephritis is very rare in children. There were no cases in one series of 121 children with membranous GN, although there have been occasional patients reported in other series.11 The incidence appears to rise in adulthood, and most of the reported cases have been over 40-50 years of age.11 Thus, it would appear, especially from more recent series, that the true incidence of cancer occuring in patients with nephrotic syndrome is much Iower than initial reports suggested. This changes the significance of nephrotic syndrome, both in a patient who is otherwise well, and in one who has a past history of malignancy. White a true incidence of cancer developing in patients with nephrotic syndrome of around 10 % might well encourage a relatively stre­nuous diagnostic activity, if the true incidence is as low as .1 % , this may not be warranted, both in terms of psychological cost to the pa­ tient and financial cost to the hospital. Never­theless, since the appearance of nephrotic syn­drome may be the harbinger of new or recurrent cancer, some index of suspicion is appropriate. lncidence of glomerulonephritis in patients with cancer Considering ali the patients seen by oncologists, it is clear that the occurrence of nephrotic syndrome in patients with cancer is strikingly low, although the exact prevalence is unknown. The incidence of sub-clinical renal abnormali­ties in patients with cancer is, however, quite considerable. Surveys of cancer patients have shown proteinuria or haematuria in between 15 % and 58 % of patients.11' 12 This seems a very high proportion, but is entirely possible that in many of these patients this is not asso­ciated with glomerural pathology. Such abnor­malities may be discovered simply because these patients tend to be supervised quite intensively. However, patients with active malignant disease who have sub-clinical proteinuria are unlikely to be subjected to intensive renal investigation including biopsy. The relationship of proteinu­ria to malignancy is probably less strong than these figures suggest. For example, in one sur­vey some degree of proteinuria was found in 22 % of hospital patients who were not suffering from malignancy.12 The incidence of GN in patients with lymphomas, as opposed to carci­nomas, has been addressed in 2 large studies, together including a total of 1700 patients with Hodgkin's Disease. Only 7 patients had evi-. dence of GN, an incidence of 0.4%.8, lO Autopsy studies have demonstrated an inci­dence of immune deposits in the kidney of 11 to 40 % , but actual glomerular changes are rare, only 1.5 % in one autopsy study. 6, 11 This latter figure is compatible with figures from the larger series of clinical GN. The pathological finding of mesangial and subendothelial depo­sits may represent non-specific trapping of im­mune complexes, or more likely tumour anti­gens, rather than indicating, or indeed causing, glomerular damage. 6 Evidence of a causal relationship between cancer and glomerulonephritis Although the association is rare, there is reaso­nable evidence of a causal relationship. The main evidence falls into 3 main categories. Firstly, a close tempom! relationship frequently exists between cancer and the GN .13 Typically, both conditions occur close together in tirne. Certainly in the majority of patients GN is apparent within 6 months either side of the appearance of the turno ur, although in one series 8-16 % of patients had greater than 1 year between the appearance of the 2 condi­tions.13 Secondly, remission of the GN is fre­quently associated with treatment of the tu­mour, and recurrence of tumour has been de­ 914 5 scribed associated with GN.2' , , Finally, antigen-antibody complexes have been found in the glomeruli, which in a minority of cases have been demonstrated to be tumour-associated an­tigen and specific antibody.6' 11 Non-Hodgkin's lymphoma is rarely associated with GN, which has been invoked as evidence against a causal relationship with this malignancy.6 Renal abnormalities in patients with malignancy Malignant disease may affect the kidney in a vari.ety of ways, directly and indirectly, at both the macroscopic and microscopic leve!. These Table 2. Causes of renal abnormalities in patients with malignancy. Direct infiltration of kidney by tumour Obstruction of urinary tract by tumour Rena! vein thrombosis due to tumour Metastases Fluid imbalance (i.e. pre-renal failure) Infection Electrolyte disturbance caused by tumour or treatment (e.g. hypercalcaemia, raised uric acid, tumour lysis syndrome) Direct toxicity of therapy (chemotherapy, radiotherapy, antibiotics, etc.) Obstruction by tumour products (e.g. light chain casts in myeloma, mucoprotein from carcinoma of pancreas) Glomerulopathy Burne/ NG et al. are laid out in Table 2. The only types