Radiol Oncol 1997; 31: 178-80. Combined radiation and chemotherapy for squamous cell carcinoma of the anal canal: Results and prognostic variables in a multiinstitutional series of 173 patients Gerhard G. Grabenbauer,1 M. Panzer,2 B. Hültenschmidt,3 R. Döker,4 K. Huber,5 H.J. Kuhne-Velte,6 M. Hutter,7 U. Rühl,8 V. Budach,9 T.G. Wendt,'0 I.H.F. Schneider," J. Druschke,1 M. Meyer,12 O. Dvorak,13 R. Sauer1 Departments of Radiation Oncology of the University Hospitals of 'Erlangen-Nürnberg, 2München-Großhadern, 3Essen, 4Diisseldorf, ''Berlin (Chanté), "' Jena, the Hospitals of 5Nürnberg, 6Konstanz, 7Frankfurt/Main-Nordwest, sBerlin-Moabit, Department of "Surgery, nBiostatistics and '-'Surgical Pathology of the University of Erlangen-Nürnberg, Germany Purpose: This retrospective multicenter study was aimed to asses the effect of combined modality therapy in patients with squamous cell carcinoma of the anal canal, stage Tl-4 NO-3 MO. Patients and methods: Between 1985 and 1994, 173 patients underwent treatment by combined radiation and chemotherapy. A median total dose of 50 Gy was delivered to the primary, perirectal, presacral and inguinal nodes, followed by a local boost in selected cases. 5-Fluorouracil was scheduled as a continuous infusion of 1000 mg/m2/24 h on days 1-5 and 29-33, and mitomycin C as bolus of 10 mg/m2 on days 1 and 29. Results: Cancer related survival (OS), NED-suivival and local control rates at 5 years were 71 ±5%, 59±4% and 67±4%, respectively. Anorectal function was preserved in 91% of the patients in whom the primary was controlled. Only 9.6% experienced severe late toxicity requiring surgery. In univariate analysis, T category (Tl/2 vs. T3/4) was predictive for OS (83+4% vs. 53+9%, p=0.01), NED-survival (75±4% vs. 36±7%, p<0.0001) and local control (81 ±4% vs. 46±7%, p<0.0001). N category (NO vs. Nl-3) influenced NED-survival (66±5% vs. 33+12%, p=0.004) and local control (76±4% vs. 37+13%, p=0.003). Treatment technique (>2 fields vs. 2 fields) was found to be of prognostic value for NED-survival (70+6% vs. 50±6%, p=0.016) and local control (77+6% vs. 58±6%, p=0.018). Only in T3/4 cases the total RT-dose (< 45 Gy vs. 45 Gy) had an impact on NED-survival (42±7% vs. 23±13%, p=0.01) and local control (52±8% vs. 45±15%, p=0.03). In multivariate analysis, the T category (UICC 1992) remained the only significant variable with impact on survival (p=0.04), NED-suiyival (p<0.001) and local control (p=0.003). Conclusion: Treatment with a combination of radiotherapy and chemotherapy is safe and effective for patients with anal canal carcinoma. The improvement of results in advanced stages is warranted. Key words: rectal neoplasms; radiotherapy; chemotherapy; treatment outcome Introduction The potential curative effects of radiation therapy (RT) alone or radio-chemotherapy (RCT) in the Correspondence to: Dr. G. Grabenbauer, Klinik und Poliklinik fuer Strahlentherapie der Univesitaet Erlangen -Nuernberg, Universitaets str. 27,91054 Erlangen, Germany. UDO: 616.351-006.6-08:615.849.1 preservation of anal function has been well established in squamous cell carcinoma of the anal canal.1"'2 The abdomino-perineal resection (APR) is reserved for patients with residual or recurrent carcinoma after primary RT or RCT. As squamous cell carcinoma of the anal canal is relatively rare, it is difficult to assemble a larger series of patients treated by a single protocol so that survival rates, primary tumor control rates, patterns 220 Grabenbauer GG et al. of failure as well as acute and late treatment-related toxicity can be evaluated. This paper gives the results after treatment of 173 patients with carcinoma of the anal canal treated in eight German centers. Patients and methods Between January 1985 and May 1994, 173 patients underwent radical treatment by combined RCT and form the study group. Patient's age, gender and histological type of the