Virtual Library
Start Your Search
M.A. Ray
Author of
-
+
P1.01 - Poster Session with Presenters Present (ID 453)
- Event: WCLC 2016
- Type: Poster Presenters Present
- Track: Epidemiology/Tobacco Control and Cessation/Prevention
- Presentations: 2
- Moderators:
- Coordinates: 12/05/2016, 14:30 - 15:45, Hall B (Poster Area)
-
+
P1.01-021 - The Impact of Smoking Status on Overall Survival in a Population-Based Non-Small Cell Lung Cancer (NSCLC) Surgical Resection Cohort (ID 5732)
14:30 - 14:30 | Author(s): M.A. Ray
- Abstract
Background:
Surgical resection is the optimal treatment modality for NSCLC, while smoking has been shown to have a negative survival impact. We evaluated smoking’s impact on overall survival within a population-based cohort of patients with surgically-resected NSCLC.
Methods:
We examined all patients who had a curative-intent NSCLC resection from 2009-2016 in 4 contiguous Dartmouth Hospital Referral Regions of the US. We compared patient and clinical characteristics among never, former (stopped >1 year prior), and active smokers using the Chi-square and ANOVA tests. Survival analyses were conducted with the Kaplan-Meier method and Cox Proportional Hazards models.
Results:
Of 2,202 patients, 206 (9%) were never, 846 (38%) were former, and 1,150 (52%) were active smokers. Significant demographic and clinical differences between cohorts included age, sex, race, insurance, comorbidities, pulmonary function, method of detection, ASA status, extent, primary site and length of resection, histology, and histologic grade (all p<0.05). Short-term post-operative mortality (at 30-, 60-, 90-, 120-days) rates for never smokers were 1%, 2%, 4%, 4%; for active smokers, 4%, 6%, 7% and 8%; and for former smokers, 5%, 7%, 9%, and 11%; and differed significantly by smoking status (p=0.0539, p=0.0316, p=0.0187, p=0.0017). At 5 years, overall survival was 69% for never smokers, 55% for active, and 49% for former smokers (p=0.0002) (Figure 1). Controlling for age, sex, race, insurance, histologic grade, extent of resection, and length of surgery, and compared with never smokers, active smokers had 1.3 times (p=0.05) the hazard of death and former smokers had 1.4 times the hazard of death (p=0.04). Figure 1
Conclusion:
In this population-based cohort, smoking is negatively associated with post-operative mortality and long-term overall patient survival; although active smokers had better survival outcomes than former smokers.
-
+
P1.01-030 - Factors Associated with Margin Positive Resections for Non-Small Cell Lung Cancer (NSCLC) in the Mid-South Region of the US (ID 5076)
14:30 - 14:30 | Author(s): M.A. Ray
- Abstract
Background:
Incomplete resection of NSCLC has a negative impact on survival. We evaluated risk factors associated with positive margins within a comparative observational population-based cohort study.
Methods:
We analyzed curative-intent resections from 2009-2016 from 4 contiguous Dartmouth Hospital Referral Regions in 3 US states. Statistical analyses were preformed using univariate and multiple logistic regression models.
Results:
Among the 2,275 NSCLC-resected patients, 52% were male, 78% white, 45% Medicare insured, and 36% privately insured, with a median age of 67 years. Factors associated with a higher margin positivity rate included male sex, large cell histology, undifferentiated tumor grade, neo-adjuvant therapy, clinical stage IIIA and IIIB, bilobectomy extent of resection, patients with abnormal diffusing capacity of the lungs for carbon monoxide (DLCO), use of bronchoscopic biopsy for diagnosis greater than 1 day before surgery, left lung resection, and tumor size >7cm (all p<0.15, Table 1). American Society of Anesthesiologists (ASA) score, prior lung cancer, smoking status, Charlson score, FEV1, PET/CT, brain scan, bone scan, mediastinoscopy, blood transfusion, and hospital were not associated with positive margins in univariate analyses (all p>0.15). Controlling for sex, histology, tumor grade, tumor size, neo-adjuvant therapy, clinical stage, extent of resection, DLCO, pre-operative bronchoscopic biopsy, and primary resection site in the multiple variable analysis, sex (p=0.0134), clinical stage (p<0.001), extent of resection(p=0.0461), DLCO (p=0.0431), and bronchoscopic biopsy (p=0.0029) were independently associated with risk for positive margins (Table 1).
