Article Text
PDF
Abstract
Objectives Low grade serous ovarian carcinoma is a rare subtype of ovarian cancer with an indolent and chemorefractory course. As such, treatment strategies among practitioners are not uniformly known. The primary objective of this study was to identify differences in practice patterns among physicians who treat low grade serous carcinoma.
Methods Materials A de novo survey was distributed to members of the Society of Gynecologic Oncology. Questions about demographics, management of primary and recurrent disease, and use of consolidation therapy were included. Statistical analyses were performed using χ2 and Fisher’s exact tests.
Results 194 gynecologic oncologists completed the survey. Approximately two-thirds of respondents practiced in a university based setting and treated a high volume of ovarian cancers, including low grade serous carcinoma. 82% recommended somatic testing during treatment and 84% routinely sent patients for genetic counseling. Treatment preferences for primary disease varied by debulking status. 48% of practitioners used hormone antagonism as consolidation after primary treatment. Secondary cytoreduction was preferred for patients with platinum sensitive recurrence and a long disease free interval following primary treatment (P<0.001). Hormone antagonism was the preferred treatment for the first platinum resistant recurrence (54%), while a BRAF inhibitor was the preferred agent in platinum resistant recurrence in the presence of a known BRAF mutation (56%).
Conclusions There was significant variation in the preferred management of low grade serous carcinoma among practitioners. Further efforts to improve knowledge of this disease, identify optimal treatment modalities, and provide guidelines for management should be encouraged.
- low grade
- serous
- ovarian carcinoma
- survey
- practice patterns
Statistics from Altmetric.com
Highlights
Management preferences for low grade serous carcinoma of the ovary vary widely across the USA.
Less than half of practitioners use hormonal consolidation after primary treatment.
The decision to offer fertility sparing surgical staging varies by region of fellowship training.
Introduction
Low grade serous ovarian carcinoma constitutes approximately 3% of all ovarian cancers, and 5–8% of epithelial subtypes.1–3 A number of key characteristics differentiate it from its more common high grade counterpart, including younger age at diagnosis, slow and indolent disease course, and longer progression free and overall survival.1 In contrast with high grade disease, in which responses to platinum approach 80%,4 responses to traditional platinum based chemotherapy in low grade serous carcinoma are poor, with rates of 4%, 4.9%, and 2.1% in the neoadjuvant, platinum sensitive, and platinum resistant recurrent settings, respectively.5 6 Other agents have shown somewhat improved response rates in the recurrent setting, including hormonal therapy (9%) and MEK inhibitors (16%).7 8 Despite these differences, the standard of care with surgical cytoreduction and adjuvant platinum based chemotherapy in the primary setting does not vary between the low grade and high grade histologies.
The known poor response to traditional agents without an obvious alternative may create a lack of consensus about the optimal treatment modality for this disease. The primary objective of this study was to investigate differences in practice patterns among physicians treating low grade serous ovarian carcinoma across the USA and, secondarily, to identify how these differences correlate with demographic factors, which may predict variations in treatment approach.
Methods
After approval by the institutional review board at the University of Miami-Sylvester Comprehensive Cancer Center (protocol ID No 20161110), a de novo, anonymous 18-question survey was developed (see online supplementary digital content 1). The survey included 7 demographic questions to obtain information about fellowship training, current practice setting, years of practice experience, and case volume, and 11 management questions addressing treatment preferences given various clinical scenarios, including primary and recurrent disease settings, use of consolidation, and surveillance practices. Recognizing that significant variability may exist, and that multimodal therapy may be included as part of a treatment approach, respondents were asked to select all items that they may consider using for that clinical scenario. The survey was iteratively reviewed by all study team members and edited until consensus was achieved on clarity of the verbiage and validity of the content.
Supplemental material
Utilizing the Society of Gynecologic Oncology (SGO) listserv, the survey was distributed via email to all full and candidate members across the USA (n=1190). During April 2017, the survey request was sent a total of three times, 1 week apart, and closed 3 weeks after the initial request. The study data were collected using REDCap (Research Electronic Data Capture) tools hosted at the University of Miami.9 Respondents were excluded if they were not gynecologic oncologists. All analyzed respondents had completed the survey in its entirety.