of abnormality which can be considered paraneo­plastic are obstruction by tumour products and glomerulopathy, and it is clear from the table that these causes are uncommon. To some extent, in the context of glomerulonephritis causing nephrotic syndrome, the term paraneo­plastic implies uncertainty about the exact aetio­logy of the condition. Of the direct effects, by far the most common is obstruction of the urinary tract, typically by pelvic and retroperi­toneal tumours. Metastatic spread to the kidney . is curiously rare, considering the very large renal blood flow. Types of glomerular disease: The main types of glomerular disease associated with malignancy are shown in Table 3. By far the most common variety is membranous GN. There are differen­ces in the types of GN found in association with carcinomas as opposed to lymphomas. Of all carcinoma patients who develop nephrotic syn­drome, 80-90 % will have membranous GN on 16, 17 biopsy . The histological appearances in these patients are identical to those of membra­nous GN of idiopathic type, with sub-epithelial electron dense deposits which usually prave to be deposits of IgG with or without complement component C3.11 Of the remaining cases, minimal change GN is the next most common type of glomerular Table 3. Glomerular changes most commonly associated with malignancy. Membranous GN -commonest type of GN associated with cancer -typically occurs with carcinoma Minimal change GN -usually associated with Hodgkin's Disease Proliferative GN Mesangiocapillary GN Crescentic nephritis Focal segmenta! glomerulosclerosis (FSGS) IgA nephropathy Amyloidisis pathology. This type is classically associated with lymphomas, particularly Hodgkin's Disea­se, and is uncommon· in patients with solid tumours.11 It appears very rare for renal disease to be associated with non-Hodgkin's lymphoma (NHL). Alpers and Contran6 reviewed the lite­rature and found only 20 cases associating NHL with any clinicalJy significant glomerular disea­se, and only 2 of these actually had minimal change GN. They suggest that this casts doubt on the existence of a clear-cut association bet­ween NHL and GN. Occasional cases of lymphoma have been shown to exhibit focal segfmental glomerulo­ 18 sclerosis (FS GS).6' A few cases of prolifera­tive GN are also seen, generalJy also associated with Hodgkin's Disease,19 and crescentic GN has also been associated with tumours.6 Pathogenesis: Tumour-associated glomerulo­nephritis is regarded as a paraneoplastic syndro­me, which implies that the pathogenesis is not understood. It seems most hkely that glomeru­lar damage is immunologically mediated, al­though disseminated intravascular coagulation may also be responsible (see Table 4). The condition is probably due to immune complexes which may include tumour-associated antigens in some cases, but in very few patients is the actual antigen identified. The prevailing view now is that immune complexes are most likely to be formed in situ by antibody and antigen deposited separately within the glomerular ap­ 6' 11 paratus. It is also trne that immune comple­xes may be demonstrated in the kidney which retains normal function and normal histology (see above). In Hodgkin's Disease it has been postulated that the development of GN may be related to deficient T-cell function. Long-lasting abnorma­lities of cell-mediated immunity occur in many patients with this condition, with reduction of the ratio of T4 (helper) to T8 (suppressor) Table 4. Pathogenesis of glomerulonephritis associated with malignancy. Tumour-associated antigen Re-expressed foetal antigen Vira! antigen Autologous non-tumour antigen Disseminated intravascular coagulation Amyloid (Defective T-cell function) cells. ll, 20 This change has been reported in patients with cured Hodgkin's Disease who subsequently developed GN.21 Whether this T-cell abnormality is the cause of GN in Hod­gkin's Disease is unknown, but it remains a possiblity, since an immunological contribution to the pathogenesis of GN is likely. Prognosis In general, the outcome of patients with cancer­associated GN is dominated by the prognosis of the malignancy, rather than the renal disease. There are rather few reported cases of clinical outcome and no clear picture emerges. There is, however, an appreciable number of cases of improvement in GN associated with response to anti-cancer treatments. This applies particu­larly to Hodgkin's Disease, where treatment of the malignancy is normally accompanied by improvement in the GN.7• 11 For example, a 13 year old boy with Hodgkin's disease who was found to have nephrotic syndrome at diagnosis was treated with mustine. The tumour respon­ded and the nephrotic syndrome resolved within 2-3 weeks. The Hodgkin's Disease relapsed after 3 years but there was no return of the nephrotic syndrome. 