- Email: [email protected]
Use of a novel sentinel lymph node mapping algorithm reduces pelvic lymphadenectomy rates in patients with low-grade endometrial cancer
Abstracts / Gynecologic Oncology 145 (2017) 2–220
135
therefore health care should be focused at optimizing other medical illnesses.
287 - Poster Session Utility of Adult Comorbidity Evaluation 27 (ACE-27) in predicting survival of stage I endometrial cancer patients P.S. Binder, D. Kallogjeri, L.M. Kuroki, L.M. Divine, L.S. Massad, A.R. Hagemann, D.G. Mutch, M.A. Powell, P.H. Thaker, C.K. McCourt. Washington University School of Medicine in St. Louis, St. Louis, MO, USA Objective: Adult Comorbidity Evaluation 27 (ACE-27) is a validated tool that reflects the number and severity of medical comorbidity. ACE-27 is an independent prognostic factor for endometrial cancer (EC) patients after adjusting for stage, grade, histology, and adjuvant therapy. Our objective was to evaluate the impact of ACE-27 score on disease-specific survival (DSS) and overall survival (OS) in stage I EC patients. Method: Surgically treated stage I EC patients from 2000 to 2012 were identified from a prospectively maintained cancer registry. Lymph node dissection was optional. Endometrioid, serous, and clear cell histologies were included. ACE-27 scores range from 0 to 3: 0 (no comorbidity), 1 (mild), 2 (moderate), and 3 (severe). The association between ACE-27 and survival was explored by multivariable Cox regression analysis after controlling for variables found to be significantly associated with DSS and OS in univariate analysis. Results: We identified 1,522 patients with a mean age of 60 years and median follow-up time of 60 months. ACE-27 score was 0, 1, 2, and 3 in 21%, 39%, 28%, and 12% of patients, respectively. Most cancers were stage 1A (81%), grade 1 (61%), and endometrioid histology (93%). Fifty-one (3.4%) patients died from cancer progression, and 196 (13%) patients died of other complications. After adjusting for age, morbid obesity, stage, grade, histology, lymphvascular space invasion, and adjuvant therapy, ACE-27 score was associated with OS (Fig. 1, P b 0.001) but not with DSS (P = 0.094). Compared to women with an ACE-27 score of 0, those with a score of 2 (adjusted hazard ratio [aHR] = 1.72, 95% CI 1.13–2.62) or a score of 3 (aHR = 2.57, 95% CI 1.58–4.12] had a higher risk of death. Death from comorbid conditions was more common than death from cancer in patients with an ACE-27 score of 1 (14% vs 3%), 2 (12% vs 3%), and 3 (22% vs 1%) (P b 0.001). Conclusion: ACE-27 score is a predictor of OS in patients with stage 1 EC after adjusting for age, morbid obesity, and cancer prognostic factors. Patients with an ACE-27 score higher than 0 are more likely to die from medical comorbid conditions than from cancer, and
Fig. 1.