Statistical analyses were completed using STATA C14 (StataCorp, College Station, Texas, USA). Year of fellowship graduation was divided into thirds (prior to 2000, 2000–2009, and 2010 and beyond) and used as a surrogate for years of practice experience. Location of practice and fellowship training were both segregated into regions per the United States Census Bureau tracts. Descriptive statistics were generated. Proportional comparisons between demographic characteristics and preferred treatment approaches were done using the χ2 test (or Fisher’s exact test, when appropriate), with statistical significance set at P=0.05.
Results
Demographics
A total of 199 surveys were completed, yielding a response rate of 16.7% (199/1190). Five respondents were excluded based on their stated specialty (4 medical oncologists and 1 “advanced pelvic surgeon”), leaving 194 gynecologic oncologists included in the analysis. Respondent demographics are shown in Table 1. Of the 194 respondents, over twice as many practiced in a university based setting compared with a community based setting (67% vs 33%). A total of 64% of participants had treated at least 25 patients with all types of ovarian cancer in the past 12 months, while 36% have treated 24 or less; 68% had treated at least 5 patients with low grade serous carcinoma in the past 5 years, while 32% had treated 4 or less.
Respondent demographics divided by region of fellowship training and region of current practice
Primary Disease
In the upfront management of primary disease, there was significant variation among practitioners based on cytoreductive status (Figure 1). When comparing optimally versus suboptimally debulked patients with stage IIIC low grade serous carcinoma, observation was favored significantly more often in patients who were optimally debulked (5.2%) versus suboptimally debulked (0.5%) (P=0.01), although this was still not common practice. Physicians with the most practice experience were significantly more likely to prefer observation, compared with those with the least experience (10.3% versus 3.8% versus 1.4%, P=0.04). Adjuvant platinum based chemotherapy alone, including standard intravenous, intraperitoneal/intravenous, or dose dense regimens, was the most commonly used management regardless of debulking status, selected in 73.7% and 79.9% in optimally versus suboptimally debulked cases, respectively. Rates of chemotherapy for primary treatment varied significantly based on region of fellowship completion, with oncologists trained in the south favoring the greatest use of chemotherapy following both optimal (P=0.02) and suboptimal (P=0.04) debulking.
Management preferences in primary stage IIIC low grade serous ovarian carcinoma based on debulking status. Respondents were able to select more than one option. VEGF, vascular endothelial growth factor.
Hormonal therapy alone was preferred more often after optimal than suboptimal cytoreduction (12.9% versus 6.7%, P=0.04), although it was used relatively infrequently overall. Additionally, 19 respondents (9.8%) stated they would treat with a combination of both platinum based chemotherapy and hormonal therapy, independent of debulking status. Vascular endothelial growth factor inhibitors were not commonly used in primary treatment although, when chosen, they were selected more often in suboptimally debulked disease compared with optimally debulked disease (5.7% versus 1.0%, P=0.01).
Recurrent Disease
In the recurrent setting, differences existed between the first platinum sensitive recurrence, first platinum resistant recurrence, and second platinum resistant recurrence (Figure 2). In platinum sensitive recurrence, secondary cytoreduction was significantly more preferred at 91.8% compared with both first and second platinum resistant recurrence at 27.3% and 17.5%, respectively (P<0.01). The preference for secondary debulking in first platinum resistant recurrence was also significant when compared with second platinum resistant recurrence (27.3% versus 17.5%, P=0.02).
Management preferences in recurrent low grade serous ovarian carcinoma based on type of recurrence. Respondents were able to select more than one option. VEGF, vascular endothelial growth factor.
Chemotherapy was less likely to be used at the time of second platinum resistant recurrence (15.5%) than in either platinum sensitive recurrence (33.5%) or first platinum resistant recurrence (26.8%) (P<0.01). Selection of an AURELIA based chemotherapy10 at the time of a second platinum resistant recurrence showed significant regional variation based on current practice location, with 12.7% of northeastern, 2.2% of midwestern, 21.1% of southern, and 31.8% of western practitioners opting for this management (p<0.01). Additionally, community based physicians were significantly more likely to utilize AURELIA based chemotherapy in these patients compared with those in a university based setting (25.0% versus 10.7%, P=0.01).
