Radiol Oncol 1996; .30: 249-54.
Technetium labeled autologous polyclonal immunoglobulin G (lgG) for scintigraphy of inflammation
Nataša V Budihna, Silvester Kladnik, Tomaž Benulič ', Metka Milčinski
Nuclear Medicine, University Medical Center, 'Institute of Oncology, Ljubljana, Slovenia
Background: Radiolabeled heterologous polyclonal human gcimmci g/olmlins (HIG) have been used for detection of inflammation. Autologous immunoglobulins have been success/u//v used _frbr defection o/ inflammation in animals.
Aim of our study was to assess feasibility <f"directly labeled autologous /gG, separatedfrom patients' sera, for' imaging of infection.
Methods: autologous /gG were separated fra/ri patients' sera using /àst protein liquid chromatography (FPLC) and directly radiolabeled with w"'7c. P/awr scintigraphy was performed in 18 patients with suspected inflammation.
Results: sensitivity of autologous /gG scintigraphy was 66 % and spea'/tcity 80% when compared with other' investigations and final diagnosis.
Conc/usion: The method seems promising fbr detection of inflammation although studies comparing radiolabeled autologous /gG compared to heterologous are necessary to prove superiority of either' method.
Key words: autologous lgG, Tc-labeled scintigraphy of inflammation - radionuclide imaging; lgG - diagnostic use; technetium - diagnostic use
Introduction
Radiolabeled heterologous polyclonal human gamma globulins (HIG) have been used several years for detection of inflammation '--1 and accuracy to detect infection is excellent in selected patients.5 Uptake of IgG in some tumors was shown as well.2 HIG are easy to prepare and thus useful for routine clinical work although radiolabeled leukocytes are superior in the diagnosis of focal purulent disease.'' Distribution of various HIG preparations in vivo does not always follow identical pattern, as shown in baboon experiments. Louw et al. believe that it
Correspondence to: Prof. Nataša V. Budihna. M.D., Ph.D., Department of nuclear medicine, University medica! center, Zaloška 7. 1000 Ljubljana. .Slovenia. Phone: +386 61 316 8.55. Fax: +386 6i 132 72 72. The work was in part supported by research grant no. B-6208-312-94, Ministry of Science and Technology, Slovenia.
UDC: 616-002.182-07-097
depends on degree of damage of commercially available lgG caused during the preparation.7
Dormehl and coworkers8 showed foci of inflammation with autologous labeled lgG in baboon experiments. They consider autologous lgG most closely related to intact lgG. Labeled autologous lgG are even less likely to cause allergic reaction in recipient than the heterologous ones and could therefore be useful for repeated investigations in humans.
Aim of our study was to evaluate the feasibility and possible contribution of autologous IgG, separated with fast protein liquid chromatography (FPLC) and directly labeled with 'WmTc, for scintig-raphy in patients with inflammation.
Patients and methods
Separation of /g G
Five to 10 ml of patient's blood were withdrawn and serum was separated. Serum was exposed to
250
Budihna V N et al.
0.3 % sodium cholate (Merck) tor 24 hours before further manipulation to prevent contamination of the chromatographic system.'1
For separation of Ig from the serum, 3.0 ml of serum were precipitated with 3.0 ml of cold saturated ammonium sulfate solution. The resulting suspension was incubated and mixed during next 30 minutes. Centrifugation at 2000 g and 4° C for I 5 minutes followed. The supernatant was discarded and the precipitate redissolved in 1.0 ml of 0.15 Mol/L and pH 6.5 phosphate buffer. The resulting solution was precipitated with 1.0 ml of saturated ammonium sulfate and centrifuged under the same conditions again. The precipitate was redissolved in 1.0 ml of phosphate buffer and passed through a 0.22 pm Millipore filter.
