Inf. .I. Radrnrron Oncolngs Biol. Phps., Vol Printed m the U S A All nghts reserved
0360.3016192 $5.00 + 00 Copyright 0 1992 Pergamon Press
22, pp. 581-584
plc
??Session C: Concepts and Developments in Cell Sensitization
FIRST ANALYSIS ON TUMOR REGRESSION FOR THE EUROPEAN RANDOMIZED TRIAL OF ETANIDAZOLE COMBINED WITH RADIOTHERAPY IN HEAD AND NECK CARCINOMAS M.D. ,* I. CHARREAU, M.D. ,1_H. SANCHO-GARNIER, M.D., M.Sc. ,t F. ESCHWEGE, M.D. * AND E. P. MALAISE, M.D., PH.D. *
D. CHASSAGNE,
Institut Gustave-Roussy, 94805 Villejuif, France From July 1987 to July 1990, 374 patients were randomized in a multicenter trial coordinated by the GustaveRoussy Institute. Patients were treated either by radiotherapy alone (RT) or by combined etanidazole with radiotherapy (ETA). The same radiotherapy protocol was used in both arms. Major deviations from the protocol occurred in 16% of the patients. Etanidazole was given at a dose of 2 g/m’, 3 times per week, for a total of 30-34 g/m’. Seventeen percent of the patients received less than 14 injections (4% refused, 12% presented a toxicity, 1% died before beginning). The rate of neuropathy was 28% in the ETA arm and 3% in the RT arm. Acute radiotherapy reactions occured in similar proportions in both arms. The 3.month rates of complete regression are presently 75.3% in the RT alone group and 77.6% in the ETA group; this difference is not significant. No definitive results are presently available and we must wait for the two-year survival results. In addition, if a meta-analysis could be performed with the parallel RTOG study, the results would be more valid. Hypoxic cell sensitizer, Randomized trial, Head and neck carcinoma. INTRODUCTION
this magnitude has also been described for human tumor xenografts (5). We have selected a group of tumors with the following characteristics for clinical trials on ETA: disease-free survival, which mainly depends on local control of the primary tumor; a tumor whose evolution is sufficiently fast that most of the results would be known within 2 years; 2-year local control of about 50%; conventional treatment without chemotherapy; a tumor that is readily observed and biopsied; an incidence rate sufficient to provide a large number of patients in a reasonably short time; a tumor location for which there are data indicating that hypoxia may influence local control. Head and neck squamous cell carcinomas, mainly T2-T3, NO, Nl, and N2, satisfy the above criteria. The aim of this randomized trial is to assess the possible role of etanidazole with two endpoints: the rate of locoregional control and survival rate. We present here the
The role of hypoxia in the loco-regional control of tumors treated with ionizing radiation is still controversial. However, there is an impressive body of data suggesting that the effect of hypoxia is significant (1, 4): misonidazole appeared to be a potentially very effective member of the 2-nitroimidazole family of radiosensitizers (6, 7). Brown has used pharmacokinetic data to develop a family of 2-nitroimidazoles (1) having the same electronic affinity (and thus the same radiosensitizing capacity) as misonidazole. One of these drugs, SR-2508 or etanidazole (ETA), may have considerable clinical potential. The results of Phase I trial indicate that the total dose which can be given under appropriate conditions is about three times that of misonidazole (2, 3). A dose of 2 g/m* provides tumor drug concentration high enough to predict a sensitizing enhancement ratio (SER) around 1.6. An SER of about
*Radiotherapy Department and U. Inserm 247. TBiostatistical Department and U. Inserm 287. Reprint requests to: D. Chassagne, Institut Gustave-Roussy, 94805 Villejuif, France. Acknowledgements-We thank all the participants in this multicentric trial: Austria: Frommhold, H. W., Innsbrtick, Hackl, A., Graz; Hammer, J., Linz; Kolbabek, Wien. France: Bey, P., Vandaeuvre-les-Nancy; Daban, A., Poitiers; Daly, N., Toulouse; Eschwege, F., Villejuif; Le Floch, O., Tours; N’Guyen, Reims; Peuvrel, P., Nantes; Rambert, P., Saint-Cloud; Resbeut, M., Marseille; Romestaing, P., Pierre-BCnite. Germany: Alberti, W.,