Hormonal therapy was the most preferred option for patients with platinum resistant recurrence, and it was significantly more likely to be used in the first platinum resistant recurrence than in a second platinum resistant recurrence or a platinum sensitive recurrence (53.6% versus 32.0% versus 32.5%, respectively, P<0.01).
MEK inhibitor use was more preferred in cases of platinum resistance, at 7.2%, 18.0%, and 15.5% during platinum sensitive, first platinum resistant, and second platinum resistant recurrences, respectively (P<0.01). Physicians who had treated ≥5 cases of low grade serous carcinoma in the past 5 years were more likely to treat with MEK inhibitors in the first platinum sensitive recurrence, compared with those who had treated <5 cases (9.8% versus 1.6%, P=0.04). Additionally, an inverse relationship was seen between the time since fellowship and the use of MEK inhibitors in a second platinum resistant recurrence, with more recent graduates more likely to opt for this management (21.1% versus 19.0% versus 6.2%, respectively, among the three groups, P=0.03). University based oncologists were also more likely to use MEK inhibitors in this setting compared with those in the community (20.0% versus 6.3%, P=0.01).
Vascular endothelial growth factor inhibitor selection showed significant variation, being utilized most often in cases of first platinum resistant recurrence by 17.0% of physicians compared with 6.2% in platinum sensitive, and 4.6% in second platinum resistant recurrence (P<0.01) (Figure 2). There was a trend toward vascular endothelial growth factor inhibitor use in platinum sensitive disease by those who treated more low grade serous carcinomas (8.3% versus 1.6%, P=0.11). In a second platinum resistant recurrence, there was also a trend toward vascular endothelial growth factor inhibitor use by those of intermediate career experience at 10.3%, compared with 1.5% of their senior, and 2.8% of their junior colleagues (P=0.08).
In the case scenario presented for a second platinum resistant recurrence, the patient was noted to have a known BRAF mutation, resulting in 55.7% of respondents selecting a BRAF inhibitor as their preferred management (Figure 2). This followed a significant geographic distribution by the region of current practice, with 48.8% of northeastern, 78.4% of midwestern, 49.2% of southern, and 57.8% of western physicians opting for this management (P=0.02).
Genetics
Somatic tumor testing was recommended by 82.5% of physicians during treatment of a patient with low grade serous carcinoma. Of these, 33.8% recommended testing at initial diagnosis versus 66.2% at the time of a first or second recurrence. No regional or experiential trends were identified for this recommendation.
While treating a patient with low grade serous carcinoma, 83.5% of oncologists 'always' or 'often' recommended a consultation with a genetic counselor to discuss germline testing, while 16.5% 'rarely' or 'never' did. Practitioners who treated a greater number of patients with low grade serous carcinoma were less likely to provide a genetics referral, compared with those who treated fewer of these tumors (79.5% versus 91.9%, P=0.03).
Consolidation and Surveillance
After optimal cytoreduction and 6 cycles of adjuvant platinum based chemotherapy with no evidence of disease, 50.5% of respondents opted for surveillance alone versus 47.9% who recommended hormonal consolidation; 1.6% recommended consolidation with paclitaxel, bevacizumab, or poly (ADP-ribose) polymerase (PARP) inhibitors.
As part of routine practice during treatment and surveillance, 98.5% of gynecologic oncologists stated that they 'always' or 'often' followed CA-125 levels if initially elevated at the time of diagnosis, while 1.5% stated that they 'rarely' or 'never' did.
Fertility Preservation
In the case of a young woman with newly diagnosed low grade serous carcinoma on excision of an intact pelvic mass, without capsular involvement, negative washings, and no lymphovascular space invasion who desires future fertility, 74.2% of practitioners elected to observe, 6.2% would proceed with a full oncologic staging procedure including hysterectomy, 4.1% elected for hormone antagonism, and 4.1% for chemotherapy; 11.3% indicated that they would offer the patient a fertility sparing surgical staging procedure to include unilateral salpingo-oophorectomy, omentectomy, peritoneal biopsies, and pelvic and para-aortic lymph node dissection, with preservation of the uterus and contralateral adnexa.