Second step of the separation procedure was IgG separation. The method was originally used for lgM isolation from human serum.10 Our modification is described in detail elsewhere.11 In short, slower flow rate than the originally proposed resulted in improved separation of lgG from lgA. For this purpose, FPLC using strong anion exchange column Mono Q HR 5/5 (Pharmacia), HPLC Gradient Pump 2249 (LKB), ultraviolet light (UV) detector Uvi-cord SD 2510 (LKB), Rheodyne injector, fraction collector RediFrac (Pharmacia) and recorder REC-481 (Pharmacia) were used.
Sample of 200 ml autologous Ig solution was used with FPLC. Following solutions were used: 0.15 Mol/l phosphate buffer pH 6.5 (buffer A) and 0.30 Mol/l phosphate buffer pH 6.5 (buffer B). Mono Q HR 5/5 was rinsed with 20 % ethanol and equilibrated with buffer A before use. First elution was performed with buffer A for I O minutes, the second elution with buffer B for 15 minutes at flow rate of 0.5 ml/min.
Different immunoglobulin classes were detected with measurement of UV absorption and separately eluted. lgG were detected at UV wavelength 277 nm and at absorbancy range 1.0 AUFS (absorbance units full scale, as defined by manufacturer). lgG were eluied at the first, lgA at ihe second and Ig M at the third UV absorption peak.
Confirmation of lgG was accomplished with im-munoelectrophoresis. The first fraction of the elu-ate, containing IgG, was collected separately and its total volume of approximately 0.4 ml was used for direct radiolabeling.
Direct /abeling of /gG with w,"Tc
The labeling was accomplished using the method of Pettit12 with stannous tartrate as reducing agent for
perteclrnetate. lgG solution was preincubated for 20 minutes at room temperature with 50 ml of 2.1 x 1()"3 Mol/L stannous tartrate solution (Sigma) and 500-800 MBq '""TcO, were added in a volume as small as possible. lncubation at 40" C for I 0 minutes followed. I 00 pl of saturated NaHCO3 were added thereafter and incubated again at 40" C for 20 minutes. The solution was passed through a 0.22 pm Millipore filter.
Quality control
The quality control of radiolabeled lgG was performed in two systems of ascending instant thin-layer chromatography (ITLC) using silica gel strips (Gelman) of I x 12 cm size. Methanol (85 %) was used as solvent in the first ITLC system. Free per-technetate moved within the solvent front, while the radiolabeled lgG and colloidal forms of techne-tium remained at the start.
Silica gel strips were soaked with serum albumin (5 g/L), rinsed in deionized water and allowed to dry. They were used in the second ITLC system as a supporting phase while the mixture of solvents (C,H,OH: NH,OH: H,O in 2:1:5 ratio) served as a mobiIe phase. The colloidal forms of technetium (hydrolyzed or reduced technetium) remained at the start, while the labeled lgG and free pertech-netate moved with the solvent's front.
The percentage of free pertechnetate was calculated from the radioactivity of the solvent's front in the first ITLC system. The radioactivity at the start of the second ITLC system represented the colloidal forms of technetium.
The difference between distribution of radioactivity in the first and in the second ITLC system gave the estimation of radioactivity bound to IgG. 11
Radiolabeled IgG was incubated in autologous serum at 37" C for 18 hours to test in vitro stability of radiolabeled autologous lgG.
Patients
The study protocol was approved by the national ethical committee. All patients gave informed consent before entering the study.
Eighteen nonselected patients with proven or suspected infection were included. Blood sample was withdrawn one day before the scintigraphic investigation.
Autologous immunoglobulins were labeled with technetium-99m (500 - 800 MBq) on the following day and reinjected intravenously. Acquisition of scintigraphic data followed.
Technetium labeled attlo/ogous polycloiia/ iiiiimiiiog/obuliii G (lgG) for scintigraphy of iii//aiiiiiiation 251
Scintigraphy
Planar spot view scintigrams or' the whole bocly were acquired in anterior ancl posterior projections with a large field of view gamma camera (Siemens Basicam), equipped with LEAP collimator and connected to a computer (Mclntosh llfx, McLear software for nuclear imaging), at .30 minutes, .3-6 hours ancl 17 hours alter intravenous injection of autologous radiolabeled IgG.
