Faculty Bibliography

Objective: To explore the feasibility of molecular tumor characterization in metastatic endometrial cancers (EC) to direct therapy using clinically actionable targets.
Method(s): In December 2017, our program prospectively instituted the Molecular Driven Endometrial Cancer Therapy (MODEL) paradigm to treat metastatic EC patients with selection of therapy based on molecular tumor characterization (in-house immunohistochemistry, IHC, or next-generation sequencing, NGS, FoundationOne) at the discretion of the treating physician. Therapies and reported predictive IHC and genomic alterations (GA) biomarkers included bevacizumab (B-catenin, CTNNB1), trastuzumab (ERBB2), temsirolimus (TSC), immunotherapy (MSI, dMMR, and Tumor Mutation Burden, TMB), and hormone therapy (estrogen, ER) and progesterone receptor (PR) for advanced disease (MODEL1). PI3K/AKT/mTOR pathway and ERBB3GA were also included for recurrent disease (MODEL2). A retrospective review of Foundation Medicine (FMD) and our institutional Endometrial Cancer Databases (IECD) was performed.
Result(s): Of 3,702 advanced EC in FMD (922 endometroid, 826 serous, 156 clear cell, and 40 mixed, 1,758 EC-NOS), 1,497 (40.4%) had at least one MODEL1 qualifying GA: ERBB2 (370, 10.0%); CTNNB1 mutation (660, 17.8%); TSC2 mutation (43, 1.2%); MSI-H (534, 14.4%), or MSS/TMB >20 (93, 2.5%). A total of 2,994 (80.9%) EC had MODEL2 qualifying GA: 2,836 (76.6%) had PI3K/AKT/mTOR and 115 (3.1%) ERBB3 GA. For MODEL1, 247 (6.7%) and for MODEL2 1,340 (36.2%) EC had alterations qualifying for more than 1 therapy arm (Table 1). Institutional results were similar with 33.8% (22/65) and 78.5% (51/65) having least one clinically actionable targets for MODEL1 and MODEL2, respectively. Twenty-two of twenty-three IECD patients were eligible and had tumor testing; 63.6% had at least one clinically actionable target; 40.9% (9/22) had ER/PR+ tumors (median ER 80%, range 20%-100%; median PR 40%, 10%-96%). The most common GA were in the PI3K/AKT/mTOR pathways (6/10, 60%). Seventeen percent have received biomarker-directed therapy. Tumor response and survival outcomes are being evaluated.
Conclusion(s): Treatment with clinically actionable genomically targeted drugs is feasible in metastatic endometrial cancer. Overlapping genomic alterations are common, and therapeutic prioritization is needed. [Figure presented]
Copyright

Objective: Lynch syndrome (LS) accounts for 3%-5% of endometrial cancers. Screening high-risk (HR) patients with endometrial cancer (EC) for LS can result in prevention of other cancers and cascade testing for family members. We implemented universal mismatch repair (MMR) immunohistochemistry (IHC) in patients undergoing hysterectomy for EC in July 2015. In April 2017, we implemented the practice of genetic counselors (GC) accessing IHC data from pathology reports and contacting physicians to approve genetic counseling for patients with a loss of MMR protein expression. By involving genetic counselors, we sought to increase rates of genetic counseling referrals (GCRs) and genetic testing (GT) in EC patients with abnormal MMR IHC.
Method(s): All women diagnosed with EC who underwent hysterectomy between July 2015 and July 2018 at a single institution were retrospectively identified. Demographic data, IHC results, rates of GCR, and GT rates were abstracted before and after implementation of GC involvement in MMR IHC review and GCR.
Result(s): Of 356 patients with EC who underwent hysterectomy, 321 (90%) had MMR IHC testing. Abnormal MMR IHC was found in 86 (27%) patients with the following distribution: MLH1 and PMS2, 68; MLH1, 2; PMS2, 3; MSH6, 8; and MSH2 and MSH6, 5. In 63 (73%) of the 86 patients, MLH1 promoter methylation was identified as the cause of the abnormal MMR IHC. Of the remaining 23 patients with abnormal MMR IHC, 18 (78%) received GCR, and 16 (70%) had GT. Comparing the time frame from July 2015 to April 2017 (prior to GC involvement) to April 2017 to July 2018 (after GC involvement), there was an increase in GCR from 10/15 (67.7%) to 8/8 (100%). GT rates for the MMR abnormal cohort were 8/15 (53%) compared to 8/8 (100%), respectively. Of the 16 patients with an abnormal MMR IHC result who underwent GT for LS, 9 (56%) were identified to have LS, 7 with MSH6 mutations and 2 with MSH2 mutations. See Table 1.
Conclusion(s): GC access to abnormal MMR IHC results in EC patients' improved rates of GCRs and GT to capture all patients with LS. GC involvement in the review of IHC results and GCR is a feasible and effective strategy to ensure both GCR and GT. Proper follow-up and GT of at-risk patients is critical to increase screening and prevention of other LS-related cancers in the proband, as well as for cascade testing of family members. [Figure presented]
Copyright