doi:10.1016/j.ygyno.2017.03.315
288 - Poster Session Use of a novel sentinel lymph node mapping algorithm reduces pelvic lymphadenectomy rates in patients with low-grade endometrial cancer E.J. Tanner IIIa, A.M. Puechlb, K. Levinsona, L.J. Havrileskyb, A.K. Sinnoa, A.A. Secordb, A.N. Faderc, P.S. Leeb. aJohns Hopkins Hospital, Baltimore, MD, USA, bDuke University Medical Center, Durham, NC, USA, cJohns Hopkins School of Medicine, Baltimore, MD, USA Objective: To evaluate the capability of a novel sentinel lymph node (SLN) mapping algorithm to reduce pelvic lymphadenectomy (PLND) rates in patients with low-grade endometrial cancer (LGEC). Method: Patients with LGEC underwent prospective evaluation according to a novel lymphatic assessment algorithm at time of hysterectomy and SLN mapping at 2 academic medical centers. Sidespecific PLND was performed only if ipsilateral SLN mapping failed and high-risk uterine features were identified on frozen section (FS). Side-specific and overall PLND rates were compared to theoretical PLND rates assuming the National Comprehensive Cancer Network (NCCN) endometrial cancer SLN mapping algorithm was used. Results: Since November 2015, 97 LGEC patients were managed according to the algorithm (median age and BMI were 61 years and 32, respectively). Four algorithm deviations occurred due to failure to perform FS appropriately when SLN mapping failed. Of the remaining 93 patients, SLN mapping was bilateral (82%), unilateral (11%), or neither (7%). Six patients (6.5%) had LN metastases identified. All metastases were found in SLNs of those who had bilateral SLN mapping. Of the 17 patients requiring intraoperative FS due to failed SLN mapping in a least 1 hemipelvis, high-risk uterine features were identified in 5 (29%). These patients underwent either bilateral (2) or unilateral (3) PLND. Side-specific and overall PLND rates were 3.7% and 5.4%, respectively. If all patients with failed mapping had undergone PLND according to the NCCN algorithm, side-specific and overall PLND rates would have been 12.9% and 18.3%, respectively. Of the 12 patients who failed to map and did not undergo side-specific PLND, all had lowrisk uterine features: low-grade histology (12), absent (5) or b25% myoinvasion (7), and no lymphvascular invasion (12). (See Fig. 1.) Conclusion: Lymphatic assessment using SLN mapping and selective intraoperative FS results in low rates of PLND in patients with LGEC.
Fig. 1.
136
Abstracts / Gynecologic Oncology 145 (2017) 2–220
When compared to the EC SLN NCCN algorithm, this novel strategy appears to eliminate PLND in patients at lowest risk for metastasis. doi:10.1016/j.ygyno.2017.03.316
289 - Poster Session Comparison of outcomes in optimally-cytoreduced ovarian cancer patients receiving postoperative intraperitoneal chemotherapy vs intravenous dose-dense chemotherapy without bevacizumab B. Schlappe, R. O'Cearbhaill, Q. Zhou, A. Iasonos, J.J. Mueller, O. Zivanovic, G.J. Gardner, K. Long Roche, Y. Sonoda, D.S. Chi. Memorial Sloan Kettering Cancer Center, New York, NY, USA Objective: To compare outcomes between optimally cytoreduced ovarian cancer patients receiving postoperative intraperitoneal cisplatin and paclitaxel (IP) or intravenous carboplatin and dosedense paclitaxel (DD) without bevacizumab. Method: Patients who had optimal (residual disease ≤1 cm) primary cytoreductive surgery for FIGO stage II/III ovarian, fallopian tube, or primary peritoneal carcinoma and received postoperative chemotherapy at our institution between January 2008 and September 2015 were identified. Patients who received at least 1 cycle of IP therapy were included in the IP cohort, and patients who received at least 1 cycle of DD therapy were included in the DD cohort. Patients who received a combination of IP and DD therapy were excluded. Clinical characteristics, treatment information, and follow-up data were recorded. Appropriate statistical tests were used. Landmark analysis was used for survival outcomes. Results: A total of 262 patients were included, 52 in the DD cohort and 210 in the IP cohort. Baseline comorbidity was absent in 36.5% of the DD cohort versus 54.8% of the IP cohort (P = 0.04). The median performance score at initial assessment for chemotherapy was 80 (range 40–100) versus 90 (range 60–100) in the DD and IP cohort, respectively (P = 0.001). Other clinical characteristics including age (P = 0.17), histology (P = 0.22), stage (P = 0.46), BRCA status (P = 1.00), and residual disease (none vs ≤1 cm, P = 0.88) were well-balanced. The presence of toxicity requiring a regimen change occurred in 15.4% of the DD cohort and 21.4% of the IP cohort (P = 0.44). Median progression-free survival (PFS) was 24.5 months (95% CI 19.0–33.1) for the DD cohort versus 25.4 months (95% CI 20.9–30.8) for the IP cohort (P = 0.89). 3-year overall survival (OS) was 81.1% (95% CI 61.5%–91.4%) for the DD cohort versus 90.1% (95% CI 84.7%–93.7%) for the IP cohort (P = 0.08). On univariate analysis,