tumors are shown in Table 1. Prior to treatment all patients underwent clinical examinations including sigmoidoscopy and biopsy. Chest X-ray, laboratory tests, abdominal ultrasound and CT- scans of the abdomen and pelvis were performed routinely. On the basis of these findings tumor stages were assigned according to the UICC-system of 1987." Data on cancers treated between 1985 and 1987 were revised to conform with these criteria. T and N categories are listed in Table 1. The standard treatment protocol is shown in Figure 1. Radiotherapy The primary tumor region including perirectal, internal-iliac and inguinal lymph nodes was irradiated using parallel opposed anteroposterior-postero-anterior fields in the early years of the study (19851988) and later using a 3- or 4-field box technique. External RT was delivered with megavoltage equipment (mostly 6-10 MV-photons) and single fractions between 1.6 and 2.0 Gy (median 1.9 Gy) in an uninterrupted course up to a median total dose of 49.5 Gy. The radiation dose was specified to the isocenter using multiple field techniques or to the midplane for parallel opposed fields. Otherwise specified doses were retrospectively assigned to the reference point according to the ICRU 50 guidelines. Forty-five patients received an additional boost of external RT, 31 patients using interstitial brachytherapy (BT), 12 patients Iridium-192 low-dose-rate, 11 patients Au-198 and 8 patients Iridium-192 high-dose-rate. Table 1 gives the dosages of external RT and of the Ir-192-BT for all 173 patients. Dosimetric details of the Au-198-BT patients were reported elsewhere.14 Chemotherapy One-hundred and seventy-three patients received concomitant chemotherapy. 5-Fluorouracil (5-FU) was scheduled as a continuous intravenous infusion for 120 hours (1000 mg/m2/24 h) to a maximum of 1800 mg/24 h on days 1-5 and 29-33. Mitomycin C (MMC) was administered on days 1 and 29 as a single bolus intravenous injection with a dosage of 10 mg/m2. The second course of chemotherapy was adjusted according to the extent of treatment related hematologic, gastrointestinal and cutaneous toxicity. A summary of the dosage of 5-FU and MMC is given in Table 1. Table 1. Patients characteristics Number of patients 173 Median age (range) 61 years (26-82) Histological type Squamous cell carcinoma 127 (73%) Cloacogenic carcinoma 46 (27%) T category (UICC 1992) Patients (%) T1 28 (16) T2 77 (45) T3 46 (27) T4 19(11) TX 3(1) N category (UICC 1992) NO 115(66) Nl-3 28(17) NX 30(17) Grading (UICC) Gl 21 (12) G2 74 (43) G3 49 (28) G4 2(1) GX 27 (16) External RT 30-44 Gy 45 (26) 45-50 Gy 59 (34) 51-55 Gy 25 (14,5) 56-60 Gy 25 (14,5) 61-66 Gy 19(11) Ir-192 low-dose-rate 12(7) 12 Gy 5(3) 15-16 Gy 7(4) Ir-192 high-dose-rate 8(5) 7-18 Gy 4(3) 20-26 Gy 4(3) Au-198 11 (6) 5-FU (mg/m-) 3000-4000 61 (35) 4500-8000 84 (49) 8200-12000 28(16) MMC (mg/m2) 7-10 67 (38) 11-20 81 (47) 24-35 17 (10) none 8(5) Combined radiation and chemotherapy for squamous cell carcinoma of the anal canal 221 62±8 % "-"i 53 ±9% T3/4 p = 0,0102 105 89 50 26 13 4 00 65____44____24____16__7____2_atrisk 0 20 40 ™60 To 100 120 Time (months) Figure 1. Cancer related survival according to T category (UICC 1992) Follow-up assessment Two months after the completion of therapy all patients were reassessed by digital examination, sigmoidoscopy and CT scans of the pelvis. In case of residual mass or suspicious ulcers multiple biopsies were taken under general anesthesia. Histologically verified lesions in the anal canal were counted as local disease, positive perirectal, iliac or inguinal nodes as regional disease. After negative biopsies patients were followed up at 3-month-intervals for two years from treatment and at 6-month-intervals thereafter. Median follow-up was 40 months with a range between 3 and 130 months. No patients were lost to follow-up. All events prior to May 1996 were included in the analysis. For calculating survival rates only cancer related deaths were counted (adjusted survival rate). The rates of survival, NED (no evidence of disease)-survival and locoregional tumor control were determined according to Kaplan and Meier.15 Differences between patient groups were assessed by the log-rank-test, multivariate analysis was performed according to the Cox-re-gression-model.'6 Results Cancer related survival-, NED-survival- and loco-regional tumor control rates were 71 ±5 %, 59±4 % and 67±4 % at 5 years. 