Objective: Few effective treatment options exist for women with advanced or recurrent endometrial cancer (EC). To explore a modification of the standard systemic treatment for advanced or recurrent EC, we sought to determine the feasibility of completing 6 cycles of nab-paclitaxel (Nab-P) and carboplatin. Unlike paclitaxel, Nab-P does not require any steroid or other premedication, an important consideration for patients with diabetes mellitus and in the investigation of combinations with immunotherapy. We prospectively evaluated safety and efficacy of a day 1, 8-dose schedule of Nab-P in combination with carboplatin day 1 q3weeks in patients with chemotherapy naive EC.
Method(s): Patients with early-stage and high-risk, advanced primary, or recurrent EC with no prior platinum and taxane exposure were enrolled at a single institution. Patients received 6 cycles of day 1 Nab-P 100 mg/m² IV with carboplatin AUC 6 IV and day 8 Nab-P 100 mg/m² IV q21days. We evaluated percentage completion of 6 cycles with standard dose reductions, as well as toxicity per CTCAE v.4. Measurable disease was not required, and efficacy was assessed by PFS rate at 6 months.
Result(s): From 2016 to 2018, 23 subjects were enrolled; median age was 65 (43-73) years. Nineteen (82%) completed 6 cycles of the doublet therapy. Eight subjects (35%) were dose-reduced 1 level, and 5 (22%) were reduced 2 levels; only 1 subject withdrew due to toxicity. Twelve subjects (52%) had at least 1 grade 3/4 treatment-related adverse event, the most common being anemia, 6 (26%); neutropenia, 4 (17%); and diarrhea, 2 (9%). Pre-existing neuropathy was an exclusion criteria, and 13 (57%) reported at least grade 1 neuropathy with treatment. After treatment, 3 (13%) deaths occurred with 2 due to disease progression and 1 to pulmonary embolism. At 6 months after treatment initiation, 19 (83%) had no evidence of disease or its progression; 4 (17%) had progressed. Kaplan-Meier analysis revealed a 6-month PFS rate of 80.5% (95% CI 65.1%-99.7%) (Figure 1).
Conclusion(s): The Nab-P/carboplatin day 1, 8 regimen met the prespecified criteria of feasibility with acceptable toxicity and efficacy. Use of Nab-P obviates steroid premedications, ideal for immune checkpoint inhibitors that target mismatch repair deficient advanced EC. A future phase II feasibility trial combining an anti-PD-1 agent with Nab-P and carboplatin is planned. [Figure presented]
Copyright

Objective: Niraparib (ZEJULA) is a selective inhibitor of PARP1/2 approved for maintenance treatment of recurrent ovarian cancer (OC) patients who are in complete or partial response to platinum therapy regardless of BRCA or homologous recombination deficiency status, based on the pivotal phase III ENGOT-OV16/NOVA trial (N Engl J Med. 2016 375:2154-2164). In NOVA, dose adjustments due to adverse events (AEs) occurred in 69% of patients and tended to occur early, with most patients reaching their individualized dose within 3 months. A retrospective analysis of NOVA showed that patients with body weight <77 kg or platelet count <150 K/muL were more likely to be dose-reduced due to hematologic AEs; importantly, efficacy was not compromised in those patients. The PRIMA study evaluates niraparib versus placebo as maintenance therapy in high-risk stage III/IV OC patients after frontline platinum therapy. Regular and independent safety data reviews have not identified any new safety issues in the trial. The study was prospectively amended to evaluate the safety and efficacy of a new dosing paradigm.
Method(s): Patients were initially randomized 2:1 to start at niraparib 300 mg QD or placebo. The protocol was amended to modify the starting dose to 200 mg QD in patients with baseline weight < 77 kg or platelet count <150 K/muL and 300 mg in all other patients. The trial remains blinded for efficacy and safety. The safety analyses were conducted to compare the AEs in patients who started the study with 300 mg prior to the amendment compared with those who started at 200 or 300 mg after the protocol was amended.
Result(s): Based on the data cutoff on August 15, 2018, 733 patients were randomized, and 727 patients were dosed. Of those dosed, =34% (n = 247 patients) were dosed based on weight and platelet count. Blinded data were pooled from niraparib and placebo. There were no major differences in key patient demographics or disease characteristics. Relevant safety data are presented in Table 1. [Figure presented]
Conclusion(s): These interim safety data prospectively confirm that niraparib tolerability is improved with the starting dose of 200 and 300 mg when dosing is based upon body weight and platelet count.
Copyright

