Improving outcomes in gynaecological cancers?

I do not believe any forward looking clinician would disagree with the ideals set out in the Calman-Hine report”, says David Luesley in his commentary (pages 1061–1063) on the recent changes in gynaecological cancer services in the NHS. Cancer networks are being set up, with cancer units in district hospitals being involved primarily with diagnosis and cancer centres in teaching hospitals primarily with treatment. This movement to treat virtually all gynaecological cancers in a few specialist centres will imitate the rest of Western Europe and North America, and it is hoped that such reorganisation will result in better outcomes.

Gynaecologists in the United Kingdom agree with the ideals of the Calman-Hine report, but may be sceptical of the new proposals. The reorganisation of services for treating gynaecological cancer will be costly and ideally there should be proof of its benefit. Obviously a randomised trial is impossible and so we may have to rely on secular trends of mortality from gynaecological cancer in order to measure the worth of cancer centres, a scientific exercise fraught with confounding factors. However this is the sort of proof which makes us sure of the benefits of cervical screening, where observational studies consistently show a strong association between the introduction of near-universal screening and a significant decrease in mortality from cervical cancer.

Professor Luesley sees other problems with the new proposals, concerning the allocation of resources, the organisation of the new service and the training of sufficient gynaecological oncologists and their supporting staff. The uncertainty over resources may have been resolved by the UK government's recent allocation of extra funds for cancer services, but it is the organisation of the services which is Luesley's main concern. Unless cancer services are a regional or national responsibility, confusion over the distribution of these new funds by hospital trusts will occur. These funds should also be sufficient to recruit additional gynaecological oncologists and their supporting staff, and until the proper establishment is achieved interim arrangements should be made using gynaecologists with an interest in gynaecological cancer to provide the service. These considerations should take precedence over arbitrary and unrealistic targets imposed by the government.

The treatment of cancer of the ovary will become a major issue in the reorganisation of cancer services, especially if the trends shown by Adcola Olaitan and colleagues (pages 1094–1096) continue. The authors analysed the information held in the Thames Cancer Registry to investigate secular trends in the incidence of carcinoma of the ovary over thirty years. The Registry contained the details of over twenty thousand women with ovarian cancer, more than one third of whom were over 70 years old at diagnosis. Between 20 and 44 years of age no secular trend was discernible; between 45 and 59 years of age there was a decreasing trend; and over 70 years of age there was an increasing trend in the incidence of carcinoma of the ovary. The authors suggest that the decreasing trend in middle life is due to the protective effect of the oral contraceptive pill taken some years earlier; but they are unable to account satisfactorily for the increasing trend in elderly women. The provision of services to treat ovarian cancer must acknowledge this rising trend in older ages, for elderly frail women may not be able to withstand the rigours of radical surgery and combination chemotherapy as well as younger fit women.

Probably the greatest advance in the control of ovarian cancer will be the development of an efficient screening test. Usha Menon and colleagues (pages 1069–1074) carried out a case-control study of inhibin and activin concentrations in 27 women with epithelial carcinoma of the ovary and 54 controls. Activin A concentrations were higher in women with ovarian cancer; however from the information in Figure 1 activin A by itself will have limited efficiency as a screening test, but may have promise in combination with other tumour markers. A larger study is required to establish the place of activin A as a screening test for ovarian cancer, with estimation of its sensitivity, specificity, positive and negative predictive values and likelihood ratios.

The NHS cervical screening programme requires that more than 85% of all cervical biopsies should contain cervical intraepithelial neoplasia, a target that most colposcopy units fail to meet. It was this which made R. E. J. Howells and colleagues (pages 1075–1082) attempt to increase the prediction of a negative histological result, so as to avoid biopsy of the cervix. The authors analysed data from 452 women who underwent large loop excision of the transformation zone, and found that age more than 50 years, low grade cytological atypia and negative colposcopic findings independently predicted negative histology. With this information, however, the overall prediction of negative histology was little better, and this improvement was at the expense of under-treatment of several women with cervical intraepithelial neoplasia. The authors conclude that the present targets for negative histology are unrealistic.

These four papers emphasise the tension between the organisation of clinical services on the one hand and the need for these services to be founded upon clinical research on the other. It would be a mistake to concentrate on the relatively inefficient treatment of advanced carcinoma of the ovary at the expense of basic research on screening for early carcinoma of the ovary. Standards which involve radical treatment of carcinoma of the ovary may be unsuitable in elderly women. And standards which are set arbitrarily may result in inappropriate treatment, under-treatment or over-treatment.