Reply

Reply Linked article: This is a response to the comment by Vercellini et al. Click here to view the Correspondence. We are grateful to Vercellini and colleagues for their interest in our paper and for their thoughtful comments. They have voiced several concerns regarding potential risks associated with expectant management of patients diagnosed with ovarian endometriomas that merit further discussion. Vercellini and colleagues are correct in saying that a significant proportion of women with endometriosis were excluded from our analysis. However, most patients were excluded because of the study design, which stipulated the requirement for long-term follow-up by a single examiner. Although it is true that many symptomatic women opt for active treatment, there is also a tendency to offer surgery or medical treatment to asymptomatic or minimally symptomatic women. This is done mainly on the premise that timely intervention could stop endometriosis from progressing and becoming symptomatic and is conducted less often for the purpose of preventing malignant transformation. The aim of our study1 was to determine whether endometriosis is likely to progress with time. Our sample size was not large enough to assess the risk of malignant transformation. We agree with Vercellini and colleagues that our study population was highly selective, and it is true that asymptomatic women account for a minority of endometriosis patients. However, we believe that they are an important group to study because, in the absence of symptoms, the hypothetical benefits of treatment should be balanced carefully against well-documented risks of both medical and surgical treatment. Bearing in mind the strong association between endometriomas and pelvic adhesions, surgery in these patients is often complex, with an increased risk of complications. In addition, surgical treatment of ovarian endometriomas may lead to further compromise of ovarian function and impact negatively on ovarian reserve2. Vercellini and colleagues also suggest that deep endometriotic lesions should be taken into account when assessing disease progression. Although this is a valid remark, there are currently no uniform criteria by which endometriosis progression can be defined. In our study, the behavior of endometriomas over time was of primary interest. The criteria used to define progression were based on the change in size and derived from our previous study of inter- and intraobserver variability in the measurement of endometriomas3. We should also add that we have reported previously how deep endometriotic nodules behave over time in women presenting with mild or no symptoms. We found that approximately one-third of these women experienced an increase in the number or size of deep nodules at follow-up4. The risk of malignant transformation of endometriomas, which was also articulated by Vercellini and colleagues, has been studied extensively in the past. They highlighted the study by Barnard et al., which reported a ten-fold increase in the risk of ovarian cancer in women with endometriosis compared to women without endometriosis5. This was much higher than the summary relative risk of 1.93 (95% CI, 1.68–2.22) reported in a systematic review and meta-analysis by Kvaskoff et al.6 The estimated prevalence of endometriosis in the study by Barnard et al.5 was only 6.3%, which is much lower compared to more recent outpatient ultrasound studies that showed that deep endometriosis was present in nearly 20% of women attending general gynecological clinics7. If superficial endometriosis was detectable on ultrasound, the reported overall prevalence would probably be even higher. This apparent underestimation of the prevalence of endometriosis in their population may explain the higher-than-expected risk of ovarian cancer found by Barnard et al.5. However, even if this hazard ratio is accepted as correct, that would translate into only 10 additional cases of ovarian cancer per 10 000 women over a mean of 12 years. Considering the associated risks and cost, it is very unlikely that prophylactic surgical bilateral salpingo-oophorectomy could be implemented into routine clinical practice in the foreseeable future. Additionally, Vercellini and colleagues refer to a study showing association of long-term use of hormonal contraception and reduced risk of ovarian cancer8. The risk reduction was highest in long-term (> 10 years) users of contraception. However, this benefit is offset by a higher risk of breast cancer, which increases with duration of use9. Another study showed an increased risk of cervical cancer with use of hormonal contraception, which also increased with length of use10. Bearing in mind that the incidence of breast cancer is several times higher compared with ovarian cancer, any benefit in preventing ovarian cancer could potentially be offset by a much greater number of additional treatment-induced cases of breast cancer. The lack of sound evidence to support the routine use of any active management strategy to prevent ovarian cancer in women with endometriosis has been summarized in the current ESHRE guideline on endometriosis11. Overall, the findings of our study1 provide evidence that, in many women, endometriosis will remain clinically silent and not progress with time. In this specific group of asymptomatic women, active management is unlikely to offer any significant benefits, and the risks of treatment probably outweigh potential benefits. Until more evidence is available, we agree that appropriate counseling regarding different management options and shared decision-making with patients is the best practice.