Essen 1; Muller, R.P., Koln; Pfaender, K., Passau; Sauerwein, W., Essen; Schmitt, G., Dusseldorf. Italy: Busutti, L., Bologna; Cionini, L., Firenze; Romanini, A., Roma; Sannazzari, G.L., Torino. United Kingdom: Bleehen, N., Cambridge; Durrant, K., Oxford; Morgan, D., Nottingham; Phillips, R.H., London; White W. J., Guildford. This work was supported by Roberts Laboratories, Eatontown, NJ 07724 USA and in part by the Ligue Nationale Fratqaise Contre le Cancer (Comite des Hauts-de-Seine) France. Accepted for publication 3 July 1991. 581
582
I. J. Radiation Oncology 0 Biology 0 Physics
Table 1. Etanidazole - Phase III initial site and TNM classification (N = 373) Tl-T2 Location Nasopharynx (45) Oral cavity (55) Oropharynx/ base of tongue (152/28) Hypopharynx (17) Larynx (76) Total
T3
T4
Eta
Rt
Eta
Rt
Eta
Rt
Total
NO NUN2 NO NUN2
6 6 1 5
2 5 1 1
2 4 5 9
1 8 10 7
3 1
4 3 2 8
18 27 19 36
NO NUN2 NO NUN2 NO NUN2
16 24 4 3 11 1 77
20 20 4 6 9 3 71
17 26 18 8 89
13 26 21 5 91
1 7
67 113 08 09 59 17 373
6 10 20
25
RTOG; we hope that in the future we will be able to pool the European and the RTOG data (8). METHODS
AND MATERIALS
Patient population The following pretreatment characteristics were observed: average age: 57, 13% and 9% female, respectively, for ETA and RT arms. One hundred percent Karnofsky performance was 95% and 89%, respectively, in the ETA and RT arms. Epidermoid carcinomas were found to be well differentiated in 73% and 77%, respectively, in ETA and RT arms. Table 1 shows the initial sites and TNM classification. Note that the TNM inclusion criteria were defined in order to provide a homogeneous probability of survival for the different patients’ groups. Patients were only included if their laboratory values fell within the following ranges: urea < 10.7 mmol/l, serum creatinine < 135 mmol/l, WBC 2000, platelets 100.000, liver function tests below 1.25 X upper normal limits. Patients with less than 116/100 ml hemoglobin were transfused to bring the level up to 11 g and maintained at this level throughout the radiation treatment period. Radiotherapy protocol 6oCo rays and photons X rays from Linac 4 to 6 MV were advised. Two opposed lateral fields were used to encompass the primary tumor region and the bilateral cervical lymph node chains. Fractionated doses were given daily, Monday through Friday, for 5 equal daily doses of 2 Gy per week. Each patient received 33-37 fractions for a total dose of 66-74 Gy delivered to the target volume over a period of 6.5 to 7.5 weeks. Boosts to primary tumors were given after 50 Gy using photons or electrons (4 to 25 MeV) or by interstitial therapy (Iridium192). Boost dose to persistent node
Volume 22, Number 3, 1992
metastases was given after 50 Gy; the dose varied with the sizes of the node but the entire head and neck region was irradiated with a dose of at least 45 Gy, even for the NO stage. Treatment should not be interrupted for more than 2 weeks; total tumor dose should reach a level of 70 to 74 Gy. Drug regimen SR-2508 was given at a dose of 2.0 g/m*, 3 times per week for 5.5 weeks (17 doses), for a total dose of 34 g/m*. The drug was dissolved in saline and infused IV, over 5-10 min. Radiotherapy was begun 30-60 min after the start of drug infusion “Force p.o. fluids” were routinely performed. There was no compensation made for “area under the curve” estimations. Drug treatment was discontinued if patients developed severe skin rash, any sign of peripheral neuropathy, CNS toxicity, or ototoxicity. Assessment criteria Toxicities were recorded using the RTOG toxicity criteria. Tumor clearance, estimated by measuring the longest axis and that at right angles to it, was assessed at the end of radiotherapy, 6 weeks and 3 months after the end of radiotherapy and recorded as complete remission (CR) or partial regression: PRl if PR > 50% and PR2 if PR < 50%. The time of recurrence was noted when it occurred after CR. The final endpoints will be loco-regional control after 24 months, and survival. Late effects are also recorded. Statistical analysis The randomization scheme used was stratified according to the treatment center and primary tumor site. Treatment group designation was provided by the