Racial health disparities in ovarian cancer: not just black and white

Ovarian cancer (OC) is the most lethal of all gynecologic malignancies [1]. Globally, OC is the 7th most commonly diagnosed and the 8th leading cause of cancer-related mortality among women [2]. According to estimates by the American Cancer Society, 22,440 new OC cases and 14,080 associated deaths will occur in 2017 [1]. It is a deadly disease with no effective screening [3]. Approximately 70% of patients with OC are diagnosed at an advanced stage, with associated poor prognosis, even after aggressive and immediate treatments [4]. Studies over past several years have revealed that besides being highly lethal, OC also disproportionately affects some distinct racial populations, particularly the black women of African American (AA) heritage, as compared to the Caucasian Americans (CA) or white women of European heritage. In fact, even with higher incidence of OC in CA women in the United States, the associated mortality is disproportionately higher in AA women [5], and such disparities are common worldwide [6]. Though the exact causes of racial disparities in OC still remain unclear, they are likely to be multifaceted and may include socio-cultural factors, acquired co-morbid conditions, increased frequency of modifiable risk factors, access to health care, diet and preventive health factors. Emerging data suggests that several biological factors, such as genetic, epigenetic etc., could be more crucial than thought for health disparities in OC incidence and outcome [5, 7, 8]. In this review article, we have discussed, in-detail, various biological and non-biological factors in racially disparate clinical outcomes of OC.

While some progress has been made in OC treatment, it has been observed that AA women with OC are not reaping the same benefits of the advances as CA women. Progress in management of OC patients has improved over time but this progress has been relatively slow for AAs [9]. The all-cause mortality of AA OC patients is 1.3 times higher, as compared to CA OC patients, even when access to care is equal [5]. According to data from SEER database, between the years 1992 and 2008, the five-year survival rate for CA women rose from 40.7% to 45.0%, while the five-year survival for AA women fell from 47.9% to 40.3%. For the years 2006 through 2012, the reported 5 year survival of AA OC patients is relatively poor, compared to CA OC patients, irrespective of cancer stage at the time of diagnosis [1] (Table 1). Although the overwhelming majority of all OC patients are diagnosed at an advanced stage, it has been suggested that AA women bear a greater burden in the late diagnosis than CA women. Increased instances of late diagnosis may be attributed to socio-economic factors that will be discussed later in this article.

This disparity in survival may also be linked to observed prevalence of comorbidities. Evidence shows that AA women diagnosed with OC are much likely to have hypertension (75.5%), renal disease (58.5%) and cardiovascular disease (63.21%) [10]. Often, AA women, post-treatment, tend to have elevated CA-125 levels [10]. CA-125 is a diagnostic marker for OC and its elevated levels indicate a lingering OC, which could implicate higher recurrence and, possibly, reduced survival. Further, earlier normalization of CA-125 levels during primary chemotherapy for epithelial OC predicts improves progression free survival (PFS), overall survival (OS) as well as platinum sensitivity [11]. For each one-cycle improvement in CA-125 normalization, there is an estimated 3.8 months increase in PFS, 8.6 months increase in OS and 8% increase in platinum sensitivity.

With the realization of race-based OC health disparities [12], efforts are now being made to better understand the etiology and progression of OC in AA populations. For example, a recent study looked at the lifetime number of ovulatory cycles and epithelial OC risk in just AA patients [13]. The study found a positive association between lifetime ovulatory cycles and epithelial OC in AA women, similar to what was already known in CA patients. The association of antioxidant selenium uptake and risk of OC in AA women was also studied in a cohort of 406 AA OC patients and 632 AA healthy (control) women that were age- and site-matched [14]. Consistent with the beneficial effects of antioxidants, women with high selenium intakes were found to be at ~30% lowered risk of OC. Further, oral contraceptive use, pregnancy and breastfeeding were observed to be inversely correlated with OC in AA women, similar to what was already known in CA women [15]. Also, high intake of dietary sugars and glycemic load were confirmed to be associated with higher OC risk in AA women, similar to other populations [16].

Socioeconomic status (SES) remains a major factor driving OC health disparities. In a study that evaluated 2432 epithelial OC patients (1989 CA vs. 443 AA) in Cook County, Illinois between the years 1998 and 2007, neighborhood SES, particularly the affluence and disadvantage, had profound effect on OC survival [17]. Even after adjusting for factors such as age, tumor characteristics, treatment and year of diagnosis, AAs were much more likely to succumb to OC, compared to CAs. This disparity could be attenuated only if SES factors were factored-in as well. On very similar lines, in a study that evaluated patients from the same locality, but from an early time-period (1994 through 1998), neighborhood disadvantage was still observed as the primary reason for OC-specific survival [18]. In an earlier report [19], AA OC patients were reported to be less likely to undergo surgery and chemotherapy. Unfortunately, the OC health disparity has increased over time and one reason for this increase is the disparate access to treatment, particularly surgery [20]. In a study that evaluated 393 OC patients (68 AA vs. 325 CA), a clear survival disadvantage was observed in AA patients [21]. The AA patients had lower SES, in terms of education, income and property, to start with. However, the survival disadvantage was evident even when controlled for these factors. Progression free survival was only 16 months in AA OC patients, as compared to 27 months in CA OC patients, whereas overall survival was less than half in AA patients (42 months), as compared to CA patients (88 months).