Dala analysis
Scintigrams were analyzed by two experienced observers, blincled to the clinical data. The lesion to background uptake of WmTc IgG was evaluated visually and the uptake intensities compared. Accumulation of labeled IgG in liver, kidneys, spleen and bone marrow as well as in the blood pool in the first phase after application was considered normal. The uptake in organs and tissues higher than the normal background was considered abnormal. Studies with nonconclusive accumulation in organs other than the normal distribution were reported as equivocal. Final decision was made by consensus in case of disagreement of separate readings of the two observers.
The results of scintigraphy with 'Wl"Tc IgG were compared with final diagnosis and with findings of all available investigations as bone scan, radiograms, computed tomography, ultrasound, erythrocyte sedimentation rate and white blood celi count.
Scintigraphy was validated using sensitivity and specificity of the new method compared with the final diagnosis for ali studies. False positive or negative results were analyzed in detail.
Results
lgG sepamlion wul labeling
IgG were completely separated from other immunoglobulins in ali sera as confirmed with chromatography. The labeling efficiency of isolated autologous lgG was above 95 %. Labeled '"'"Tc-IgG were stable in vitro. Only 2.5 % of free pertechnetate and 4.6 % of colloid was detected after 18 hours' incubation at .37 "C.
Patients
No adverse reaction was observed after intravenous application of autologous w"'Tc-lgG to the patients.
Seven men and 11 women, in average 46 (±15.6) years old, all with suspected inflammatory lesions were included. Two patients had spondylodiscitis, single patients had discitis and paravertebral abscess, cliscitis and psoas abscess, paravertebral, suspected pancreatic, gluteofernoral and paranephric abscess. Osteomyelitis was suspected in four patients, pulmonary sarcoidosis in 2 and 4 patients were investigated for suspected myocarditis (Table 1). Average duration of inflammation in patients (except in patients with sarcoidosis and myocarditis) was 2.8 (± 3.1) months. Most patients were treated with antibiotics at the time of scintigraphy. Only 2 patients with suspected osteomyelitis, 2 with sarcoidosis and 3 with myocarditis did not have antibiotic therapy.
Eight suspected inflammatory lesions were true positive and 4 true negative. In one patient with suspected inllammation in recent bone fracture the lesion was false positive. False negative were 3 lesions in patients with psoas or paravertebral abscesses ancl in one patient with spondylodiscitis in previously irradiated area. Results of scintigraphy in two patients with dilatative cardiomyopathy after myocarditis and in one patient with sareioclosis were equivocal.
Eight of 17 lessions in suspected inflammation were true positive, 4 were true negative, 4 were false positive ancl I false negative: calculated sensitivity of our method (without the three equivocal cases) in detection of inflammation is 66 % and specificity 80
Discussion
Detection of inflammation with heterologous polyclonal human gamma globulins (HIG) is well accepted. As autologous IgG are considered to be most closely related to the intact IgG,' they might also have more specific affinity for patient's antigens than the heterologous ones and could therefore be suitable for investigations in humans. Radiolabeled autologous IgG are not likely to cause allergic reaction in the recipient than heterologous poly-clonal HIG.
Autologous technetium labeled lgG coulcl be therefore useful for detection and follow-up of patients with inflammation or tumors without risk of immu-nologic reactions and with low racliation exposure.
Our study was planned to evaluate feasibility of autologous lgG, separated with fast protein liquid
Table 1. Patient's data.
Patient. Gender		Age	Therapy	Final	Disease	Radiogram	Ultrason nd	Cytology	Autologous	Wm TC-HMPAO	Bone	Laboratory Tests
Number				Diagnosis	Duration	Computed			99m Tc"-lgG	Labeled	Scintigram	
						Tomography			Scintigram	Leucocyte		
I	Female	34		Discitis L4/5	2	Discitis			Uptake	Uptake	ESR 22.	