34 patients died of anal cancer, 14 of intercurrent disease. The survival rate for patients with Tl/2-tumors was 83+4 % compared to 53+9 % in patients with T3/4-tumors (Figure 2, p=0.0102). Significant prognostic factors for all three endpoints are shown in Table 2. Only the T- and N category had a highly significant prognostic impact on NED-survival. For T3/4-tumors a total RT-dose of less than 45 Gy led to a significantly inferior NED-survival being 23±13% at five years compared to a survival rate of 42±7% following higher doses (p=0.01). Multiple field arrangements were associated with a better NED-survival of 70±6% in comparison to parallel opposed fields leading to survival rates of 50+6% (p=0.016). Primary and regional tumor control Two months after completion of therapy 134 (77%) patients had a clinical complete remission, 38 (22%) patients a partial remission, one patient no change. All patients with partial remission had positive biopsies, either performed as local excisions (12 patients) or multiple needle biopsies (14 patients). Significant prognostic factors for the loco-regional tumor control are shown in Table 4. Patients with smaller lesions up to a maximum diameter of 5 cm (Tl/2) had a loco-regional tumor control rate of 81+4% (Figure 3) as compared to patients with larger tumors (T3/4), in whom a control rate of 46±7% was achieved (p<0.0001). Positive regional lymphnodes were associated with a poor loco-regional tumor control rate of 37±13% (Figure 4). By contrast a 76±4%-loco-regional control rate was noted in NO-cases (p=0.003). For patients with larger primaries (T3/4) a total RT-dose of less than 45 Gy led to a significantly lower tumor control rate of 45±15% as compared to 52±8% for doses above or equal to 45 Gy (p=0.03). Tumor control was also influenced by the treatment technique. The use of parallel opposed anterior and posterior fields was associated with a control rate of 58±6% which was significantly inferior to the results following treatment with multiple fields leading to a control rate of 77±6% (p=0.018). Patterns of failure Forty-nine (28 %) of the 173 patients experienced a local and/or regional tumor recurrence. In 26 cases there was an isolated local failure, in 7 cases a combined loco-regional failure and in another 8 cases a regional failure alone. Recurrences were noted after a time interval between 2 and 58 months (median 12 months). Twenty-six (15 %) patients experienced distant metastases between 1 and 35 months after completion of therapy. Distant metastases were combined with a local and/or regional failure in 9 cases. 222 Grabenbauer GG et al. 46 ±7 % T 3/4 p< 0,0001 Time (months) Figure 2. Locoregional tumor control according to T category. 76 ±4 % NO ^ 59 ±10% H p = 0,0030 J 37 ±13 »/ N + 115 77 46 26 15 5 28 12 6 3 1 at risk Time (months) Figure 3. Locoregional tumor control according to N category. Table 2. Significant prognostic factors for survival, NED-survival and locoregional tumor control at 5 years. Factor Overall survival NED-survival Tumor control T category Tl/2 T3/4 83±4% 53±9% (p=0.0102) 75±4% 36±7% (p<0.0001) 81 ±4% 46±7% (pcO.OOOl) N category NO Nl-3 73±5% 66+11% (n.s.) 66±5% 33±12% (p=0.004) 76±4% 37±13% (p=0.003) RT Dose (T3/4) >45Gy <45Gy 57±9% 67±16% (n.s.) 42+7% 23+13% (p=0.01) 52±8% 45+15% (p=0.03) RT/CT Dose >45Gy, >6g 5-FU <45Gym <6g 5-FU 73±6% 69+14% (n.s.) 61 ±6% 