Racial disparity has been noted in the diagnosis of OC [22] with AA women being frequently diagnosed with advanced stage OC. However, interestingly, in one study, no race-based differences in clinical outcome were reported in patients with advanced stage epithelial OC, provided they received similar treatment [23]. While elsewhere, it has also been reported that AA OC patients are much less likely to receive standard guideline-recommended care, as compared to CA patients [24]. Although these studies seem to support the notion that equal access to healthcare is an important factor in determining disparate OC outcomes, it is also possible that this was because of the fact that these studies focused on advanced stage OCs, where the disease was too progressed and aggressive in the patient populations to reveal a meaningful correlation of progression and/or survival with the race. Access to healthcare has often been discussed as an important factor that influences cancer health disparity. In OC, access to health care and its impact on survival of OC patients was studied in a cohort study in Northern California [10]. Since the cohort comprised of members of Kaiser Permanente, it was assumed that the access to health care was uniform for all subjects. Even with equivalent access to health care, AA women had worst survival, which was, in part, due to treatment delay and early discontinuation.

Geographical proximity to a tertiary hospital system and travel distance are also factors that play a role in disparate outcome of OC in AA women [25]. Availability of reliable mode of transportation is part of the SES, and it is believed that AA OC patients are much less likely to travel more than 20 miles to a big hospital for treatment, compared to CA patients. While ~22% CA patients would travel more than 20 miles for treatment, only about 14% AA patients would likely travel this distance. This results in increased risk of non-adherent care and the resulting sub-optimal treatment and follow-up. AA patients are also significantly less likely to undergo initial surgical intervention for OC at big, high-volume hospitals, compared to CA patients [26]. Participation of minorities in clinical trials is yet another challenge. It has been reported that the overall enrollment of AA populations in gynecologic oncologic group (GOG) trials is dismal, and is probably going down further, according to the most recent data [8]. In this data, generated from 445 GOG publications between 1985 and 2013, the overall participation of AA patients was just 8%, compared to 83% CA patients. The more disturbing observation was that the situation is probably getting worse as the enrolment of AA patients has gone down 2.8 folds for the years 2009–2013, compared to the enrolment in years 1994 through 2002. The various non-biological factors that contribute to OC health disparities, as discussed in this section above, are summarized in Fig. 1.

While non-biological factors play an important role in OC health disparity, there is evidence to suggest a critical role of biological factors in OC health disparity as well [6]. In fact, enhanced incidence and mortality in AA women with OC has been observed, as compared to EA counterparts, even when socioeconomic factors have been accounted for. For example, in a study that evaluated 393 OC patients (68 AA vs. 325 CA), a clear survival disadvantage was observed in AA patients [21]. The AA patients had lower SES, in terms of education, income and property, to start with. However, the survival disadvantage was evident even when controlled for these factors. Progression free survival was only 16 months in AA OC patients, as compared to 27 months in CA OC patients, whereas overall survival was less than half in AA patients (42 months), as compared to CA patients (88 months). Possible involvement of biological factors in OC health disparity was also suggested in a study that factored for access to healthcare, and still observed poor survival of AA OC patients, compared to CA patients [10].

Higher adiposity is a known OC risk [27]. Since AA women have disproportionately high rate of obesity [28], a study was carried out to evaluate whether body mass index (BMI) or the weight gain since the age of 18 years can be independent OC risk factors [27]. It was observed that a BMI ≥ 40 posed a significant OC risk, compared to BMI < 25, thus connecting obesity with increased OC risk. Similarly, gain of weight since age 18 also correlated positively with OC risk [27]. In a similar study, where BMI ≥ 30 was considered obese, AA women with higher BMI reported having OC symptoms much before the formal diagnosis [29]. Severe obesity is a critical risk factor for OC [30], underlining the importance of healthy and active lifestyle.

Several molecular alterations such as gene-mutations, genetic polymorphism, epigenetic alterations etc. have been associated with OC racial health disparities (Fig. 2). In this section we discuss the role of these molecular factors in the racially disparate outcome of OC.

The existence of OC health disparities is undeniable, as per the evidence that has accumulated over last decade or so. For records, it needs to be acknowledged that it’s not just the AA women, but OC health disparities are evident in other populations as well, such as, Hispanics and Asian/Pacific Islanders [10, 26, 83]. What is not clearly understood is the basis of such disparities. Non-biological factors, particularly the socioeconomic ones, make sense but emerging literature clearly hints at existence of defining biological factors as well. While non-biological factors might appear to be simple and straight-forward, a closer look reveals that OC health disparities involving SES factors such as neighborhood-related, income, access and affordability of treatment etc. are not easy to be addressed. A number of steps need to be taken to tackle these issues [84]. These will invariably involve some grass-root level efforts with involvement of community outreach programs, counseling initiatives and, possibly, social media. The establishment of partnerships such as Georgia Breast Cancer Genomics Health Consortium to assist young AA women at increased risk of hereditary breast and OC is a step in this direction [85]. In addition to the various non-biological and biological bases of OC health disparities, there is an added burden of factors that are difficult to classify one way or the another. An example is the suggestion that more prevalent use of genital powder in AA women is a risk factor for epithelial OC [86]. Thus, an interplay of biological as well as non-biological factors, with some role of environmental factors, is possibly a more acceptable explanation for observed OC health disparities. This brings in epigenetic factors that provide a logical connection between biological, non-biological as well as environmental factors. When compared to other cancers with documented health disparity, the literature on OC is still in its infancy. However, we need to remind ourselves that OC is the most lethal gynecological malignancy. It’s about time that more sincere efforts are made to better understand and target OC health disparities, so that the dismal screening and outcome can be substantially improved.