Systemic Scleroderma with Portal Hypertension

whenever feasible. Failing this, we regard high C.I. readings after completion of radical treatment (allowing a threemonths period to elapse in the evaluation of pre-menopausal cancer patients after radiotherapy) as suggestive of persistence of neoplastic disease, provided, of course, that no oestrogens had been administered. A low C.I. reading during the observation period after completion of treatment is regarded as a favourable prognostic omen only in cases in which pre-treatment C.I.s had been high. It is not yet established whether our original hypothesis, that malignant tissue is capable of oestrogen production. is correct. In a few of our own cases urinary oestrogen estimations have been undertaken, and we are greatly indebted to Dr. Ian Ferguson Sommerville for kindly carrying out these investigations at the Endocrine Laboratories of the Institute of Obstetrics and Gynaecology in the Chelsea Hospital for Women. In the few patients investigated urinary oestrogen excretion was high and compared well with the high C.I. readings. However, these data are as yet incomplete and no conclusions can be drawn from them, although other authors (Young et al., 1957) showed similar good correlation on a larger material. The frequency with which a high C.I. reading is associated with malignant disease in the female genital tract is significant enough to suggest the use of it as a test for cure. Summary and Conclusions C.I. estimations in patients who had received radical treatment for malignant disease in the genital tract seemed to be of great diagnostic help in the recognition of failure of response or in the detection of early recurrence. Accurate forecasts, relying solely on C.I. readings, had been made in 94.8% of 165 cancer followup patients. Nine patients with a poor cytological prognosis and no clinical signs of recurrence are still under observation. The cytological prognosis was incorrect in eight patients with persistent or recurrent disease and low C.I. readings. The possible causes of this prognostic error are discussed. The technique of C.I. estimation is easy and can be learned in a short time. We suggest that C.I. estimations should be included in routine cytological cancer followup investigations in view of the important implication of this test.