45±11% (p=0.042) 67±6% 54±12% (p=0.11) Treatment technique >2 fields 2 fields 74±9% 71+6% (n.s.) 70±6% 50±6% (p=0.016) 77+6% 58±6% (p=0.018) n.s. not significant Table 4. Multivariate analysis on prognostic factors for cancer related suvival, NED survival and local tumor control. Cancer related survival VariableBeta 95% C.I. Exp (Beta) P T category -0,2992 0,5546-0,9910 0,7414 0,043 Technique -0,0562 0,6880-1,2990 0,9457 0,72 RT-dose 0,0116 0,0759-1,0487 1,0016 0,53 MMC-dose -0,0265 0,9264-1,0237 0,9738 0,29 NED-survival T category -0,5591 0,4397-0,7434 0,5717 0,001 Technique 0,1781 0,9004-1,5869 1,1954 0,217 RT-dose -0,0228 0,9447-1,0114 0,9775 0,19 MMC-dose -0,0164 0,9436-1,0256 0,9837 0,44 Local tumor control T category -0,5458 0,4302-0,7804 0,5794 0,003 Technique 0,2674 0,9436-1,8090 1,3065 0,10 RT-dose -0,0100 0,9530-1,0285 0,9901 0,60 MMC-dose -0,0046 0,9503-1,0427 0,9954 0,84 77 ±6 % > 2 fields 58 ±6 % 2 fields p = 0,0189 Time (months) Figure 4. Locoregional tumor control according to treatment technique. Toxicity The various levels of acute toxicity were classified as stated by the WHO 17 and are shown in Combined radiation and chemotherapy for squamous cell carcinoma of the anal canal 223 Table 3. Acute toxicity (WHO) and late treatment related toxicity (Eschwege) among 124 patients* following radio-chemotherapy of anal canal carcinoma. Grade Dermatitis Diarrhea Anemia Leucopenia Thrombocytopenia Late Toxicity 0 10(6%) 20(11%) 99 (56%) 59 (33%) 93 (53%) 77 (62) 1 13(7%) 28 (15%) 18(10%) 21 (12%) 19(11%) 9(7) 2 42 (24%) 45 (25%) 7 (4%) 27(15%) 7 (4%) 26 (21) 3 56 (32%) 31 (18%) 0 16(9%) 3 (2%) 12(10) 4 3 (2%) 1 (1%) 0 1 (1%) 2(1%) * Data on toxicity from one center not available Table 3. The frequency of severe hematological toxicity (grade 3/4) was 22/124 patients (18%). Thirty-four percent and 19% of the patients experienced a dermatitis and enteritis of at least grade 3 (WHO), respectively. Late toxicity was scored according to Eschwege et al. (1985) and is shown in Table 5. Only 12 of 124 patients (9.6%) had late sequelae of grade 3, requiring surgery. The use of interstitial BT using Ir-192 high-dose-rate and Au-198 in two centers had a significant negative impact on survival without grade 3 toxicity, which was 92% without and 74% with interstitial BT of this type. ease. Adequate local excision is therefore not feasible under the prerequisite of preserving an intact anal sphincter. Local excision as a treatment strategy appears to be only appropriate for small lesions (< 2 cm) involving the anal margin and the perianal skin.18 APR was the standard treatment for anal cancer in many centers until the mid-eighties. Five-year-survival-rates between 40% and 70% have been reported. Obvious disadvantages of the APR include the permanent loss of the anal sphincter function, a significant postoperative morbidity and mortality as well as impotence.'9"23 Presetyation of anorectal function A major objective of the conservative treatment of anal carcinoma by RCT or RT is the preservation of anorectal function. APR was performed in 51 (29%) of the patients, in 41 patients after local recurrence and in 10 patients for other reasons. The remaining 12 patients were scored as being partially incontinent. Thus a functioning anorectal sphincter was preserved in 110 of 173 patients (64%), and in 110 of 121 patients (91%) in whom the primary tumor was controlled by RCT. Multivariate analysis The following variables were included in the analysis: Total RT dose, total dose of MMC (continuous variables), treatment technique, T category (Tl/2 vs. T3/4) and N category (NO vs. Nl-3), as catego-rial variable. The only independent and significantly related factor for survival, NED-survival and loco-regional tumor control was the T category with a p-value of 0.04, 0.001, and 0.003, respectively (see alsoTable 4). Discussion Carcinoma of the anal canal tends to