Fertility treatment and cancers-the eternal conundrum: a systematic review and meta-analysis.
Barcroft JF., Galazis N., Jones BP., Getreu N., Bracewell-Milnes T., Grewal KJ., Sorbi F., Yazbek J., Lathouras K., Smith JR., Hardiman P., Thum M-Y., Ben-Nagi J., Ghaem-Maghami S., Verbakel J., Saso S.
STUDY QUESTION: Does fertility treatment (FT) significantly increase the incidence of breast, ovarian, endometrial or cervical cancer? SUMMARY ANSWER: Overall, FT does not significantly increase the incidence of breast, ovarian or endometrial cancer and may even reduce the incidence of cervical cancer. WHAT IS KNOWN ALREADY: Infertility affects more than 14% of couples. Infertility and nulliparity are established risk factors for endometrial, ovarian and breast cancer, yet the association with FT is more contentious. STUDY DESIGN, SIZE, DURATION: A literature search was carried out using Cochrane Library, EMBASE, Medline and Google Scholar up to December 2019. Peer-reviewed studies stating cancer incidence (breast, ovarian, endometrial or cervical) in FT and no-FT groups were identified. Out of 128 studies identified, 29 retrospective studies fulfilled the criteria and were included (n = 21 070 337). PARTICIPANTS/MATERIALS, SETTING, METHODS: In the final meta-analysis, 29 studies were included: breast (n = 19), ovarian (n = 19), endometrial (n = 15) and cervical (n = 13), 17 studies involved multiple cancer types and so were included in each individual cancer meta-analysis. Primary outcome of interest was cancer incidence (breast, ovarian, endometrial and cervical) in FT and no-FT groups. Secondary outcome was cancer incidence according to specific fertility drug exposure. Odds ratio (OR) and random effects model were used to demonstrate treatment effect and calculate pooled treatment effect, respectively. A meta-regression and eight sub-group analyses were performed to assess the impact of the following variables, maternal age, infertility, study size, outliers and specific FT sub-types, on cancer incidence. MAIN RESULTS AND THE ROLE OF CHANCE: Cervical cancer incidence was significantly lower in the FT group compared with the no-FT group: OR 0.68 (95% CI 0.46-0.99). The incidences of breast (OR 0.86; 95% CI 0.73-1.01) and endometrial (OR 1.28; 95% CI 0.92-1.79) cancers were not found to be significantly different between the FT and no-FT groups. Whilst overall ovarian cancer incidence was not significantly different between the FT and no-FT groups (OR 1.19; 95% CI 0.98-1.46), separate analysis of borderline ovarian tumours (BOT) revealed a significant association (OR 1.69; 95% CI 1.27-2.25). In further sub-group analyses, ovarian cancer incidence was shown to be significantly higher in the IVF (OR 1.32; 95% CI 1.03-1.69) and clomiphene citrate (CC) treatment group (OR 1.40; 95% CI 1.10-1.77), respectively when compared with the no-FT group. Conversely, the incidences of breast (OR 0.75; 95% CI 0.61-0.92) and cervical cancer (OR 0.58; 95% CI 0.38-0.89) were significantly lower in the IVF treatment sub-group compared to the no-FT group. LIMITATIONS, REASONS FOR CAUTION: The large, varied dataset spanning a wide study period introduced significant clinical heterogeneity. Thus, results have to be interpreted with an element of caution. Exclusion of non-English citations, unpublished work and abstracts, in order to ensure data accuracy and reliability was maintained, may have introduced a degree of selection bias. WIDER IMPLICATIONS OF THE FINDINGS: The results for breast, ovarian, endometrial and cervical cancer are reassuring, in line with previously published meta-analyses for individual cancers but the association between IVF and CC treatment and an increase in ovarian cancer incidence requires additional work to understand the potential mechanism driving this association. In particular, focusing on (i) discriminating specific treatments effects from an inherent risk of malignancy; (ii) differential risk profiles among specific patient sub-groups (refractory treatment and obesity); and (iii) understanding the impact of FT outcomes on cancer incidence. STUDY FUNDING/COMPETING INTEREST(S): This study did not receive any funding. The authors have no financial, personal, intellectual and professional conflicts of interest to declare. PROSPERO REGISTRATION NUMBER: CRD42019153404.