statistical coordinating unit in response to telephone/telex requests. It was calculated that a total of 400 patients was required (one-tailed binomial with (Y 0.05 and p 0.20) to demonstrate an enhancing effect of SR-2508 on 2-year survival from 35% to 47%. Informed consent Written informed consent was obtained from each patient. The nature of the trial, including the procedures to be used, their purpose, any potential discomfort or risk, and the expected benefit, was explained to each patient. RESULTS Toxicity Fifty-two peripheral neuropathies were noted in the ETA group, and 5 in the RT group. Among these 52, 35 were grade 1, 11 Grade 2, and 2 Grade 3. Three were not graded. Forty-six of these 52 appeared either during or after the fifth week of infusion (28 being diagnosed after the last day of infusion). Twenty-five of these 52 peripheral neuropathies disappeared during the first year of follow-up, 6 patients died from cancer evolution with concomitant neuropathies, and
Etanidazole and radiotherapy in head and neck carcinoma 0 D.
583
CHASSAGNE et al.
% CR
Fig. 1. Complete
regression
rate by site 3 months after the end
of radiotherapy (N = 339).
20 patients are living with different follow-up durations and persistent neuropathies, but usually with an important regression of their neuropathy symptoms. In the ETA arm no modifications of the biological parameters were noted after the infusions of etanidazole. Acute skin and mucosa reactions, which normally occurred during and after the end of radiotherapy, were slightly higher in the ETA group (32 grade 3 or 4 reactions in the ETA group vs 21 grade 3 or 4 reactions in the RT group). However, the difference is not statistically significant. Allergic toxicity. In the ETA group, nine skin rashes have forced interruption of the etanidazole infusions, with no further complications after the stop. Refusal of treatment. After having signed the informed consent, eight patients refused the etanidazole infusions. Tumor regression Three months after the end of radiotherapy is the first endpoint of the trial; the final results will be based on the 2-year survival rate. The overall complete regression rate (CRR) is 76%, 3 months after the completion of radiotherapy; 10% of the patients were either not assigned (7%) or not evaluable (3%). In addition, 2% in the ETA group and 2% in the RT group had salvage surgery. No difference in the CRR 3 months after the end of radiotherapy can be observed between the two arms. Figures 1 and 2 show that both the regression rate and the regression speed significantly differ from one primary site to another. The lowest rate of CRR three months after the end of radiotherapy is in the oral cavity site group (51 cases); the CRR is 58% in the ETA group and 41% in the RT group (Table 2), but this difference is not statistically significant.
DISCUSSION In our Phase II trial we did not find any statistical correlations between the total area under the curve and the risk of peripheral neuropathy. Consequently, in our Phase III
0
IO
100
TIME
Fig. 2. Cumulative (CR).
distribution
100
IS0
IN DAYS
of time to complete
response
trial, we discontinued these dosages. The drug was to be stopped immediately if there was any sign of grade 2 neuropathy and/or skin rash. The three main causes for stopping the radiosensitizer infusion before the 15th dose have been: skin rash (5%), neuropathies (3%), and patient refusal (3%). Overall, only 15% of the ETA patients were given less than 15 infusions. The peripheral neuropathies are in their great majority characterized by only sensitive and subjective symptoms that appeared around the last week of infusion. Half of them are transient, and the symptoms usually disappeared after some duration. Some of these neuropathies were
Table 2. Complete regression rate 3 months after the end of radiotherapy (N = 339)
Site
Eta %
Rt %
Nasopharynx (40) Oral cavity (51) Orophamynx (152)Ibase of tongue (28) Hypopharynx ( 16) Larynx (68) Total (339)
81 58 76 100 88 77
84 41 71 100 97 74
Note: 10% of the patients were either not assigned (7%) or not evaluable (3%), 2.1% in Eta group and 2.4% in RT group had surgery on tumor or nodes.