residual disease ≤1 cm versus no residual disease remained associated with PFS (HR = 1.61, 95% CI 1.18–2.18) and OS (HR = 1.96, 95%CI 1.17–3.26). (See Fig. 1.) Conclusion: No statistically significant difference was seen in PFS or OS between patients receiving postoperative DD or IP chemotherapy without bevacizumab, but given sample size limitations, further exploration of the effect on OS is warranted. Specific differences in toxicity may guide regimen selection in specific patients. doi:10.1016/j.ygyno.2017.03.317
290 - Poster Session A single institution pilot study for universal Lynch syndrome screening: A key step towards statewide screening and care C.M. Cosgrovea, F.J. Backesa, H. Hampelb, R. Salania, L.J. Copelanda, D.M. O'Malleyc, J.M. Fowlera, R. Pearlmand, D.E. Cohne, P.J. Goodfellowf. aThe Ohio State University, James Cancer Hospital, Columbus, OH, USA, bThe Ohio State University Comprehensive Cancer Center, Columbus, OH, USA, cThe Ohio State University, Columbus Cancer Council, Hilliard, OH, USA, dJames Cancer Hospital, Columbus, OH, USA, eThe Ohio State University, Columbus, OH, USA, fThe Ohio State University Medical Center, Columbus, OH, USA Objective: Universal Lynch syndrome (LS) screening has been suggested for endometrial cancer (EC) patients. We sought to determine best approaches for screening at a single, high-volume National Comprehensive Cancer Network center as a first step towards population screening. Method: All molecular testing was performed in CLIA-approved laboratories. Tumor microsatellite instability (MSI) and IHC for MLH1, PMS2, MSH6, and MSH2 were performed for subjects. MLH1 promoter methylation was then assessed in cases with MSI and/or IHC abnormalities. Tumors were classified as mismatch repair (MMR) normal, epigenetic MMR defect, or probable MMR mutation. Germline and tumor mutation testing were performed. Results: A total of 295 cases were recruited from January 2013 to May 2016, comprising b40% of total EC cases treated in the time period. LS screening and mutation testing have been completed for 275 cases. A total of 198 patients (72%) were MMR normal; 61 (22.2%) had an epigenetic MMR defect (MLH1 hypermethylation); and 16 (5.8%) were classified as probable MMR mutation. Of the 16 “probable MMR mutation” cases, 9 carried LS mutations, 6 of which were in MSH6. Mutation carriers were younger (49.1 vs 59.5 years, P b 0.01) and had lower BMIs (29.1 vs 34.6, P b 0.05). One MSH6 mutation carrier had a clear cell cancer, and all other LS cases had endometrioid cancers. Of the 7 probable MMR mutation cases that lacked germline LS mutation tested, 5 had somatic mutations in MMR genes explaining the tumor defect. Conclusion: Universal LS screening for EC patients is feasible with ~6% of patients’ tumors classified as probable MMR mutation. This translates to ~1 in 17 cases being referred for clinical LS gene testing. The majority of women will have LS, and the others have tumor-specific somatic mutations in MMR genes. Our study further underscores the importance of the MSH6 mutation in LS in EC patients, with two-thirds of LS cases having MSH6 mutations. In light of the fact that tumor sequencing effectively identifies both inherited mutations and tumor-specific mutations, we propose universal tumor sequencing as an alternative to current IHC and MSI screening that is followed by genetic testing. No additional consent or specimens are required for tumor-based LS mutation testing. Tumor sequencing diagnostics and strategies for follow-up care of patients and their family members are currently being developed. doi:10.1016/j.ygyno.2017.03.318
Fig. 1. Progression-free Survival by Chemotherapy Regimen.