15 Glomerulonephritis as­sociated with solid tumours may also remit with treatment of the cancer. Cantre1114 reported a patient with carcinoma of the stomach who had GN at the time of diagnosis. Complete surgical resection was carried out with resolution of the GN. Lee et al.2 also described a patient whose GN "led to the discovery of carcinoma of the colon", and in whom the GN improved after resection of the primary tumour. Where GN has improved with treatment of the tumour, recurrence of the malignancy has often, though not invariably, been associated with recrudes­cence of the GN_ll The leve! of proteinuria may be a prognostic factor. In cancer patients with GN, Row et al. 5 found a worse prognosis in patients with greater proteinuria. Sawyer et al. 12 extended this, com­paring prognosis with the leve! of proteinuria in 504 patients, regardless of whether they had GN or not. The median survival of patients with proteinuria was only 4.5 months, compared to 10 months for patients without. In addition, the higher the leve) of proteinuria, the worse the prognosis. The oncology details were not presented and the low median survival implies that the patients in this study had advanced disease. It is quite possible that patients with aggressive disease were more likely to sustain renal complications causing proteinuria, both from GN and other causes such as infection, so that the proteinuria may be purely a marker, and not a cause, of poorer prognosis. Case discussion Patient 1 developed nephrotic syndromt: severa! years before presenting with carcinoma of the bronchus. Although nephrotic syndrome may precede the appearance of cancer by a surprisin­gly long time, the interval in this case exceeds what can be regarded as likely for a causal association. Whilst he had the commonest va­riety of tumour-associated GN, the majority of patients presenting to renal physicians with membranous GN never develop cancer. No change was observed in the patient's renal status with the development or treatment of the cancer. Case 2 illustrates the classical dilemma, which leads to a difficult mamagement problem. Al­though the patient had early carcinoma of the breast, she stili has a significant risk of relapse. The development of nephrotic syndrome, due to membranous GN, might presage recurrent disease. Is re-staging or further investigation warranted or necessary in this situation? The cost of re-staging this patient is shown in Table 5. The financial cost is relatively low, a total of f 184 at our hospital, but the psycho- Table 5. Cost of re-staging Patient 2. Investigation Cost in pounds sterling FBC s Biochemistry 13 CXR 8 Mammogram 19 Ultrasound Scan Abdomen 19 Bone Scan 120 TOTAL f184 Psychological cost Not assessable ! Burnet NG et al. logical cost is impossible to assess. Since it may take in excess of 1 year after the appearance of nephrotic syndrome for tumour to present, this patient could be faced with a long period of uncertainty about her carcinoma if re-staged. Note that for this patient the more expensive radiological investigations (CT of 1 region, without contrast f90; and MRI of 1 region f300) are not required. The cost of re-staging patient 3, including CT and bone marrow investiga­tions, would have been approximately three times as expensive. Part of the answer to the question of whether to re-stage depends on whether a tumour recur­rence detected earlier yields a higher chance of cure or long-term remission. At present, recur­rent carcinoma of the breast is incurable, and nothing is gained by earlier detection. Regretta­bly, this is true for most solid tumours at the present tirne. This fact considerably weakens the argument for extensive reinvestigation of patients developing nephrotic syndrome, as op­posed to simple, non-invasive routine tests, such as haematology, biochemistry and CXR. One factor which militates in favour of re-sta­ging is the severity of th nephrotic syndrome. Since this may respond to treatment of the malignancy, in some cases of association, inve­stigation in the hope of this may be appropritate if the severity of the nephrotic syndrome is extreme. In case 3, the duration of follow up exceeds 9 years from the completion of treatment. In the case of high grade non-Hodgkin's lympho­ma, this should render the risk of recurrence extremely low. Literature review also suggests that a causal association between nephrotic syndrome and NHL is unlikely. On two counts, therefore, recurrence of disease can be clinically excluded in this patient. Conclusions An association between nephropathy and malig­nancy is well established. However, the true incidence is probably much lower than was once believed, perhaps 1 % or less. In addition, there are circumstances