Further analyses of the material contained in trials of the hyperbaric oxygen chamber in the radiotherapy of carcinoma of the cervix have shown that patients who were severely anaemic prior to radiotherapy, and who required blood transfusion, showed very poor local tumour control when conventionally treated after transfusion, but very good local tumour control when treated in hyperbaric oxygen. The finding of a special sub-group where hypoxia would seem to be an important cause of radiation failure, and where hyperbaric oxygen was successful in overcoming it, may have importance in the evaluation of other methods for overcoming the hypoxia, including the use of chemical sensitising agents.

No abstract

In a randomized controlled clinical trial of hyperbaric oxygen in the radiotherapy of advanced carcinoma of the uterine cervix a total of 320 cases were contributed by four radiotherapy centres in the United Kingdom. The use of hyperbaric oxygen resulted in improved local control and survival. The benefit was greatest in patients under the age of 55 who presented with stage III disease. There was a slight increase in radiation morbidity but it seemed that the benefit of hyperbaric oxygen outweighed this increase in morbidity and that there was a true improvement in the therapeutic ratio.

In a randomized controlled clinical trial of hyperbaric oxygen in the radiotherapy of carcinoma of the bladder a total of 241 cases were contributed by four radiotherapy centres in the United Kingdom. In this trial where in each centre identical radiotherapy was employed for both oxygen and air cases, no benefit was shown with the use of hyperbaric oxygen.

Summary Forty patients with stage III and 15 patients with stage IVa carcinoma of cervix have been treated with two pulses of cisplatin, vincristine and bleomycin combination chemotherapy before full dose radical radiotherapy. Twenty-seven of 51 (53%, 95% confidence interval 40-67%) had an objective response to chemotherapy and all chemotherapy responders had a complete response to radiotherapy. The actuarial survival at 24 months of responders to chemotherapy is 71% against 37% for non-responders. The responding patients had an estimated reduction in mortality to 36% (P=0.014, 95% CI 15-81%) of that of the non-responders to chemotherapy. The incidence of tumour recurrence or distant metastases showed a similar reduction to 34% (P=0.006, 95% CI 14-73%) of that of the non-responders. The data strongly suggest that response to chemotherapy in the initial treatment of advanced cervical cancer is associated with an improvement in survival following subsequent radical radiotherapy.Both clinical studies (Shukovsky & Fletcher, 1973) and animal experiments (Fowler et al., 1963) have shown that radiation doses that will consistently eradicate small tumours will only be effective in a minority of larger lesions. The maximum radiation dose that can be given to patients with advanced carcinoma of cervix is limited by the tolerance of surrounding organs, such as the bladder and bowel.Even employing what are currently regarded as optimal fraction policies in combination with intracavity treatment, local control is achieved in only about half of patients with advanced carcinoma of cervix. Typical local recurrence-free rates are 51% for stage III (Adams & Kerby, 1983) and 43% for stage IVa patients (Upadhyay et al., 1988). The relative radio-resistance of bulky tumours may in part be due to the presence of hypoxic cells within the tumour as radiotherapy in hyperbaric oxygen increased local control in one large randomised series (Watson et al., 1978), but other factors could also be important. Local control using conventional radiotherapy schedules may be increased if the tumour size could be reduced by chemotherapy given before radiotherapy. Advanced carcinoma of cervix may be particularly susceptible to this approach as the tumour is moderately chemosensitive (Guthrie, 1985) and a proportion of lesions which are locally far advanced may not have spread beyond the true pelvis.As a first step to testing this hypothesis we carried out a pilot study of cisplatin-based chemotherapy given before full dose radiotherapy. Materials and methodsThe first 55 patients who entered this pilot study have been followed up for up to 51 months since the end of treatment (median follow-up 21 months). Forty patients were assigned to stage III and 15 to stage IVa; all had squamous carcinomas. All the relevant biopsy material has been retrospectively reviewed (without knowledge of which patients responded to chemotherapy) using the grading system of Reagan & Wentz (Gunderson et al., 1974). The mean age of patients was 49 years (range 29-70) and all ...