Published recently in : 2010 World Endometriosis Society World Endometriosis Society e-Journal Volume 12 No 1, 2010 8 ; editor L. Rombauts.
The association between ovarian cancer and endometriosis has become highlighted over the last years and the last eJournal adequately details the available data. The interpretation of the data with RR ratios and confidence intervals, however, unfortunately leaves a feeling of ‘How to lie with statistics’. We know the limitations of cohort studies and of associations. Since our interpretation of data as emerging from the published articles seems at least insufficiently prudent, biased and incomplete – if not incorrect – and potentially harmful, inducing fear in women with endometriosis, some discussion seems appropriate. Biochemical similarities Often biochemical similarities between endometriosis and cancer are described (Prowse et al, 2006; Mandai et al, 2009).
The pathophysiology of endometriosis clearly is not well understood. Considering typical, cystic, and deep endometriosis as a benign tumour caused by a genomic incident (Koninckx et al, 1997) (and subtle lesions as a physiologic condition occurring intermittently in all women) remains the easiest way to explain hereditary influences and effects of dioxins and total body radiation. It, moreover, could explain the many described differences in the endometrium of women with and without endometriosis: the slight differences could point to some genetic disturbances making them more susceptible to develop endometriosis. The Sampson hypothesis seems attractive since viable endometrial cells, with implantation capacity are found in the peritoneal fluid of most women. To the best of our knowledge, however, a progression from subtle to typical to cystic or deep lesions has never been observed nor demonstrated. The Sampson hypothesis, and the metaplasia theory which is a variant, can explain implantation but should be considered as speculation since it cannot, does not even attempt to, explain progression. If, however, endometriosis is considered a benign tumour, and if the endometrial differences signal some genetic predisposition (and are not a consequence of the existing endometriosis) it is not surprising that some biochemical similarities exist between endometriosis and ovarian cancer and cancer in general. Association statistics
The case control study of Brinton et al (1997) finds an odds ratio of 1.9 to more than 4 for developing ovarian cancer in women with endometriosis or ovarian endometriosis respectively, a conclusion based upon a Swedish hospital discharge diagnosis of endometriosis. The study of Melin et al (2006) finds an odds ratio of 1.2 equally based on a hospital discharge diagnosis of endometriosis. It can be anticipated that this group of women, in comparison with women without a diagnosis of endometriosis, had had more interventions, more pelvic pain, more medical treatments of endometriosis and probably more infertility treatments. Having been diagnosed with ovarian endometriosis means probable surgery of the ovary, very often focal coagulation leaving behind at least carbon deposits. It also means more adhesion formation etc. It is intellectually unfair to pick one aspect only –endometriosis– for the comparison of two groups of women who differ in many other aspects. This indeed is a fundamental problem of correlation and association statistics. Logistic regression and statistical model building were developed to find out which are independent and which are important factors. This obviously has not been done for the association of endometriosis and cancer. A second bias might have been introduced when analysing national discharge statistics. In most women ovarian endometriosis is diagnosed by ultrasound; and in a recent IOTA review (Van Holsbeke, in press) of over 2500 cysts, a risk of an (borderline) ovarian cancer was found in 1%. This could seem to support the argument, but analysing the data in detail, this is true only for women after menopause, a period of life that seems hardly associated with endometriosis. It would not be surprising if many women who did not have ovarian endometriosis by pathology ended up with a primary hospital discharge diagnosis of endometriosis. Tubal ligation Tubal ligation is associated with a decreased risk of ovarian cancer, a lower incidence of PID, and absence of retrograde menstruation. Tubal ligation, however, also influences in some women ovarian function and blood supply at least as evidenced by irregular cycles. Today we do not know how to explain exactly the decreased incidence of ovarian cancer in these women. The absence of retrograde menstruation does not seem logical as an explanation; indeed women with severe cystic ovarian endometriosis will have a higher incidence of infertility and thus a lower probability of undergoing tubal sterilisation. Accuracy of diagnosis It is surprising that the diagnosis of endometriosis is so easily accepted in the articles using hospital discharge records, knowing that the diagnosis of endometriosis is bound to have biases. Whereas for typical and deep lesions