in which a causal rela­tionship is unlikely. Por example, GN is very rare in association with NHL, suggesting a Jack of causality. The clinical behaviour of a particu­lar tumour type may also make the appearance of GN unlikely to be due to relapse. In spite of the clear association between glomerulonephritis and cancer, the vast majo­rity of patients with nephrotic syndrome never develop a malignancy. In patients with malig­nant disease, the incidence of glomerulonephri­tis is extremely low. The glomerulonephritis may act as a marker of disease activity, but in the majority probably does not alter the prog­nosis, provided that the renal condition is cor­rectly treated. Nevertheless, it seems prudent to carry out simple screening investigations in patients who develop nephrotic syndrome de novo, especially in patients over 50 years of age. Por patients with a past history of malig­nancy, the need or desirability of investigation should probably be assessed on an individual basis, rather than managed with a standard policy. Regrettably, in most tumours, recur­rence is usually not curable even if discovered early, and there may be little to gain, and much to lose from over-zealous re-investigation. Acknowledgements This paper is based on a clinical presentation. We are grateful to Dr Viljem Kovac for encou­ragement to commit the presentation to paper, and to Professor GH Neild and Professor DC Linch for suggestions and comments on the presentation. The review of Davison and Thom­son has been most helpful, and contains a fuller reference list on a number of points relating to this topic. References l. Galloway J. Remarks on Hodgkin's Disease. Brit Med J 1922; 2: 1201-4. 2. Lee JC, Yamuchi H and Hopper J Jr. The associa­tion of cancer and the nephrotic syndrome. Ann Intern Med 1966; 64: 41-51. 3. Beaman M, Howie AJ, Hardwicke J, Michael J and Adu D. The glomerular tip Iesion: a steroid responsive nephrotic syndrome. Clin Nephrol 1987; 27: 217-21. 4. Hopper J Jr. Tumour-related renal lesions. Ann lntern Med 1974; 81: 550--1. 5. Row PG, Cameron JS, Turner DR, Evans DJ, Ogg CS, Chantler C and Brown CB. Membranous nephropathy. Long-term follow-up and association with neoplasia. QJ Med 1975. 44: 207-39. 6. Alpers CE and Cotran RS. Neoplasia and glome­rular injury. Kidney lnternational 1986; 30: 465-73. 7. Kaplan BS, Klassen J and Gault MH. Glomerular injury in patients with neoplasia. Annual Review of Medicine 1976; 27: 117-25. 8. Heneghan W, Rao TKS, Nicastri AD and Fried­man EA. The incidence of malignancy associated with nephrotic syndrome in the elderly. Am Soc Nephrol 13th annual meeting, Washington DC, 1980, p 20A. 9. Plager J and Stutzman L. Acute nephrotic syn­drome as a manifestation of active Hodgkin's Disease. Am J Med 1971; 50: 56-66. 10. Kramer P, Sizoo W and Twiss EE. Nephrotic syndrome in Hodgkin's Disease. Report of 5 cases and review of the literature. Netherlands J Med. 1981; 24: 114--9. 11. Davison AM and Thomson D. Malignancy-asso­ciated glomerular disease. In: Oxford Textbook of Clinical Nephrology. Cameron JS, Davison AM, GrUnfeld JP, Kerr D and Ritz E. (Eds.) Oxford: Oxford University Press, 1992: 475-86. 12. Sawyer N, Wadsworth J, Wijnen M and Gabriel R. Prevalence, concentration and prognostic im­portance of proteinuria in patients with malignan­cies. Brit Med J 1988; 296: 1295-8. 13. Keur I, Krediet RT and Arisz L. Glomerulopathy as a para-neoplastic phenomenon. Netherlands J Med. 1989; 34: 270--84. 14. Cantrell EG. Nephrotic syndrome cured by remo­val of gastric carcinoma. Brit Med J 1969: 2: 739-40. 15. Ghosh L and Muehrcke RC. The nephrotic syn­drome: a prodrome to lymphoma. Ann lntern Med 1970; 72: 379-82. 16. Gagliano RG, Costanzi JJ, Beathard GA, Sarles HE and Bell JD. The nephrotic syndrome associa­ted with neoplasia: An unusual paraneoplastic syndrome: Report of a case and review of the literature. Am J Med 1976; 60: 1026--31. 17. Bunn PA Jr and Ridgway EC. Paraneoplastic syndromes. In: De Vita VT, Hellman S and Rosenberg SA (Eds.) Cancer: Principles and Prac­tice of Oncology. Philadelphia: J. B. Lippincott, 1989: 1923-4. 18. Phillips M, Neild GH, Cameron JS, Williams DG and Amlot P. Hodgkin's Disease and focal glome­rulosclerosis. N Engl J Med 1983; 309: 1389-90. 19. Lokich JJ, Galvanek EG and Moloney WC. Ne­phrosis of Hodgkin's Disease. An immune com­plex-induced lesion. Arch lntern Med 1973; 132: 597---600. 20. Hellman S, Jaffe ES and DeVita VT. Hodgkin's Disease. In: DeVita VT, Hellman S and Rosen­berg SA (Eds.) Cancer: Principles and Practice of Oncology. Philadelphia: J. B. Lippincott, 1989: 1707-10. 