Objective: Obesity is a significant risk factor for perioperative morbidity and mortality. Outcomes can be improved with standardized protocols including preventive measures and specialized surgical equipment and personnel. We sought to evaluate the outcomes of a surgical safety protocol for all patients with a body mass index (BMI) of >=40 undergoing planned gynecologic surgery.
Method(s): The high BMI pathway (HBP) was developed by a multidisciplinary team of gynecologic oncologists (GO), anesthesiologists, and ancillary surgical and nursing staff based on the most current recommendations from the literature and instituted as a quality improvement project. It was implemented for all morbidly obese patients undergoing planned surgery by a GO. Patients who underwent robotic hysterectomies (RH) on the HBP from 2016 to 2018 were compared with consecutive historical controls who had RHs from 2014 to 2015 prior to HBP implementation. Standard two-sided statistical analyses were performed.
Result(s): Of the 80 patients who successfully completed surgery on the HBP, 55 patients (68.8%) underwent RH and were included in this analysis. These patients were compared to 48 historical controls prior to HBP initiation. There were no significant differences in patient factors or perioperative times between pre- and post-HBP groups (Table 1). Since implementing HBP, there were fewer anesthesia-related complications (ARC) in HBP patients after RH compared to pre-HBP patients (0.0% vs 12.5%, P = 0.02). Among the control patients with ARC, two had respiratory distress requiring pharmacologic intervention, two had increased postoperative nausea and vomiting, and two had intractable postoperative pain. There was also an increase in same-day discharges among patients who underwent RH (65.5% vs 41.7%, P = 0.03), but no difference in hospital readmission rates. There were no differences in intraoperative and 30-day postoperative complications.
Conclusion(s): A HBP to improve perioperative safety for morbidly obese patients undergoing RH resulted in fewer ARCs and increased rates of same-day discharge without increasing perioperative times or intraoperative and postoperative complications. [Figure presented]
Copyright

Objective: The Gynecologic Oncology Group (GOG) study 240 demonstrated a 3.5-month improvement in overall survival when bevacizumab (bev) was added to a combination chemotherapy regimen. This study established a bev-containing regimen as standard therapy for women with recurrent, persistent, or metastatic cervical cancer (CC). Gastrointestinal fistula (GIF) formation is a known complication of bev, and the long-term data of GOG 240 reported that a GIF rate of 15% in women who were treated with bev compared to 1% in the control group women. We sought to evaluate our experience with women treated with bev for CC and to identify associated risk factors for GIF formation.
Method(s): All patients who have received bev for CC from 2012 to 2018 at two academic institutions were identified, and their records were reviewed. Standard two-sided statistical analyses were performed.
Result(s): A total of 43 women were treated with a bev-containing chemotherapy regimen; among them, 34 (79.1%) were treated for CC recurrence, and the remaining were treated for metastatic disease at initial presentation or persistent disease following primary treatment. Thirty-three women (76.6%) received prior radiation therapy (RT); of these, 10 (32.3%) received external beam radiation therapy (EBRT), and 21 (67.7%) had prior EBRT and brachytherapy (BT). The median dose of bev was 15 mg/kg for both EBRT only and EBRT and BT groups. Eleven women developed GIF after bev treatment (11/43, 25.6%). All 11 (100%) had been previously treated with RT, and six (54.5%) had received EBRT plus BT. This resulted in rates of 33.3% (11/33) for GIF formation among women who received EBRT, and 28.6% (6/21) for GIF formation among women who received EBRT plus BT. The median number of bev cycles prior to GIF development was 8 (1-29), and 7 (7/11, 63.6%) received the dose of bev (15 mg/kg) as prescribed in GOG 240. See Table 1.
Conclusion(s): In our cohort of women with CC who were treated with bev, over 25% developed GIF. This is more than expected based on the 15% seen in GOG 240. Notably almost all who developed GIF had recurrent disease and were treated with prior RT. A third of women treated with RT followed by bev formed GIF, representing a considerable proportion of the cohort. GIF development and the possibility of requiring a colostomy should be a part of counseling prior to bev initiation especially in those who have had prior RT. [Figure presented]
Copyright

