Introduction Ovarian cancer is the leading cause

Introduction Ovarian cancer is the leading cause of death among gynecologic malignancies, chiefly due to the large proportion of cases diagnosed at advanced stage [[1], [2], [3]]. Optimal debulking (<1 cm residual disease) is an important predictor of survival for patients with advanced ovarian cancer, and an additional survival benefit is conferred for those with resection to no gross residual disease [[4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14]]. The standard treatment for advanced ovarian cancer in the U.S. has traditionally been primary debulking surgery (PDS) – which aims to remove the majority of disease upfront – followed by adjuvant chemotherapy. However, optimal debulking often requires aggressive surgery that is associated with substantial morbidity, including complications that can delay chemotherapy [11,[15], [16], [17]]. Neoadjuvant chemotherapy (NACT) followed by interval debulking surgery and adjuvant chemotherapy has become an increasingly accepted alternative to PDS for advanced ovarian cancer [[18], [19], [20]]. NACT aims to reduce tumor burden and improve performance status preoperatively. This decreases the need for aggressive procedures, reduces post-operative complications, and increases the probability of optimal debulking [17,21,22]. However, it Conessine is unclear whether NACT provides a survival benefit. In the U.S., the role of NACT in advanced ovarian cancer remains controversial [[23], [24], [25], [26], [27], [28], [29]]. Although two randomized controlled trials (RCTs) outside the U.S. observed non-inferior survival with NACT versus PDS [10,30], the generalizability of these trials to U.S. practice has been questioned due to lower median overall survival and lower rates of optimal debulking compared to U.S. populations [28,29]. A less aggressive surgical effort compared to that commonly performed in the U.S. has been postulated as the cause of the observed non-inferiority results [27,[24], [25], [26], [27], [28], [29]]. Among U.S. gynecologic oncologists surveyed in 2010, 82% did not believe there was sufficient evidence to justify NACT [31]; conversely, a European survey found that 70% of gynecologic oncologists believed there was sufficient evidence for use of NACT in advanced ovarian cancer [32].
Materials and methods We conducted a PubMed search on February 5th, 2017 using keywords selected based on prior knowledge and examination of previously-identified studies (Fig. 1). Eligible studies met the following criteria: (1) compared the effectiveness of NACT and PDS on overall survival for advanced ovarian, fallopian, or primary peritoneal cancer; (2) conducted in the U.S.; (3) published in English, and (4) published 2010 or later (reviews published in 2007 and 2011 outline evidence before 2010) [33,34]. We screened titles and abstracts identified through the electronic search for eligibility. The full text of potentially-eligible articles was then reviewed to confirm eligibility.
Results Nine observational studies met eligibility criteria for review (Table 1) [19,20,[35], [36], [37], [38], [39], [40], [41]]. Data from 1991 [20] through 2013 [41] were used to conduct these studies. Most studies included women with stage IIIC-IV ovarian cancer [19,36,37,39,41], one included only stage IV [35], and others included earlier stage disease [20,38,40]. Four studies included only NACT patients who received interval debulking surgery [35,37,39,41]. Three studies [20,36,40] used nationally-representative registry data, while the remaining studies used institutional medical record data. All studies compared overall survival between patients treated with NACT versus PDS using hazard ratios and/or Kaplan-Meier survival estimates, except one study [40] that used incidence rate ratios (Table 1). Studies used various methods to control for confounding, including multivariable adjustment [20,35,[39], [40], [41]], propensity score methods [19,20,36], instrumental variable analysis [20], stratification [19,36,37,40], and restriction [36]. Two studies did not report any confounder control when comparing overall survival [37,38]. Though measured confounders were listed for most studies, covariates used for confounder control were unclear for some studies [20,35,39,41]. Four studies conditioned on residual disease, a variable on the causal pathway and one that is not known preoperatively when treatment decisions are being made [19,35,39,41].