Results from the Kentucky Ovarian Cancer Screening Study at first glance look incredibly promising. Among the over 37 thousand women who underwent annual pelvic sonograms, the 5-year survival rate for all women with ovarian cancer in the screened group was 75% compared with 54% for unscreened women with ovarian cancer from the same institution treated exactly the same otherwise. The investigators attribute this increased survival to earlier detection – 70% of the screened group were diagnosed at stage I or II, compared with only 27% in the un-screened group. Stage III cancers tended to be earlier (IIIa and IIIB instead of IIIC), and there were no stage IV cancers among women who were screened.
The investigators markedly improved on the positive predictive value of screening by boldly refusing to go where others have always gone before – to the operating room. They stood firm and watched cysts grow to as large as 10 cm before intervening, provided those cysts did not bear the defining characteristics of malignancy – namely solid areas and papillary internal growths. They also were not afraid to tweek their triage algorithm as experience with sonography improved. This is perhaps the biggest contribution from the study – permission to watch and wait.
Following a mean of 5.5 screens in 37,293 women, the authors achieved a specificity of 98.5% and a PPV of 8.9% with 11.1 operations per case of primary invasive epithelial ovarian cancer. This compares with a specificity of 98.4% and 19.5 operations per case of primary invasive epithelial ovarian cancer in the Prostate, Lung, Colorectal and Ovarian Cancer Screening trial, in which both ultrasonography and CA 125 were used as first-line tests.
But a closer look reveals important questions that must be answered before we can begin to recommend screening in the general population.
1. Could the results be explained by the healthy volunteer effect? This was not a randomized trial, just a comparison between women in the screening program and the rest of the population who got ovarian cancer in the same time frame outside the program. We all know that folks who volunteer for studies such as this tend to be healthier in general than the overall population, thus skewing survival statistics in their favor. In this study, however, survival was equivalent between control and screened groups diagnosed in early stages, suggesting that it was indeed the stage shift that led to higher survival in screened groups and not just a healthy volunteer effect.
2. How about lead time effect? This happens when cancer is identified a little earlier, giving the false impression that folks are living longer when it is really that they have just learned a little earlier about the diagnosis that ultimately will lead to their demise. All screening studies have this potential bias. This is why overall mortality and not just survival time must be the relevant statistic to compare between screened and unscreened groups.
3. Not all cancers were caught by sono. Twelve women developed cancer in the year after a normal screening test, with 7 deaths due to cancer in this group. Such aggressive tumors may never lend themselves to early detection, no matter what modality is used.
4. Major surgery remains the only way to ultimately diagnose ovarian cancer. In the Kentucky trial, 523 women, or about 1.4% of participants screened ended up in the OR, and 86% of these women did not have cancer. Until we have a less invasive was to get reliable pathology on ovarian cysts, we are going to be exposing healthy women to unnecessary surgery while chasing the elusive early diagnosis. While this may be marginally acceptable in high risk women, expanding screening to the general population will lead to millions of avoidable operations, with their consequent risks, costs and mortality.
___________________________________________________________________________
Long-term survival of women with epithelial ovarian cancer detected by ultrasonographic screening. van Nagell JR, Miller RW, DeSimone CP, Ueland FR, Podzielinski I, Goodrich ST, Elder JW, Huang B, Kryscio RJ, Pavlik EJ Obstet Gynecol. 2011 Dec; 118(6):1212-21
Jacobs,I; Menon,U. Can Ovarian Cancer Screening Save Lives? The Question Remains Unanswered. Obstet & Gynecol. 118(6):1209-1211, December 2011.