584
I. J. Radiation Oncology 0 Biology 0 Physics
noted either very soon or very lately, and consequently it is very doubtful that all “declared” neuropathies are linked to etanidazole toxicity. We are in the process of extensive revision of these neuropathies. The 3-month regression rate varies from one site to another (Figs. 1 and 2). This variation in immediate response to radiotherapy is a well known fact and does not preclude the definitive results at 2 years. The high regression rate observed in both arms might be caused by the excellent cooperation achieved by all participants in the quality assurance program, which has permitted a high and homogeneous quality level of the irradiation technique.
Volume 22, Number 3, 1992
In the T3-T4 3-month level of level observed in that in these two of the ETA arm; significant.
of oro-pharynx and oral cavity, the complete regression is lower than the other sites, and it must be emphasized locations a difference appears in favor but this difference is not statistically
CONCLUSION No definitive answers can yet be given about the results of this randomized trial, we must wait for 2-year survival results. A possible meta-analysis with the results of the parallel RTOG study (8) will bring more powerful results.
REFERENCES Brown, J. M. Clinical trials of radiosensitizers: what should we expect? Int. J. Radiat. Oncol. Biol. Phys. 10:425429;
lationship to outcome of radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 14:831-838; 1988.
1984. Chassagne, D.; Malaise, E. P.; Sancho-Gamier, H.; Bonnay, M.; Charreau, I. ; Feilleux, C.; Eschwege, F.; Daly, N. ; Bey, P.; Busutti, L.; Nguyen, T.; Sauerwein, W. Preliminary results of a phase II multicenter trial of “etanidazole” hypoxic cell sensitizer combined with radiotherapy in head and neck squamous carcinomas. Proceedings of the 6th Chemical Modifiers of Cancer Treatment, Paris, 1988. Coleman, C. N.; Halsey, J.; Cox, R. S.; Hirst, V. K.; Blaschke, T.; Howes, A. E.; Wasserman, T. H.; Urtasun, R. C.; Pajak, T.; Hancock, S.; Phillips, T. L.; Noll, L. Relationship between the neurotoxicity of the hypoxic cell radiosensitizer SR-2508 and the pharmacokinetic profile. Cancer Res. 47:319-322; 1987. Gatenby, R. A.; Kessler, H. B.; Rosenblum, J. S.; Cola, L. R.; Moldofsky, P. J.; Hartz, W. H.; Broder, G. J. Oxygen distribution in squamous cell carcinoma metastases and its re-
Guichard, M.; Malaise, E. P. Radiosensitizing effects of Misonidazole and SR-2508 on human melanoma transplanted in nude mice. Influence on repair of potentially lethal damage. Int. J. Radiat. Oncol. Biol. Phys. 8:465-468; 1982. Overgaard, J.; Sand Hansen, H.; Andersen, A. P.; HjelmHansen, M.; Jorgensen, K.; Sandberg, E.; Berthelsen, A.; Hanner, R.; Pedersen, M. Misonidazole combined with splitcourse radiotherapy in the treatment of invasive carcinoma of larynx and pharynx. Report from the Dahanca 2 study. Int. J. Radiat. Oncol. Biol. Phys. 16:1065-1068; 1989. Overgaard, experience.
J. Sensitization of hypoxic tumor cells. Clinical Int. J. Radiat. Biol. 56:801-811; 1989.
Wasserman, T. H.; Lee, D. J.; Cosmatos, D.; Coleman, N.; Phillips, T.; Davis, L.; Martial, V.; Stetz, J. Clinical trials with etanidazole (SR-2508) by the radiation therapy oncology group (RTOG). Radiother. Oncol. 2O(Suppl.):129-135; 1991.