Br. J. Cancer (1991), 64, 775 779 © Macmillan 1991 A Medical Research Council phase II trial of alternating chemotherapy and radiotherapy in small-cell lung cancer N.M. Bleehen', D.J. Girling', A. Gregor2, R.C.F. Leonard2, D. Machin', C.G. McKenzie3, D.A.L. Morgan4, J.F. Smyth2, M.F. Spittle5, R.J. Stephens', H.M.A. Yosef6, on behalf of the Medical Research Council Lung Cancer Working Party* 'MRC Clinical Oncology and Radiotherapeutics Unit, Addenbrooke's Hospital, Hills Road, Cambridge; 2Department of Clinical Oncology, Western General Hospital, Crewe Road, Edinburgh; 3Department of Radiotherapy and Oncology, Hammersmith Hospital, Du Cane Road, London; 4Hogarth Centre of Radiotherapy and Oncology, General Hospital, Park Row, Nottingham; 5Department of Radiotherapy and Oncology, Middlesex Hospital, Mortimer Street, London; and 6Belvidere Hospital, Glasgow, UK. Summary In a non-randomised study in six centres in the UK, 24 patients with previously untreated small-cell lung cancer of limited extent were treated with a regimen of alternating chemotherapy and radiotherapy to assess response, toxicity, and the feasibility of applying such a regimen on a multicentre basis in the UK. The intention was to give six courses of chemotherapy on five consecutive days at 4-week intervals: etoposide 75 mg m2 on days 1, 2, and 3; doxorubicin 40mg m2 on day 1; cisplatin 100mg m2 on day 2; and cyclophosphamide 300 mg m-2 on days 2, 3, 4 and 5. A dose of 20 Gy thoracic radiotherapy was to be given following the 2nd and the 3rd courses, and one of 15 Gy following the 4th course. After 12 patients had been admitted, the cisplatin dosage was reduced to 80 mg m-2 because of unacceptable toxicity. Two patients were withdrawn during treatment on review of their histology because their diagnosis was found to be incorrect. Only one patient of the 12 treated with cisplatin 100 mg m2 was able to complete treatment, compared with five of the eligible ten given the lower dosage. Among the 22 patients with confirmed small-cell disease, a complete response was reported in 14 (64%) and a partial response in a further three (total response rate 77%). Myelosuppression was the commonest serious adverse effect. It occurred in 19 of the 24 patients and gave rise to septicaemia in five, four of whom were receiving the higher cisplatin dose. Sixteen patients required blood transfusion and ten platelet transfusion. Vomiting, oesophagitis, and peripheral neuropathy occurred in 12, four and four patients, respectively, and radiation pneumonitis developed in two. Treatment was considered a contributory cause of death in four. The working party concluded that the alternating regimen was feasible in only a small proportion of centres in the UK, and decided not to embark on a multicentre randomised trial comparing alternating with conventional scheduling. Small-cell lung cancer responds well to combination chemo- therapy (Seifter & Ihde, 1988). Objective response rates (World Health Organization, 1979) of around 80% are typi- cal in published reports, as are median survival times of approximately 12 months in patients with limited disease and 6 months in those with extensive disease (Leonard, 1989). In patients with limited disease, the inclusion of thoracic radio- therapy in the treatment regimen both improves local control of the cancer and prolongs survival (Bleehen, 1986; Arria- gada et al., 1989a). Nevertheless, 3-year survival rates are low and the great majority of patients die from their lung cancer. At the time this study was planned, however, not only high response rates but also substantial 2-year and 3-year survival rates were being reported by Arriagada and his colleagues in non-randomised phase II trials using regimens of alternating chemotherapy and radiotherapy. Thirty-five patients less than 70 years of age, with small-cell lung cancer of limited extent, and good performance status, were treated with a regimen of three doses of mediastinal radiotherapy (total dose, 55 Gy) and six courses of chemotherapy using doxorubicin, etopo- side, cyclophosphamide, and cisplatin. The radiotherapy was given following the 2nd, the 3rd, and the 4th courses of Correspondence: D.J. Girling, MRC Cancer Trials Office, 1 Brook- lands Avenue, Cambridge CB2 2BB, UK. *Members: N.M. Bleehen (Chairman until October 1989), J.J. Bol- ger, D.J. Girling (Secretary), P.S. Hasleton, P. Hopwood, F.R. Macbeth, D. Machin (Statistician), K. Moghissi, M.I. Saunders, R.J. Stephens, N. Thatcher (Chairman from October 1989). Received 4 April 1991; and in revised form 11 June 1991. chemotherapy which were given at 4-week intervals (Arria- gada et al., 1985a). The complete response rate, broncho- scopically confirmed, was 91%, the local recurrence rate was 22%, and the relapse-free survival rate at 2 years was 32%. Haematological toxicity, oesophagitis, and infectious bronchopneumonia were common, but were considered acceptable. Subsequently, in 109 similarly treated patients (Le Chevalier et al., 1987), the complete response rate was 79%, the local recurrence rate was 25%, and the survival rate at 3 years was 26%. Lethal toxicity was reported in 3% of the patients. It therefore appeared that alternating scheduling might improve long-term survival rates, although the patients selected for the above trials were a group known to have a relatively good prognosis (Rawson & Peto, 1990). The rationale for alternating the chemotherapy and radio- therapy is to make optimum use of these two modalities from the start of treatment, without incurring the unacceptable levels of toxicity that have been reported when the two are given concurrently (Arriagada et al., 1985b; 1989a; Bunn et al., 1987). Early use of both modalities rapidly reduces tumour bulk, and is presumed to lessen the risk of the emergence of resistant cells. Early use of chemotherapy ensures that occult distant metastases are suppressed from the start, and an alternating schedule allows the chemo- therapy to be given without interruption in regular (4-weekly) courses. The present trial was conducted to determine whether the regimen used by Arriagada and his colleagues was logistically feasible in centres in the United Kingdom, whether the toxi- city was acceptable, and whether high complete response rates could be achieved. The intention was that a large multicentre randomised trial to compare alternating with conventional scheduling should then be considered. Br. J. Cancer (1991), 64, 775-779 4" Macmillan Press Ltd., 1991