21. Shapiro CM, Vander Laan BF, Jao W and Sloan DE. Nephrotic syndrome in two patients with cured Hodgkin's Disease. Cancer 1985; 55: 1799­1804. Radio! Onco/ 1993; 27: 214-22. Neoadjuvant intraarterial chemotherapy of malignant soft tissue tumors. A new treatment protocol Andras Konya, 1 Zoltan Vigvary, 1 P. Rah6ty2 1 Department of Radiology, Semmelweis University of Medicine 2 Department of Surgery, National Institute of Oncology, Budapest, Hungary A group of 58 consecutive patients with extremity sarcomas of soft tissue @rigin (23 primaries, 35 recurrences) received preoperative intraarterial chemotherapy so that the prerequisities for limb salvage surgery were created. The average size of the tumors was 140cm2 (25-520cm2), thus the only criteria for patients to get into this study was the absence of any sign of distant spread or such local invasion of the tumor that would have rendered any efforts for limb sparing surgery aimless. To achieve the best possible local control, the intraarterial cytostatic infusions were complemented by different interventional radiological procedures including superselective catheterization, chemoem­bolization, temporary balloon blockage together with applying distal tourniquet in combination with one another establishing a new treatment protocol based on the characteristics of the tumor's vascular supply. The overwhelming majority of 49 patients who received multiple (2-5) treatment cycles showed favourite response rate (4 complete remissions, 16 partial remissions, 14 minimal response, while only one tumor progressed. In 51/58 patients successful limb salvage surgery was performed. The follow up tirne was 8-79 months (average 26.2 months) for 57 patients (one was lost to follow up due to early postoperative death); 34 of them developed local recurrences after a period of 7-28 (average 12.3) months. Thirty-five patients died of diseminated disease, their survival was 8-42 (average 18.9) months. At present, 22 patients are stili alive. Ten underwent another one or two surgeries (follow up 18-48, average 31, months), while 12 patients are alive with no evidence of disease (their survival is 17-79, average 33.2, months) Key words: soft tissue sarcomas, soft tissue neoplasms -drug therapy, neoadjuvant treatment; infusions, intra-arterial, intraarterial chemotherapy; radiology, interventional; limb salvage surgery Introduction With intraarterial (i.a.) chemotherapy the cyto­statics are delivered in a greater concentration Correspondence to: Andras K6nya M.D.Ph.D., Asso­ciate Professor, Department of Radiology, Semmel­weis University of Medicine, H-1082 Ull6i ut 78/a, Budapest, Hungary. to the tumor bearing region to reduce the size of the tumor for limb salvage surgery. The benefits of exposing tumor tissue directly to a higher concentration of drug have been confir­med by experimental studies showing a concen­tration gradient of 3.5-9 between the tumor tissue and the venous blood,1• 2 whereas this factor was 40-50 between the general venous route and the leve! of subterminal arteries.1 UDC: 616-006.3.04-08 The gradient is further increased by supplemen­ Neoadjuvant intraarterial chemotherapy of malignant soft tissue tumors tary methods such as superselective catheteriza­tion, temporary balloon blockage using an oc­cluding balloon catheter, chemoembolization and applying tourniquet.2---{i This study is aimed to summarize our prelimi­nary results obtained by interventional radio­logy procedures to improve local tumor control and thus promote limb salvage. It is now appa­rent that nonamputative surgery with neoadju­vant treatment is possible in many patients with malignant soft tissue tumors, with a local con­trol rate that approaches, and in many instances exceeds, that achieved by amputation.7• 8 • 9 We have also tried to establish a new neo­adjuvant treatment protocol including a multi­cycle, combined i.a. chemotherapy complemented by interventional radiology pro­cedures which is based upon the consideration of the characteristics of the tumor's vascular supply. Material and methods A group of 58 consecutive patients with extre­mity soft tissue sarcomas of differing histology received preoperative i.a. chemotherapy via a catheter placed percutaneously under fluorosco­py. The localization of tumors is shown in Table l. Two thirds of the Iesions developed in the thigh and leg (25 and 11, respectively) while true extracompartmental location was encoun­tered only in 10 patients ( axillary, inguinal and popliteal regions). The tumor size was establish­ed by arteriography. In ali tumors, the greatest diameter exceeded 5 cm. According to WHO criteria the correct tumor size was determined as the product of two greatest diameters per- Table l. Localisation of the tumours (n = 58) Axillary region 3 Upper arm 6 Forearm 3 Hand 2 Subinguinal/Inguinal region 4 Thigh 25 Popliteal region 3 Leg 11 Foot 1 Tota! 58 Table 2. Distribution of treatment modalities. Group 1 