Conclusion:
This detailed evaluation in a large regional cohort indicates patient-level characteristics are associated with positive surgical resection margins. Our recently published evaluation of the National Cancer Database also identified institutional factors that impact the rates of positive margins. Patient-level, surgeon-level, and institutional-level factors should be considered jointly to fully understand factors impacting margin positivity rates. Figure 1
-
+
P1.03 - Poster Session with Presenters Present (ID 455)
- Event: WCLC 2016
- Type: Poster Presenters Present
- Track: Radiology/Staging/Screening
- Presentations: 1
- Moderators:
- Coordinates: 12/05/2016, 14:30 - 15:45, Hall B (Poster Area)
-
+
P1.03-067 - Validation of the IASLC 8th Edition (8E) TNM Classification for Non-Small Cell Lung Cancer by the Quality of Surgical Resection in a US Cohort (ID 6237)
14:30 - 14:30 | Author(s): M.A. Ray
- Abstract
Background:
We compared the prognostic impact of 8E to 7[th] Edition(7E), using sequentially-defined surgical quality cohorts.
Methods:
We analyzed curative-intent resections for non-small cell lung cancer from 2009-2016 in a population based cohort from 4 Dartmouth Referral Regions in 3 US states. Patients were re-staged by 8E criteria. Survival analyses used Kaplan-Meier estimates and Proportional Hazards models with adjusted hazard ratios(aHR) controlling for age, histology, grade, and comorbidities.
Results:
548 of 2226 patients were stage-redistributed: 525(24%) up, 23(1%) down-staged. The largest shifts were from IB to IIA (76/522 [15%]);IIA to IIB (238/251[95%]); IIB to IIIA (88/217 [41%]); IIIA to IIIB (59/277[21%]). We found no difference in unadjusted survival in patients upstaged to IIA compared with those remaining in IB (p=0.55). Patients upstaged from IIB to IIIA had similar survival to those remaining IIB (p=0.4884), but were similar to patients already IIIA by 7E (p=0.8152). However, patients upstaged from IIIA to IIIB had worse survival than those remaining in IIIA (p=0.0360). Sub-classification of IA had no prognostic value when comparing IA1 vs. IA2 (p=0.74), but patients in IA3 had significantly worse survival than those in IA2 (p=0.0177). 5-year survival estimates for IA1/IA2/IA3 were 65%, 68%, and 61% in our cohort, compared to 92%, 83%, and 77% in the IASLC database. Adjusted models indicate 8E stage as a significant prognostic factor (p<0.0001), with increasing hazards of death with each progressive stage beyond IA2 (Table 1). This result was reasonably consistent as the quality of resection increased incrementally from: All Patients, excluding margin-positives, excluding margin-positives and pNX resections, excluding margin-positives and resections without mediastinal nodes(MedNX).IASLC 8th-EditionStage 3-Year SurvivalEstimate (95% CI) 5-Year SurvivalEstimate (95% CI) IA1(N=91) 0.80(0.69-0.88) 0.65(0.48-0.77) IA2(N=454) 0.80(0.75-0.84) 0.68(0.62-0.73) IA3(N=312) 0.71(0.65-0.76) 0.61(0.54-0.68) IB(N=509) 0.67(0.63-0.72) 0.55(0.49-0.60) IIA(N=81) 0.66(0.53-0.76) 0.61(0.48-0.72) IIB(N=375) 0.59(0.53-0.64) 0.45(0.39-0.52) IIIA(N=302) 0.50(0.43-0.56) 0.41(0.34-0.48) IIIB(N=62) 0.39(0.26-0.52) 0.29(0.18-0.42) IV(N=40) 0.44(0.26-0.61) 0.44(0.26-0.61) Adjusted Hazard Ratios by Quality Parameters All Patients (N=2195) Exclude Margin+ (N=2090) Exclude Margin+/pNX (N=1939) Exclude Margin+/MedNX (N=1656) IA1 1.00 1.00 1.00 1.00 IA2 0.83 0.83 0.70 0.75 IA3 1.12 1.11 1.00 1.13 IB 1.30 1.26 1.11 1.23 IIA 1.34 1.27 1.11 1.25 IIB 1.72 1.69 1.56 1.69 IIIA 2.40 2.31 2.11 2.45 IIIB 3.73 3.21 2.95 3.41 IV 3.76 3.43 2.62 3.11
Conclusion:
8E was generally supported by our data, although modifications for Stage IA1-IIB patients were not fully evident, even in high-quality resections. The survival disparity with IASLC data suggests that unidentified confounding factors are impairing survival in this early-stage US NSCLC cohort.