cause sphincter muscle invasion in the early course of the dis- Treatment options, results and late effects It has been the ongoing and continuing policy of several French centers to use RT alone or in combination with interstitial BT for conservative treatment of anal cancer with excellent results.,3,10' 12.18,24,25 rt alone proved to be very effective for small lesions of less than 4 cm in diameter: Local tumor control rates between 76% and 91% were reported.3'9 For adequate local control of larger primaries, however, relatively high total doses between 60 and 65 Gy had to be applied.3'10'" Consequently late treatment related toxicity was noted more frequently (9-13% grade 3 according to Eschwege) requiring APR for the control of distressing symptoms. In a series reported by Touboul et al. only 65% of the patients with no evidence of local tumor had an intact anal sphincter." By contrast, local complication rates as low as 3% together with local tumor control rates of 88% (23,24) were reported by centers using RCT with and without interstitial BT.2'5'6'27 Nigra et al. (1974) pioneered in the clinical use of concomitant RCT as a neoadjuvant treatment strategy for downstaging of anal carcinoma. After 30 Gy total dose and one course of MMC/5-FU histologically negative resection specimens were obtained in almost 60% of the patients. During the following years numerous phase-II studies were conducted. 224 Grabenbauer GG et al. Our results of 173 patients treated between 1985 and 1994 in eight Radiation Centers compare very favorably with the literature data on survival-, NED-survival- and loco-regional control rates at of 71%, 59% and 67%, respectively. The anal sphincter function was preserved in 110 of 121 patients (91%) in whom the local tumor was permanently controlled by RCT. Noteworthy is the fact that in other series using RCT the sphincter preservation rate was as high as 80%5-12,18 in comparison to the results following RT alone.10'"'24 A recent randomized EORTC study provided preliminary data concerning the issue whether RCT is superior to RT alone in advanced anal carcinoma (either node positive or > 4 cm). Both locoregional tumor control and colostomy-free survival were significantly improved in the concomitant arm.26 In an RTOG/Intergroup study the value of additional MMC as part of the concomitant chemotherapy was investigated.4 As has been pointed out earlier in a retrospective series by Cummings et alP NED-survival and local tumor control rates were significantly lower after regimens omitting MMC, but using RT and 5-FU alone. Prognostic factors Our results clearly demonstrate that the tumor volume represented by the T category (UICC 1992) remained the only independent significant prognostic factor for survival, NED-survival and local tumor control. This was also noted in a larger series of 242 patients treated with RT alone.10 In patients treated by RT alone and studied by multivariate analysis the size of the primary tumor was predictive for both survival and local control.'o^8-29 Local tumor control rates of 91% and 89% for tumors of less than 4 cm in diameter versus control rates of 73% and 71% for larger primaries were reported W.I8 In patients treated by RCT, local control rates for primary tumors up to 2 cm in diameter were in the range of 95-100%, for 2-5 cm 80-95% and for larger than 5 cm 65-80%.2.5.27,30 pjve.year survival rates, corrected for death from intercurrent disease, were about 95-100% when the primary was < 2 cm and in the range of 60-70% for larger tumors reflecting the good surgical salvage rates for local relapse after treatment with RCT. It has been suggested by others" that a combination of preoperative RT and surgery or a more ag- gressive RCT including cisplatin based chemotherapy12 could improve local tumor control and survival for patients with T4 tumors. In our series a total RT dose of less than 45 Gy