These management preferences varied based on region of fellowship training (Figure 3). Observation was more preferred by those who trained in the northeast and midwest (78.2% and 87.0%, respectively) compared with the south and west (69.0% and 54.5%, respectively) (P=0.02). Although not reaching statistical significance, there was notable regional variation in the recommendation for full oncologic staging with hysterectomy, with 9.1% and 13.5% of those who trained in the northeast and west opting for this management, respectively, compared with 0.0% and 5.6% of those who trained in the midwest and south, respectively (P=0.07). There was also a trend in preference for fertility sparing surgical staging, which was most commonly recommended by practitioners from the west (27.3%) compared with 7.3% from the northeast, 8.7% from the midwest, and 11.3% from the south (P=0.11). No factors other than region of fellowship training were correlated with the management preferences in this case scenario.
Management preferences in low grade serous ovarian carcinoma for young women desiring future fertility, based on region of fellowship training.
Discussion
Given the unique disease course of low grade serous ovarian carcinoma, known poor response to traditional therapies, and lack of consensus on optimal treatment recommendations for this tumor type, management decisions are left to the preferences of individual practitioners. This inherently creates a large degree of variation in practice patterns between practitioners. As we have demonstrated here, there are significant differences in the management of this disease based on location of training, practice environment, and experience treating women with low grade serous carcinoma.
Hormone antagonistic therapy has demonstrated significant benefit in prolonging time to disease progression in retrospective studies.11 However, respondents in this survey did not commonly elect to use hormone antagonism as consolidation therapy after primary treatment with no evidence of disease. No demographic information was found to correlate with the decision for consolidation therapy or surveillance alone. The large dichotomy here is likely related to a lack of rapid dissemination of new information, as the study identifying the benefits of hormonal consolidation in this setting was published less than 2 months prior to the distribution of this survey. This highlights the need for a method of prompt circulation and adoption of updated practice recommendations in the presence of cutting-edge therapeutic approaches that can positively impact patient outcomes.
The majority of participants recommended somatic tumor testing and genetics referrals for their patients with low grade serous carcinoma. At this time, somatic tumor testing is not part of the routine management recommendations for patients with this disease and is generally utilized as part of an individualized approach in the recurrent setting. Germline BRCA mutations have rarely been identified in cases of low grade serous carcinoma or their serous borderline precursors and, when present, are generally considered to be an incidental finding, rather than the causative aberration.12–16 However, the National Comprehensive Cancer Network does recommend a genetics referral to discuss germline testing in all patients with ovarian cancer, including low grade serous carcinoma.17–19
Research into the vascular endothelial growth factor inhibitor, bevacizumab, has been undertaken in ovarian cancer in general through several phase III clinical trials, and demonstrated small improvements in progression free survival.10 20–24 In low grade serous carcinoma, retrospective analyses identified a partial response rate of 40–55% when bevacizumab was used, and complete response rate of 0–7.5%.25 26 However, these studies suffered from relatively low numbers of low grade serous carcinoma cases, did not perform subset analysis based on debulking status or tumor burden, and in some cases reported relatively discordant outcomes, making the use of bevacizumab in low grade serous carcinoma controversial.