135
therefore health care should be focused at optimizing other medical illnesses.
287 - Poster Session Utility of Adult Comorbidity Evaluation 27 (ACE-27) in predicting survival of stage I endometrial cancer patients P.S. Binder, D. Kallogjeri, L.M. Kuroki, L.M. Divine, L.S. Massad, A.R. Hagemann, D.G. Mutch, M.A. Powell, P.H. Thaker, C.K. McCourt. Washington University School of Medicine in St. Louis, St. Louis, MO, USA Objective: Adult Comorbidity Evaluation 27 (ACE-27) is a validated tool that reflects the number and severity of medical comorbidity. ACE-27 is an independent prognostic factor for endometrial cancer (EC) patients after adjusting for stage, grade, histology, and adjuvant therapy. Our objective was to evaluate the impact of ACE-27 score on disease-specific survival (DSS) and overall survival (OS) in stage I EC patients. Method: Surgically treated stage I EC patients from 2000 to 2012 were identified from a prospectively maintained cancer registry. Lymph node dissection was optional. Endometrioid, serous, and clear cell histologies were included. ACE-27 scores range from 0 to 3: 0 (no comorbidity), 1 (mild), 2 (moderate), and 3 (severe). The association between ACE-27 and survival was explored by multivariable Cox regression analysis after controlling for variables found to be significantly associated with DSS and OS in univariate analysis. Results: We identified 1,522 patients with a mean age of 60 years and median follow-up time of 60 months. ACE-27 score was 0, 1, 2, and 3 in 21%, 39%, 28%, and 12% of patients, respectively. Most cancers were stage 1A (81%), grade 1 (61%), and endometrioid histology (93%). Fifty-one (3.4%) patients died from cancer progression, and 196 (13%) patients died of other complications. After adjusting for age, morbid obesity, stage, grade, histology, lymphvascular space invasion, and adjuvant therapy, ACE-27 score was associated with OS (Fig. 1, P b 0.001) but not with DSS (P = 0.094). Compared to women with an ACE-27 score of 0, those with a score of 2 (adjusted hazard ratio [aHR] = 1.72, 95% CI 1.13–2.62) or a score of 3 (aHR = 2.57, 95% CI 1.58–4.12] had a higher risk of death. Death from comorbid conditions was more common than death from cancer in patients with an ACE-27 score of 1 (14% vs 3%), 2 (12% vs 3%), and 3 (22% vs 1%) (P b 0.001). Conclusion: ACE-27 score is a predictor of OS in patients with stage 1 EC after adjusting for age, morbid obesity, and cancer prognostic factors. Patients with an ACE-27 score higher than 0 are more likely to die from medical comorbid conditions than from cancer, and
Fig. 1.