with concomitant chemotherapy does not seem to be adequate for larger tumors (> 5 cm). Control rates and NED-survival were only 45% and 23% at 5 years. Following higher doses, however, results could be significantly improved (p=0.01). Surprisingly in our series refinements in the treatment technique had a significant impact on NED-survival and local control in univariate analysis. This could not be demonstrated by others. In surgical series histopathologically confirmed involvement of perirectal, superior hemorrhoidal, pelvic or inguinal node groups was associated with 5-year survival rates of about 50%, being 25% worse than those of patients without nodal involvement31 The presence of regional lymphnode metastases did not correlate with control of the primary tumor in patients managed with RT- or RCT protocols in the early series.9-28 However, we found a striking difference in local tumor control for patients without regional metastases being 76% vs. 37% for patients with positive regional nodes (p=0.003). It appears that the application of routine CT-scanning of the pelvis for staging purposes may have detected enlarged nodes more accurately. Recently Myerson et al.1 reported a worse disease-free survival of 52% for Tl-3 Nl-3 patients as compared to that of patients in Tl-3 NO stages which was as high as 88% at 10 years (p=0.03). No data exists on patients treated by RT/RCT concerning the prognostic value of tumor DNAcontent. In one large surgical series of flowcyto-metric DNA analysis of paraffin embedded tissue, ploidy was found to be strongly predictive for outcome, patients with diploid tumors having a 5-year survival of 75% compared to 55% for patients with aneuploid tumors." Among serum markers of interest, only serum squamous cell carcinoma antigen (SCCAg) provided significant influence in one multivariate analysis predicting tumor specific death rates and recurrent carcinoma.2'-' In conclusion the results from the current study and others strongly suggest that with respect to certain prognostic factors further improvement in the therapy of anal canal cancer is possible, particularly modifications of the chemotherapeutic regimen and the radiation dose. Refinements in radia- Combined radiation and chemotherapy for squamous cell carcinoma of the anal canal 225 tion technique and fractionation schedules need to be prospectively evaluated to minimize late sequelae and to preserve anal sphincter function even in advanced cases. References 1. Allal A, Kurtz JM, Pipard, G et al. Radiochemotherapy versus radiotherapy alone for anal cancer: a retrospective comparison. hit J Radiat Oncol Biol Phys 1993; 27: 59-66. 2. Cummings BJ. Anal cancer. Int J Radiat Oncol Biol Phys 1990; 19:1309-15. 3. Eschwege F, Laser P, Chavy A, Wibault P, Kac J, Rougier Ph. Squamous cell carcinoma of the anal canal: treatment by external beam irradiation. Radiother Oncol 1985; 3: 145-50. 4. Flam MS, John M, Pajak T et al. Role of mitomycin in combination with fluorouracil and radiotherapy, and of salvage chemoradiation in them definitive nonsurgical treatment of epidermoid carcinoma of the anal canal: results of a phase III randomized intergroup study. J Clin Oncol 1996; 14:2527-39. 5. Grabenbauer GG, Schneider IHF, Gall FP, Sauer R. Epidermoid carcinoma of the anal canal: treatment by combined radiation and chemotherapy. Radiother Oncol 1993; 27: 59-62. 6. Martenson JA, Lipsitz SR, Lefkopoulou M et al. Results of combined modality therapy for patients with anal cancer. Cancer 1995; 76:1731-6. 7. Myerson RJ, Shapiro SJ, Lacey D et al. Carcinoma of the anal canal. Am J Clin Oncol 1995; 18:32-9. 8. Nigro ND, Vaitkevicius VK, Considine B. Combined therapy for cancer of the anal canal. Dis Colon Rectum 1974; 27: 763-6. 9. Papillon J, Montbaron JF. Epidermoid carcinoma of the anal canal: a series of 276 cases. Dis Colon Rectum 1987; 5: 324-33. 