Background: Few data on the performance of screening strategies for endometrial cancer exist, perhaps because long term survival of >70% can be achieved with current treatment at the time of clinical diagnosis. However, this tumour is now the commonest gynaecological malignancy and the incidence continues to relentlessly rise worldwide. Furthermore the greatest success of gynaecological oncology remains the dramatic reduction in the burden of cervical cancer that has occurred in countries able to implement a successful screening strategy.
Methods: During UKCTOCS, the primary endpoint of which was mortality from ovarian cancer, sonographers performing annual TVS also recorded the endometrial thickness (ET) in the participants, all of whom were over 50 years. 3,646 women with an intact uterus were recruited to the ultrasound arm at the Portsmouth centre of the study. Individuals found to have an ET > 10mm were offered a hysteroscopy and biopsy.
Results: In the first 8 years of the trial 19,886 scans were performed. 250 individual women had a reading of >10mm. In 18 of these biopsy revealed invasive cancer, one of which was a secondary from a primary breast tumour. Eight endometrioid endometrial cancers were diagnosed during the prevalence screen and 9 stage Ia endometrioid tumours during subsequent incidence screen scanning. Only 5 of these women reported bleeding prior to their TVS, 12 were asymptomatic. Median follow up of these 17 women with primary endometrial cancer is now 8.8.years from diagnosis. One death has occurred from a well documented cerebrovascular event with no evidence of metastatic disease being present. No one has died of endometrial cancer in this group of 17 women, whereas perhaps 4 deaths might be predicted from a similar number of a series of clinical patients.
Conclusions: These survival results could be explained by a predisposition to develop a better prognosis variant of endometrial cancer among women participating in UKCTOCS. Also women who accepted a random invitation to participate in a national screening trial may be better placed to look after their health following a diagnosis of cancer, than such similarly diagnosed patients in the general population. In addition it is possible that TVS cut off >10mm suffers from length bias as a screening test, in that slow growing low grade better prognosis lesions are preferentially detected. Alternatively, attending a screening centre annually for a scan might favourably attenuate the prognosis of endometrial cancer. Further investigation of this cohort in total and the randomised un screened control group of this trial for mortality from endometrial cancer would help clarify these possibilities.