the probability of histological confirmation is very high, the rate of histological confirmation of subtle lesions is low, mainly dependent upon the expertise of the surgeon and the pathologist. Also for cystic ovarian endometriosis the histological confirmation of endometriosis, defined as stroma and glands, is not that well established and often pathology returns as ‘compatible with endometriosis’; most of these women, however, will be discharged as having endometriosis. Especially after menopause the diagnosis of endometriosis risks to be erroneous. Indeed, the occurrence of ‘cystic ovarian endometriosis’ or ‘large cystic ovarian endometriosis’ in women who did not have an existing cystic ovarian endometriosis before menopause is something rare and suspicious by definition. From the available data it is difficult to judge how strict the diagnosis of endometriosis was. Anyway, it would not be surprising, that many of these women were erroneously diagnosed as having had endometriosis. For recent prospective studies where women were followed by serial ultrasound Kobayashi et al (2009, 2008) the bias is obvious given that the ultrasound diagnosis of endometriosis has a specificity of 88% only. Moreover in the group of the 12% false positives the ovarian cancer incidence was 18% after menopause versus 0.6% before menopause (Van Holsbeke, in press). It is moreover questionable not to perform surgery in menopausal women with an ultrasound diagnosis of endometriosis. Menopause Unopposed oestrogen administration following menopause is associated with an increased risk of endometrial hyperplasia and adenocarcinoma. Following menopause, existing endometriosis will become inactive but will not disappear. Considering the prevalence of endometriosis in women before menopause it is very surprising that oestrogen-only therapy following menopause (in women without an endometrium) is not associated with an epidemic of adenocarcinomas originating from the endometriosis. This even looks as if endometriotic cells, although hormonally responsive, are pretty resistant to malignant transformation. Metastatic behaviour Especially deep endometriosis has invasive and „metastatic‟ behavior because of the positive lymph nodes. In our series of over 2000 deep nodules, only one however has turned out to be malignant, and this in a postmenopausal woman (unpublished data). In conclusion, the data describing an association between endometriosis, in particular cystic ovarian endometriosis, and ovarian cancer can be interpreted in many different ways. Today we do not consider the data sufficient to consider women with endometriosis or cystic endometriosis at risk for developing ovarian cancer, a conclusion already reached by others (Somigliana et al, 2006). It reminds us of the fact that so many ovaries have been removed prophylactically after age 50 to find out recently that women with ovaries live longer than women without ovaries (Parker et al, 2009)
„It ain’t so much the things we don’t know that gets us into trouble. It’s the things we do know that just ain’t so‟ Josh Billings
Philippe R Koninckx and Anastasia Ussia
KULeuven, Belgie en Gruppo Italo Belgo, Roma Italia
REFERENCES Brinton LA, Gridley G, Persson I, Baron J, Bergqvist A. Cancer risk after a hospital discharge diagnosis of endometriosis. Am J Obstet Gynecol 1997; 176(3):572-579. Kobayashi H, Sumimoto K, Kitanaka T, Yamada Y, Sado T, Sakata M et al. Ovarian endometrioma–risks factors of ovarian cancer development. Eur J Obstet Gynecol Reprod Biol 2008; 138(2):187-193. Kobayashi H. Ovarian cancer in endometriosis: epidemiology, natural history, and clinical diagnosis. Int J Clin Oncol 2009; 14(5):378-382. Koninckx PR, Barlow D, Kennedy S. Implantation versus infiltration: the Sampson versus the endometriotic disease theory. Gynecol Obstet Invest 1999; 47 Suppl 1:3-9. Mandai M, Yamaguchi K, Matsumura N, Baba T, Konishi I. Ovarian cancer in endometriosis: molecular biology, pathology, and clinical management. Int J Clin Oncol 2009; 14(5):383-391. Melin A, Sparen P, Persson I, Bergqvist A. Endometriosis and the risk of cancer with special emphasis on ovarian cancer. Hum Reprod 2006; 21(5):1237-1242. Parker WH, Broder MS, Chang E, Feskanich D, Farquhar C, Liu Z et al. Ovarian conservation at the time of hysterectomy and long-term health outcomes in the nurses’ health study. Obstet Gynecol 2009; 113(5):1027-1037. Prowse AH, Manek S, Varma R, Liu J, Godwin AK, Maher ER et al. Molecular genetic evidence that endometriosis is a precursor of ovarian cancer. Int J Cancer 2006; 119(3):556-562. Somigliana E, Vigano P, Parazzini F, Stoppelli S, Giambattista E, Vercellini P. Association between endometriosis and cancer: a comprehensive review and a critical analysis of clinical and epidemiological evidence. Gynecol Oncol 2006; 101(2):331-341. Vanholsbeke C, Van Calster B, Guerrierro S, Savelli L, Leone F, Fischerova D, Czekierdowski A, Fruscio R, Veldman J, Van De Putte G, Testa AC, Bourne T, Valentin L, Timmerman D. Imaging in gynaecology: How good are we in identifying endometriomas? F, V & V in ObGyn, 2009, 1 (1): in press.