10. Schlienger M, Krzish C, Pene F et al. Epidermoid carcinoma of the anal canal: treatment results and prognostic variables in a series of 242 cases. Int J Radiat Oncol Biol Phys 1989; 17:1141-51. 11. Touboul E, Schlienger M, Buffat L et al. Epidermoid carcinoma of the anal canal. Cancer 1994; 73: 156979. 12. Wagner JP, Mahe MA, Romestaing P, Rocher PF, Berger C, Trillet-Lenoir V, Gerard J-P. Radiation therapy in the conservative treatment of carcinoma of the anal canal. Int J Radiat Oncol Biol Phys 1994; 29:17-23. 13. Hermanek P, Sobin LH, eds. UICCTNM classification of malignant tumors. 4th ed., 2nd revision. Beriin, Heidelberg, New York: Springer, 1992. 14. Panzer M, Sutter T, Wendt T, Jauch KW. Radiochemotherapie mit und ohne Radikaloperation bei Analkarzi- nom: GroBhaderner Langzeitergebnisse. Tumordiagn Ther 1993; 14: 167-74. 15. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 457-81. 16. Cox DR. Regression models and life tables. J R Stat Soc B 1972; 34:187-202. 17. Sischy B. The use of radiation therapy combined with chemotherapy in the management of squamous cell carcinoma of the anus. Int J Radiat Oncol Biol Phys 1985; 11:1587-93. 18. Papillon J, Montbaron JF, Gerard JP, Chassard JL, Ar-diet JM, Touraine-Romestaing P. Role of combined radiation and chemotherapy: the Lyon experience. In: Sauer R, ed. Interventional radiation therapy. Berlin: Springer; 1991. 19. Boman BM, Moertel CG, O'Connel MJ et al. Carcinoma of the anal canal - a clinical and pathologic study of 188 cases. Cancer 1984; 54: 114-25. 20. Frost DB, Richards PC, Montague ED, Giacco GG, Martin RG. Epidermoid cancer of the anorectum. Cancer 1984; 53: 1285-93. 21. Hager T, Hermanek P. Maligne Tumoren der Analregion. In: Gall FP, Hermanek P, Tonak J, eds. Chirurgische onkologie. Berlin: Springer, 1987:176-83. 22. Pintor MP, Northover JM, Nicolls RJ. Squamous cell carcinoma of the anus 1948-1984. Br J Surg 1989; 76: 806-10. 23. Schneider IHF, Grabenbauer GG, Reck T, Kockerling F, Sauer R, Gall FP. Combined radiation and chemotherapy for epidermoid carcinoma of the anal canal, hit J Colorectal Dis 1992; 7: 192-6. 24. Ng YK, Pigneux J, Auvray H, Brunet R, Thomas L, Denepoux R. Our experience of conservative treatment of anal canal carcinoma combining external irradiation and interstitial implant: 32 cases treated between 1973 and 1982. Int J Radiat Oncol Biol Phys 1988; 14: 2539. 25. Pipard G. Combined therapy of anal canal cancer: a report on external irradiation with or without chemotherapy followed by interstitial Ir-192. In: Sauer R, ed. Interventional radiation therapy. Berlin: Springer, 1991. 26. Roelofsen F, Bosset J, Eschwege F, Pfeiffer M, van Glabbeke M, Bartelink H. Concomitant radiotherapy and chemotherapy superior to radiotherapy alone in the treatment of locally advanced anal cancer; results of a phase Ill-randomized trial. Proc ASCO 1995; 14: 194. 27. Cummings BJ, Keane TJ, O'Sullivan B, Wong CS, Catton CN. Epidermoid canal cancer: treatment by radiation alone or by radiation and 5-fluorouraciI with and without mitomycin C.IntJ Radiat Oncol Biol Phys 1991; 21:1115-25. 28. Cummings BJ. Anal canal carcinoma. In: Hermanek P, Gospodarowicz MK, Henson DE, Hutter RVP, Sobin LH, eds. Prognostic factors in cancer. Berlin, Heidelberg, New York: Springer, 1995: 81-7. 226 Grabenbauer GG et al. 29. Goldman S, Svensson C, Bronnergard M, Glimelius B, Wallin G. Prognostic significance of serum concentration of squamous cell carcinoma antigen in anal epidermoid carcinoma. Int J Colorectal Dis 1993; 8: 98102. 30. Tanum G, Tveit K, Karisen KO, Hauer-Jensen M. Chemotherapy and radiation therapy for anal carcino- ma: survival and late morbidity. Cancer 1991; 67: 2462-6. 31. Shephard NA, Scholefield JH, Love SB, England J, Northover JMA. Prognostic factors in anal squamous cell carcinoma: a multivariate analysis of clinical flow cytometric and pathologic factors in 235 cases. His-topathology 1990: 16:545-55.