					months					L4/5	L4/5	Leucocytes 8.5
2	Female	35	Anti-	Paravertebral	1	Paravertebral			Uptake		Uptake	ESR 102.
			biotics	abscess, disciti	week	lesion, discitis			L4/5		L4/6	Leucocytes 9.1
						L4/5						
3	Female	58	Anti-	Osteosynthesis	2	Fragments			Uptake in	Negative	Uptake in	ESR 45.
			biotics		months	of bone without			fracture		fracture	Leucocytes 8.1
						callous						
4	Female	65	Anti-	Spondylo	10	L3/5 degenerative			Negative		Low uptake	EST 120
			biotics	discitis	days	changes					postirradition	
5	Female	59	Anti-	Paranephritic	2		Parane-	Pus	Paranephric			ESR 32
			biotics	abscess	months		phric fluid		uptake			
6	Female	36	Anti-	Post osteo-	5				Negative			
			biotics-	myelitis	months							
7	Female	71	Anti-	Discitis. psoas	6	Inflammation left		Pus in	Uptake	Negative	Uptake	ESR 60
			biotics	abscess	months	SIS. L4/5. psoas		psoas	L4/5		L4/5	
						abscess (CT)						
8	Female	58	Anti-	Osteomyeitis	3	Partially healed			Uptake		Uptake	Hemoculture
			biotics	in femoral	months	fracture			fem ui-		femur	Streptococcus
				fracture								viridans
9	Male	64	Anti-	Gluteofemoral	1	Discitis L4/5.			Uptake	Uptake	Uptake	Positive
			biotics	abscess	month	pus in thigh			L4/5.	glu teal	L4/5	hemoculture
									gluteal			
10	Male	56		Pancreatic cyst	2	Cyst in pancreas	Cyst in		Negative	Negative		
					months		pancreas					
11	Male	49	Anti-	Paravertebral	3	Spondylo-	Heart.		Negative			Hemoculture
			biotics	abscess	weeks	lysis L5	abdomen					Staphyloccocus
							normal					aureus
I 2	Male	45		Dilatative	2	Enlarged heart	No pericardial	Few	Equivocal in	Antimyosin		
				myocardiopathy	months		effusion	signs of	myocardium	antibodies		
				post myocarditis				myocar-		positive		
								ditis				
I 3	Female	20		Postpartal	1	Enlarged heart	Intracardial		Uptake in the			
				myocardiopathy	month		thrombus		heart			
14	Male	25		Dilative	18	Enlarged heart	No pericardial	Few	Equivocal in	Antimyosin		
				myocardiopathy	months		effusion	signs of	myocardium	antibodies		
				post myocarditis				myocar-		equivocal		
								ditis				
15	Male	74		Pulmonary	120	Thorax radiogram			Equivocal			ESR 4
				sarcoidosis	months	normal						
16	Female	44		Pulmonary	24	Hilar lymph nodes			Equivocal			
				sarcoidosis	months							
17	Female	43	Anti-	Sepsis.	1				Negative			Leucocytes 13.8
			biotics	suspected	month							
				myocarditis								
18	Male	37		Osteoarthritis	12 mnnthc	Osteoarthritis			Negative		Osteoarthritis	
L4/L5 or = fourth to fifth lumbar vertebrae. SIS = sacroiliac joint. CT = computed tomography. L = left. ESR = erythrocyte sedimentation rate (nim/h). Leukocytes = number x 10!7[.. normal 4.0-10.0. '""Tc-HMPAO = scan with "'"Tc-HMPAO labeled leucocytes.
Technetium labeled autologous polyclonal immunoglobulin G (lgG) for .scintigraphy of iiiflaiiiiiiatioii 253
chromatography (FPLC) and directly labeled with ,J,J"'Te, for scintigraphy of inflammation. Modified flow rate with FPLC allowed complete separation of IgG from other serum proteins. In vivo instability of technetium labeled IgG was shown to be a problem with direct labeling via stannous ion reduction. 14 Hnatowich and coworkers found evidence of in vivo transchelation of w"Tc to cistein as a cause of high renal radioactivity. They also observed IgG fragmentation after stannous ion reduction. Fragments can be excreted via kidneys as well. Our studies in vivo" demonstrated high radioactivity in kidneys and in urinary bladder but not in the thyroid, what is consistent with stable binding of technetium. Stability of directly labeled ,J,J"Tc-IgG complex was in our study proved in vitro. Concentration of radioactivity in lesions was high enough to allow scintigraphic visualization of most inflammatory lesions in spite of shortcomings of direct labeling of autologous IgG with '"""Tc.