MEK inhibitors, which target the downstream effectors of the RAS GTPases and RAF kinases in the mitogen activated protein kinase pathway, are still early in their investigation for low grade serous carcinoma, and therefore were not commonly selected for management in this study. The prevalence of KRAS, NRAS, and BRAF mutations in low grade serous carcinoma is 19–41%, 15–33%, and 2–33%, respectively.16 Given their mechanism of action, improved tumor response would be expected in cases of RAS/RAF mutation compared with wild-type status. The phase II trial, GOG-239, investigating the use of a MEK inhibitor in recurrent low grade serous carcinoma, revealed an improvement in response rate and stable disease rate over hormonal therapy and cytotoxic chemotherapy (16% versus 9% versus 5%, and 65% versus 62% versus 60%, respectively).6–8 27 However, no correlation between response rate and mutational status was identified in this trial.8 Unfortunately, the subsequent phase III MILO trial closed early, after the hazard ratio for progression free survival dropped below the predetermined futility boundary.28 Further trials into MEK inhibitors in low grade serous carcinoma are currently ongoing, including GOG-281, a phase II/III trial which has recently completed accrual.29
BRAF inhibitors, which target the RAF kinases of the mitogen activated protein kinase pathway, are another group of agents currently being investigated. BRAF inhibitors have demonstrated an initial improvement in outcomes for melanoma and are routinely used in this tumor.30 They have also been investigated in a number of other BRAF mutated solid tumor types, including colorectal cancer, non-small cell lung cancer, and thyroid cancer, with mixed results.30–32 BRAF inhibitors have demonstrated a paradoxical activation of the mitogen activated protein kinase pathway in RAS mutated BRAF intact melanomas, likely due to an intracellular RAF autoregulation system that inhibits RAS.33 As a result, use of BRAF inhibitors should be tailored to the known mutational status of the tumor. Several case reports using this targeted therapy in low grade serous carcinoma have been reported in the literature with promising results, although no formal studies looking specifically at BRAF inhibitors in this disease have been published at present.34
The variations in fertility sparing approaches observed correlated with the region of fellowship training. Although the specific reasons for this variation cannot be definitively known from this study, prior studies have also identified regional variations in practice patterns of fertility sparing treatment. In a review of the SEER dataset, the western region was found to be significantly more likely to perform fertility sparing treatment without hysterectomy in early stage epithelial ovarian cancer than the southern region.35 In another survey based study of SGO members investigating the practice patterns of fertility sparing treatments, the influence of socioeconomic status on the decision for fertility sparing management was found to be significantly higher in the midwest compared with the south. They also found that the majority of providers considered patient age, number of children, history of infertility, cancer stage, and prognosis when deciding to offer fertility sparing treatment, independent of practice region.36
To our knowledge, this is the first study examining the treatment differences between practitioners in low grade serous carcinoma. The major strength of this study is that it provides the opportunity to not only identify the management variations that exist between practitioners, but also to correlate these divergences with categorical factors, thus creating awareness of certain trends and biases and reducing any unintended effect on patient care in the future.
As with any survey based study, this study is subject to certain limitations. One such limitation is the possibility of selection bias in the setting of our survey completion rate of only 16%. However, this is comparable with other recent published survey based studies of SGO member practice patterns, with response rates ranging from 14% to 24%.36 37 Another potential limitation is of reporting bias, as some of the management choices included therapy options that have shown success in the laboratory setting or in case studies, but that have not been adequately studied for routine in vivo use. In some cases, these may have been selected based on a theoretical therapeutic advantage but may not actually be utilized in the respondent’s current practice. Lastly, this survey did not undergo an instrument validation process, which could have potentially resulted in biased responses to some questions.
When treating low grade serous ovarian carcinoma, it is important to consider the unique characteristics that separate it from the other more common subtypes. Without recommendations that address these differences, management often requires a more individualized approach on a case by case basis, compared with other gynecologic cancers. Selecting an appropriate regimen can be difficult given the rare nature of the disease, as any individual’s level of experience will likely be low. One potential solution would be to create centers of excellence that specialize in low grade serous ovarian cancer, thus providing a location with a greater collective experience for patient referrals, as well as facilitating research progress through larger cohorts of patients to identify superior treatment modalities. A similar approach was taken in he UK for the treatment of colorectal cancer to address variations in patient care across the country, resulting in an improvement of a number of metrics for these patients.38
In conclusion, there is a large degree of variation in practice patterns between gynecologic oncologists across the USA in the management of low grade serous ovarian carcinoma. These preferences were found to be associated with a number of factors, including practice volume, region of fellowship training and current practice, practice setting, and years of practice experience. Further research is required to identify the optimal treatment modalities for this disease but a greater emphasis should currently be placed on providing management recommendations that are specific to low grade serous carcinoma and its unique characteristics.
References
Footnotes
Contributors BS; developed concept, reviewed and approved manuscript. DG: developed concept, reviewed and approved manuscript. JS: developed concept, performed data collection and analysis, wrote manuscript text, approved manuscript. MS: developed concept, performed data collection and analysis, wrote manuscript text, approved manuscript.
Funding This study was funded through MS’s faculty research fund, and carried out with the use of software licenses provided through the University of Miami-Sylvester Comprehensive Cancer Center.
Competing interests None declared.
Ethics approval The study was approved by the institutional review board at the University of Miami-Sylvester Comprehensive Cancer Center (protocol ID# 20161110).
Provenance and peer review Not commissioned; externally peer reviewed.