Superselective catheterization (n = 19) + Embolization/chemoembolization ± Temporary balloon blockage Group 2 Selective catheterization of the main artery (n = 19) of the tumour-bearing region with distal tour­ niquet + superselective approach of feeder(s) ± Embolization/chemoembolization ± Temporary balloon blockage Group 3 Selective catheterization of the main artery (n = 20) of the tumour-bearing region with distal tour­niquet pendicular to one another. On the basis of this calculation the size of the tumors varied bet­ween 25-520cm2 (average 140cm2). As to the treatment modality three groups were formed according to the characteristics of tumor's vascular supply (Table 2). Group 1 (n = 19) All the supplying arteries could be approached superselectively. Superselective catheterization was preferred if it was feasible and in the majority of cases this was supplemented by embolization/chemoembolization and/or inter­mittent balloon blockage using an occluding balloon catheter ( e.g. in deep femoral artery, anterior tibial artery). Group 2 (n = 19) If the tumor had multiple vascular supplies which could be catheterized only in part, after treating the neoplasm through the approachable branches by means of i.a. infusions completed by chemoembolization and/or balloon blockage, the main artery of the tumor-bearing region was catheterized. To protect the healthy tissues from toxic-irritative effect of the drugs a pne­umatic tourniquet was placed around the affect­ed Iimb distally to the tumor and was inflated intermittently above the systolic pressure (200­ 250 Hgmm) during i.a. infusions (e.g. most of­ten in the cases where the tumor of the thigh had supplying branches from deep femoral as well as superficial femoral arteries). Group 3 (n = 20) Tumors without selectively approachable arte­ries were treated by placing the catheter into K6nya A et al. the main supplying artery of the affected region as close to the feeders as possible and so called "distal tourniquet" was a!so applied (e.g. mainly in the supplying area of the axillary artery, superficial femoral and popliteal arteries). Treatments were performed as a series of chemotherapy cycles with an interval of 3 weeks in between. The number of courses ranged from one (n = 9) to five (average 3), giving a total of 154 cycles. These cycles were carried out in the form of short-term i.a. infusions which usually lasted for 20-40 min. The basic drug used was doxorubicin (ADM) which was administered as monotherapy in 11 patients. ADM was applied in combination with cisplatin (CDDP) in 34 patients, while a combination of ADM, CDDP and dacarbazine (DTIC) was used in 13 cases. Both ADM and CDDP were applied in concentrated solution (20-25 mg/10 ml). DTIC was solved in its sol­vent and 1 % lidocaine in a ratio of 1 : l. The to tal amount of drug was 50-450 mg ( average 220 mg) for doxorubicin, 50-220 mg (average 150 mg) for cisplatin and 400-1200 mg ( average 800 mg) for dacarbazine depending mainly on the number of cycles per patient. In severa! cycles of 16 patients embolization and mainly chemoembolization was carried out using application of Gelfoam or a mixture of Gelfoam and 10-20 mg doxorubicin and/or cis­platin. In 5 patients (13 cycles) balloon blockage (so called occlusion-infusion) was performed after inserting a balloon catheter into the given sup­plying artery. It was carried out in 10 cycles in a branch of deep femoral artery while in 3 in the anterior tibial artery. Tumor response was established by compute­rized tomography (CT), arteriography and pa­thological evaluation and was classified as fol­lows: complete tumor response (CR), partial tumor response (PR) where there was 50 % or more regression in size. The group of minimal response (MR) was further divided into 2 sub­groups: MR50_25 where tumor regression was between 50 % and 25 % , and MR<2s where tumor response was less than 25 % . Besides the generally used category of overall response rate (CR + PR) we recommend the use of the notion of favourable response rate which included overall response rate with MRso-2s• Results For different reasons 9 out of 58 patients receiv­ed only one chemotherapy cycle so the evalua­tion of treatment in these cased had to be based exclusively on pathological examination of the surgical specimen (6 patients refused additional cycles whereas 3 patients appeared again far beyond the optimal tirne for the repetition of cycle). That is the reason why only the data on other 49 patients, who received 2-5 chemothe­rapy courses and were followed up also by imaging modalities (CT, arteriography), are dis­cussed here. Table 3. Relationship between the tumour size and the efficacy of treatment consisting of multiple (2-5) cycles (n = 49). Size before treatment (cm2) CR PR MRs0-2s MR