The mechanism of IgG accumulation in inflammatory and neoplastic lesions is not entirely understood. One of the most important reasons for IgG uptake in intlammatory lesions is increased vascular permeability and the diffusion at the site of inflammation." Some authors suppose that the at-, traction between Fc region of lgG and Fc receptors on the leukocytes and bacteria allows accumulation of labeled substances. "'-'s The type of immunoglobulin was shown to play a role in accumulation at the site of inflammation.1'1
The results of scintigraphy in our patients with infection were comparable to the results of studies with heterologous technetium labeled IgG.4"20 Autologous 'w,"Tc-lgG are not equally successful in detection of all types of inflammation. Satisfactory results were achieved in patients with spondylodis-citis (Figure 1 ). The fact that labeled leucocytes have limitations for scintigraphy of inflammation in spinal region5- 12 warrants further study of this specific application of radiolabeled w'"Tc-lgG. On the other hand, because of high blood pool activities the results in patients with myocarditis and sarcoidosis were equivocal. False negative autologous WmTc-lgG scintigram in paravertebral or psoas abscesses was the rute in our study. It was probably due to high background of the spine, liver and kidneys in comparison with the uptake in the abscess. The false negative scan in spondylodiscitis in a patient with previously irradiated spine was probably due to low bone metabolism and lowered perfusion of that area. False positive uptake in a recent
unhealed fracture and in some osteoarthritic joints showed the low specificity of the new method. The results of our study are encouraging in patients with spondylodiscitis. Larger group of patients should be studied for more reliable evaluation of sensitivity and specificity of scintigraphy with autologous <»«Tc-IgG. Furthermore, tomographic techniques would improve diagnostic accuracy of our method. Studies comparing technetium labeled autologous with heterologous IgG are necessary to prove possible differences between the two methods. Newer direct labeling techniques12 probably have advantages for scintigraphy with autologous and hetero-logous polyclonal IgG.
Figure l. Scintigram with autologous [gG labeled with technetium-99m in a patient with discitis in the lumbar spine. High uptake of technetium-99m labeled autologous IgG is shown in 4th and 5th lumbar vertebrae.
Conclusions
The present study proves safe and feasible use of autologous lgG, directly labeled with technetium, for scintigraphy in patients. Convincing uptake of autologous 'w"Tc-IgG was shown in inflammatory lesions like spondylodiscitis. It is suitable for repeated use because there is no danger of allergic reaction in the recipient. The method of separation and labeling is to complicated to be used unselec-tively in routine clinical practice.
Reference
1.	Rubin RH, Fishman AJ, Needlman M et al. Radiolabeled, nonspecific, polyclonal human immunoglobulin in the detection of focal inflammation by scintigraphy: comparison with Gallium-67 citrate and technetium-99m-labeled albumin. .l Nuc/ Med 1989; 30: .385-9.
2.	Rubin RH, Fishman AJ, Callahan JR et al. mIn-la-beled, nonspecific, immunoglobulin scanning in the
254
Budi/mu V N et al.
detection of focal infection. N Engl.! Med 1989; 321: 935-40.
3.	Buscombe JR, Lui D, Ensing G, de Jong R, Ell PJ. "''"'To-human immunoglobulin (HIG) - first results of a new agent for localization of infection and inflammation. Eur J Nucl Med 1989; 16: 649-55.
4.	Corstens FHM. Oyen WJG, Becker WS. Radioimmu-noconjugates in the detection of infection and inflammation. Seiiiin Nucl Med 199.3; 13: 148-64.
5.	Serafini AN. Garty 1, Vargas-Cuba R et al. Clinical evaluation of a scintigraphic method for diagnosing inflammations/infections using indium-111-labeled nonspecific human Ig G. ./ Nucl Med 1991; .32: 222732.