Objective: The use of medical marijuana (MM) in cancer patients was legalized in New York state in 2016. Reported benefits of MM include reduction of cancer-associated pain, nausea, vomiting, fatigue, and improved appetite, as well as mitigating other side effects of cancer treatment. Tetrahydrocannabinol (THC) and cannabidiol (CBD), both components of MM, have been shown to have therapeutic benefit for cancer-related symptoms. Dosing of MM is described in terms of the THC:CBD ratio. While the general therapeutic benefits have been documented, there is a paucity of data on MM use in patients with gynecologic cancers (GC). We sought to evaluate the effect of medical marijuana for symptom management in patients with GC at our institution.
Method(s): We retrospectively identified women with GC using MM between May 2016 and August 2018 from the medical record. Demographic data, cancer diagnosis, dosage form, quantity, frequency, length of treatment, indication, and reported efficacy of MM were collected.
Result(s): We identified 34 patients using MM for a diagnosis of GC. Table 1 lists the patient characteristics and indications for MM and dosing. Median age of patients using MM was 60 years (46-78 years). Median length of treatment with MM was 189 days (11-761 days). Response to MM included 17(50%) improvement in symptoms; 3 (9%) minimal change; 10 (29%) not reported yet; and 4 (12%) unknown. Overall, among patients for whom symptom improvement was available to date, 17/20 (85%) noted improvement. Four (12%) patients reported unwanted side effects, which were euphoria, dizziness, fatigue, and paranoia. Three patients who reported side effects received MM with a 1:1 THC:CBD ratio; the other patient received a low THC:high CBD dosing. Only one (3%) patient ceased medical marijuana use because of untoward adverse effects. All the patients who reported an improvement in symptoms while using MM were concurrently receiving chemotherapy. Improvement in symptoms was not related to formulation, form, or length of treatment of MM.
Conclusion(s): An improvement in symptoms associated with cancer and treatment was noted in over 80% of GC patients using MM, and very few reported adverse side effects. MM is likely to be a promising treatment for pain, nausea, and anorexia in GC patients and should be considered in the armamentarium of palliative cancer treatment options. [Figure presented]
Copyright

OBJECTIVES/OBJECTIVE:Neoadjuvant chemotherapy (NACT) followed by interval debulking surgery (IDS) is a treatment strategy for ovarian cancer patients with unresectable disease or poor performance status (PS). This strategy has been used in the treatment of advanced endometrial cancer and a survival benefit has been shown in patients who are subsequently able to undergo interval cytoreduction. This study sought to review our single institution experience with NACT for advanced endometrial cancer. METHODS:We conducted a retrospective review of all patients who received NACT for advanced endometrial cancer at two institutions in New York City between 2002 and 2016. RESULTS:We identified 39 patients (median age 61, range 35-89). The histologic subtype distribution was: serous (44%), endometrioid (28%), carcinosarcoma (10%), clear cell (8%), mixed (8%), neuroendocrine (3%). Contraindications to primary surgery included: unresectable disease (72%), poor PS (15%), unresectable disease and poor PS (13%). Twenty-three patients (59%) did not undergo IDS due to: progression of disease (70%), medical ineligibility (4%), unresectable disease (17%), lost to follow-up (4%), death (4%). Sixteen patients (41%) underwent IDS, 81% had an optimal cytoreduction. Disease status at NACT completion was: partial response (56%), stable disease (3%) and progression of disease (41%). There were no complete responses. Patients who responded to NACT had a significantly longer overall survival than those who did not (15 vs. 5 months. P = 0.015). IDS was also associated with an improvement in overall survival versus no surgery (16 vs. 6 months, P = 0.04). CONCLUSIONS:Unlike ovarian cancer, less than half of the patients undergoing NACT for endometrial cancer underwent IDS, none had a complete response, and 41% had disease progression during NACT. However, endometrial cancer patients who underwent IDS had a high rate of optimal cytoreduction. Both response to NACT and IDS were associated with improved survival.

Brachytherapy is well-established as an integral component in the standard of care for treatment of patients receiving primary radiotherapy for cervical cancer. A decline in brachytherapy has been associated with negative impacts on survival in the era of modern EBRT techniques. Conformal external beam therapies such intensity modulated radiation therapy (IMRT) or stereotactic body radiation therapy (SBRT) should not be used as alternatives to brachytherapy in patients undergoing primary curative-intent radiation therapy for cervical cancer. Computed tomography or magnetic resonance image-guided adaptive brachytherapy is evolving as the preferred brachytherapy method. With careful care coordination EBRT and brachytherapy can be successfully delivered at different treatment centers without compromising treatment time and outcome in areas where access to brachytherapy maybe limited.