doi:10.1016/j.ygyno.2017.03.315
288 - Poster Session Use of a novel sentinel lymph node mapping algorithm reduces pelvic lymphadenectomy rates in patients with low-grade endometrial cancer E.J. Tanner IIIa, A.M. Puechlb, K. Levinsona, L.J. Havrileskyb, A.K. Sinnoa, A.A. Secordb, A.N. Faderc, P.S. Leeb. aJohns Hopkins Hospital, Baltimore, MD, USA, bDuke University Medical Center, Durham, NC, USA, cJohns Hopkins School of Medicine, Baltimore, MD, USA Objective: To evaluate the capability of a novel sentinel lymph node (SLN) mapping algorithm to reduce pelvic lymphadenectomy (PLND) rates in patients with low-grade endometrial cancer (LGEC). Method: Patients with LGEC underwent prospective evaluation according to a novel lymphatic assessment algorithm at time of hysterectomy and SLN mapping at 2 academic medical centers. Sidespecific PLND was performed only if ipsilateral SLN mapping failed and high-risk uterine features were identified on frozen section (FS). Side-specific and overall PLND rates were compared to theoretical PLND rates assuming the National Comprehensive Cancer Network (NCCN) endometrial cancer SLN mapping algorithm was used. Results: Since November 2015, 97 LGEC patients were managed according to the algorithm (median age and BMI were 61 years and 32, respectively). Four algorithm deviations occurred due to failure to perform FS appropriately when SLN mapping failed. Of the remaining 93 patients, SLN mapping was bilateral (82%), unilateral (11%), or neither (7%). Six patients (6.5%) had LN metastases identified. All metastases were found in SLNs of those who had bilateral SLN mapping. Of the 17 patients requiring intraoperative FS due to failed SLN mapping in a least 1 hemipelvis, high-risk uterine features were identified in 5 (29%). These patients underwent either bilateral (2) or unilateral (3) PLND. Side-specific and overall PLND rates were 3.7% and 5.4%, respectively. If all patients with failed mapping had undergone PLND according to the NCCN algorithm, side-specific and overall PLND rates would have been 12.9% and 18.3%, respectively. Of the 12 patients who failed to map and did not undergo side-specific PLND, all had lowrisk uterine features: low-grade histology (12), absent (5) or b25% myoinvasion (7), and no lymphvascular invasion (12). (See Fig. 1.) Conclusion: Lymphatic assessment using SLN mapping and selective intraoperative FS results in low rates of PLND in patients with LGEC.
Fig. 1.
136
Abstracts / Gynecologic Oncology 145 (2017) 2–220
When compared to the EC SLN NCCN algorithm, this novel strategy appears to eliminate PLND in patients at lowest risk for metastasis. doi:10.1016/j.ygyno.2017.03.316
289 - Poster Session Comparison of outcomes in optimally-cytoreduced ovarian cancer patients receiving postoperative intraperitoneal chemotherapy vs intravenous dose-dense chemotherapy without bevacizumab B. Schlappe, R. O'Cearbhaill, Q. Zhou, A. Iasonos, J.J. Mueller, O. Zivanovic, G.J. Gardner, K. Long Roche, Y. Sonoda, D.S. Chi. Memorial Sloan Kettering Cancer Center, New York, NY, USA Objective: To compare outcomes between optimally cytoreduced ovarian cancer patients receiving postoperative intraperitoneal cisplatin and paclitaxel (IP) or intravenous carboplatin and dosedense paclitaxel (DD) without bevacizumab. Method: Patients who had optimal (residual disease ≤1 cm) primary cytoreductive surgery for FIGO stage II/III ovarian, fallopian tube, or primary peritoneal carcinoma and received postoperative chemotherapy at our institution between January 2008 and September 2015 were identified. Patients who received at least 1 cycle of IP therapy were included in the IP cohort, and patients who received at least 1 cycle of DD therapy were included in the DD cohort. Patients who received a combination of IP and DD therapy were excluded. Clinical characteristics, treatment information, and follow-up data were recorded. Appropriate statistical tests were used. Landmark analysis was used for survival outcomes. Results: A total of 262 patients were included, 52 in the DD cohort and 210 in the IP cohort. Baseline comorbidity was absent in 36.5% of the DD cohort versus 54.8% of the IP cohort (P = 0.04). The median performance score at initial assessment for chemotherapy was 80 (range 40–100) versus 90 (range 60–100) in the DD and IP cohort, respectively (P = 0.001). Other clinical characteristics including age (P = 0.17), histology (P = 0.22), stage (P = 0.46), BRCA status (P = 1.00), and residual disease (none vs ≤1 cm, P = 0.88) were well-balanced. The presence of toxicity requiring a regimen change occurred in 15.4% of the DD cohort and 21.4% of the IP cohort (P = 0.44). Median progression-free survival (PFS) was 24.5 months (95% CI 19.0–33.1) for the DD cohort versus 25.4 months (95% CI 20.9–30.8) for the IP cohort (P = 0.89). 