6.	Hovi 1, Taavitsainen M, Lantlo T. Vorne M, Paul R, Remes K. Technetiurn-99m-HMPAO-labeled leukocytes and technetium-99m -labeled human polyclonal immunoglobulin G in diagnosis of focal purulent disease. J Nucl Med 1993; 34: 1428-34.
7.	Louw WKA, Dormehl IC. Hugo N. Redelinghuys IF. Species and immunoglobulin preparation related effects on the biodistribution of technetium labeled immunoglobulin G in a baboon model. Eur J Nucl Med 1993; 20: 96-100.
8.	Dormehl IC, Louw WKA, Hugo N. Biodistribution and accumulation in inflammatory lesions of different thiol reduction-mediated '''''"Te-Ig G preparations in baboon models. Nucl Med Commun 1994; 15: 475-82.
9.	Horowitz B. Wiebe ME, Lippin A, Stryker MH. Inacti-vation of viruses in labile blood derivatives. Transfusion 1985; 25: 516-22.
10.	Sampson IA. Hodgen AN. Arthur 11-1. The Separation of immunoglobulin M from human serum by fast protein liquid chromatography. J /imumwl Methods 1984; 8: 9-15.
11.	Kladnik S, Budihna NV, Batagelj 1, Gubina M. Separation and technetium labeling of autologous polyclonal immunoglobulin G. Nucl Med Biol 1996 (in press).
12.	Pettit WA. DeLand FH, Bennett SJ, Goldenberg DM. Improved protein labeling by stannous tartrate reduction of pertechnetate. ./ Nucl Med 1980; 21: 59-62.
1.3. Thrall JH, Freitas JE, Swanson D et al. Clinical comparison of cardiac blood pool visualisation with tech-netium-99m red celi labeled in vivo and with techne-tium-99m-humaii serum albumin. J Nucl Med 1978; 19: 796-80.3.
14.	Hnatowich DJ, Mardirossian G. Rusckowski M, Fogar-asi M, Virzi F. Winnard PJr. Directly and indirectly technetium-99iii-labeled antibodies - A comparison of in vitro and animal in vivo properties. ./ Nucl Med 199.3; .34: 109-19.
15.	Oyen WJG. Claessens RAMJ. van der Meer JWM. Corstens FHM. Biodistribution and kinetics of radiolabeled proteins in rats with local infection. J Nucl Med 1992: 33: 388-94.
16.	Fischman AJ. Rubin RH, White AA et al. Localization of Fc and Fab fragments of nonspecific polyclonal Ig G at focal sites of inllammation. J Nucl Med 1990; 31: 1199-205.
17.	Demacker PNM, Dormans TPJ, Koenders EB, Corstens FHM. Evaluation of lndium-111-polyclonal immunoglobulin G to quantitate atherosclerosis in Watanabe heritable hyperlipidemic rabbits with scintigraphy: Effects of age and treatment with antioxidants or ethi-nylestradiol. J Nucl Med 199.3: 34: 1.316-21.
18.	Calame W, Feitsma HIJ, Ensing GJ et al. Detection of a local staphylococcal infection in mice with technetium-99m-labeled polyclonal human immunoglobulin ./ Nucl Med 1991; 32: 468-7.3.
19.	Fritzberg AR, Beaurnier PL. Targeted proteins for diagnostic imaging: does chemistry make a difference'' ./ Nucl Med 1992; 33: 394-7.
20.	Sciuk J. Brandau W, Vollet B et al. Comparison of technetium 99m polyclonal human immunoglobulin and technetium 99m monoclonal antibodies for imaging chronic osteomyelitis. Eur J Nucl Med 1991: 18: 401 -7.
21.	Goldenberg DM, Larson SM. Radioimmunodetection and cancer identification. J Nucl Med 1992;33: 803-14.
22.	Sykes T R, Woo T K, Baum R P, Qi P Noujaim A A. Direct labeling of monoclonal antibodies with pho-toactivation. J Nucl Med 1995; 36: 191.3-22.