3-year overall survival (OS) was 81.1% (95% CI 61.5%–91.4%) for the DD cohort versus 90.1% (95% CI 84.7%–93.7%) for the IP cohort (P = 0.08). On univariate analysis,
residual disease ≤1 cm versus no residual disease remained associated with PFS (HR = 1.61, 95% CI 1.18–2.18) and OS (HR = 1.96, 95%CI 1.17–3.26). (See Fig. 1.) Conclusion: No statistically significant difference was seen in PFS or OS between patients receiving postoperative DD or IP chemotherapy without bevacizumab, but given sample size limitations, further exploration of the effect on OS is warranted. Specific differences in toxicity may guide regimen selection in specific patients. doi:10.1016/j.ygyno.2017.03.317
290 - Poster Session A single institution pilot study for universal Lynch syndrome screening: A key step towards statewide screening and care C.M. Cosgrovea, F.J. Backesa, H. Hampelb, R. Salania, L.J. Copelanda, D.M. O'Malleyc, J.M. Fowlera, R. Pearlmand, D.E. Cohne, P.J. Goodfellowf. aThe Ohio State University, James Cancer Hospital, Columbus, OH, USA, bThe Ohio State University Comprehensive Cancer Center, Columbus, OH, USA, cThe Ohio State University, Columbus Cancer Council, Hilliard, OH, USA, dJames Cancer Hospital, Columbus, OH, USA, eThe Ohio State University, Columbus, OH, USA, fThe Ohio State University Medical Center, Columbus, OH, USA Objective: Universal Lynch syndrome (LS) screening has been suggested for endometrial cancer (EC) patients. We sought to determine best approaches for screening at a single, high-volume National Comprehensive Cancer Network center as a first step towards population screening. Method: All molecular testing was performed in CLIA-approved laboratories. Tumor microsatellite instability (MSI) and IHC for MLH1, PMS2, MSH6, and MSH2 were performed for subjects. MLH1 promoter methylation was then assessed in cases with MSI and/or IHC abnormalities. Tumors were classified as mismatch repair (MMR) normal, epigenetic MMR defect, or probable MMR mutation. Germline and tumor mutation testing were performed. Results: A total of 295 cases were recruited from January 2013 to May 2016, comprising b40% of total EC cases treated in the time period. LS screening and mutation testing have been completed for 275 cases. A total of 198 patients (72%) were MMR normal; 61 (22.2%) had an epigenetic MMR defect (MLH1 hypermethylation); and 16 (5.8%) were classified as probable MMR mutation. Of the 16 “probable MMR mutation” cases, 9 carried LS mutations, 6 of which were in MSH6. Mutation carriers were younger (49.1 vs 59.5 years, P b 0.01) and had lower BMIs (29.1 vs 34.6, P b 0.05). One MSH6 mutation carrier had a clear cell cancer, and all other LS cases had endometrioid cancers. Of the 7 probable MMR mutation cases that lacked germline LS mutation tested, 5 had somatic mutations in MMR genes explaining the tumor defect. Conclusion: Universal LS screening for EC patients is feasible with ~6% of patients’ tumors classified as probable MMR mutation. This translates to ~1 in 17 cases being referred for clinical LS gene testing. The majority of women will have LS, and the others have tumor-specific somatic mutations in MMR genes. Our study further underscores the importance of the MSH6 mutation in LS in EC patients, with two-thirds of LS cases having MSH6 mutations. In light of the fact that tumor sequencing effectively identifies both inherited mutations and tumor-specific mutations, we propose universal tumor sequencing as an alternative to current IHC and MSI screening that is followed by genetic testing. No additional consent or specimens are required for tumor-based LS mutation testing. Tumor sequencing diagnostics and strategies for follow-up care of patients and their family members are currently being developed. doi:10.1016/j.ygyno.2017.03.318
Fig. 1. Progression-free Survival by Chemotherapy Regimen.