-
+
P1.05 - Poster Session with Presenters Present (ID 457)
- Event: WCLC 2016
- Type: Poster Presenters Present
- Track: Early Stage NSCLC
- Presentations: 1
- Moderators:
- Coordinates: 12/05/2016, 14:30 - 15:45, Hall B (Poster Area)
-
+
P1.05-044 - The Impact of IASLC 8th Edition Updates for T-Classification for Lung Cancer in a US Population-Based Surgical Resection Cohort (ID 6241)
14:30 - 14:30 | Author(s): M.A. Ray
- Abstract
Background:
Accurate staging of non-small cell lung cancer (NSCLC) is vital for prognostication and treatment selection. We evaluated the impact of the 8[th] Edition TNM (8E) T-classification in a US regional NSCLC resection database.
Methods:
Curative-intent NSCLC resections from 11 hospitals in 4 contiguous Dartmouth Hospital Referral Regions within 3 US states from 2009-2016 were re-staged based on 8E T-categorization. Survival analyses were conducted using the Kaplan-Meier method and proportional hazards models with adjusted hazard ratios (aHR) controlling for age, histology, grade, pN-category, and comorbidities. M1 patients and those who received neoadjuvant therapy were excluded.
Results:
The 2245 patients had a median age of 65, were 48% female, 78% white, 21% black. The 961 pT1 (8E) distribution was 10% pT1a, 52% pT1b, and 39% pT1c. The 793 pT2 (8E) patients were 82% pT2a and 18% pT2b. Of the 318 patients with pT3 (8E), 134 (42%) were pT2b based on the 7[th] Edition TNM (7E); of the 152 with pT4 (8E), 107 (70%) were pT3 based on 7E. There was no survival difference between pT1a and pT1b (p=0.83); pT1c had worse survival than pT1b (p<0.01; Figure 1a). Of the 145 patients previously classified as pT2b by 7E, 134 (92%) were upstaged to pT3. They had similar survival to those classified as pT3 in 7E (p=0.75). Of the 296 patients previously classified as pT3, 107 (36%) were upstaged to pT4. The upstaged patients had worse survival than 7E pT3 patients who were not upstaged, although not statistically significant (aHR:1.32, Figure 1b). Adjusted models confirm an increasing trend in the hazard of death with increasing stage, with the exception of pT1b. (aHR: pT1a=1.00, pT1b=0.89, pT1c=1.15, pT2a=1.38, pT2b=1.54, pT3=1.86, pT4=2.44). Figure 1
Conclusion:
This analysis independently corroborates the 8E re-classification for late stage patients in the US. However, we found no survival differentiation between tumors less than 2cm.
-
+
P1.08 - Poster Session with Presenters Present (ID 460)
- Event: WCLC 2016
- Type: Poster Presenters Present
- Track: Surgery
- Presentations: 3
- Moderators:
- Coordinates: 12/05/2016, 14:30 - 15:45, Hall B (Poster Area)
-
+
P1.08-020 - The Effect of Two Interventions on Attainment of Surgical Quality Measures in Resected Non-Small Cell Lung Cancer (NSCLC) (ID 5694)
14:30 - 14:30 | Author(s): M.A. Ray
- Abstract
Background:
Better pathologic staging improves early-stage NSCLC survival. We sought to measure the impact of complementary surgery (lymph node specimen collection kit) and pathology (a novel gross dissection method) interventions on attainment of guideline-recommended surgical staging quality.
Methods:
We analyzed curative-intent resections from 2004-2016 from 4 contiguous Dartmouth Hospital Referral Regions in 3 US states. Preoperatively-treated patients were excluded. Patients were categorized into groups based on whether a lymph node specimen collection kit was used during surgical resection, and whether a novel, anatomically-sound gross dissection method was used to retrieve intrapulmonary lymph nodes. Chi-squared tests were used to examine differences in demographic and disease characteristics and surgical quality parameters across implementation groups.
Results:
Of 2,094 patients, 1,492 received neither intervention; 152 received only the pathology intervention; 161 received only the surgery intervention; 289 had both (Table 1). Attainment of surgical quality guidelines significantly increased in ascending order of the pathology, kit, and combined interventions (Table 2). Figure 1 Figure 2
Conclusion:
The combined effect of two interventions to improve pathologic lymph node examination has a greater effect on attainment of a range of surgical quality parameters than either intervention alone.
-
+
P1.08-021 - Predictors of Post-Operative Mortality in Non-Small Cell Lung Cancer (NSCLC) in a High Mortality Region of the US (ID 4447)
14:30 - 14:30 | Author(s): M.A. Ray
- Abstract
Background:
Surgical resection is recommended for most patients with early-stage NSCLC. High postoperative mortality risk diminishes the benefit of curative-intent surgery. We examined factors associated with mortality within 120 days of curative-intent resection in a population-based cohort.
Methods:
We examined all NSCLC patients with curative-intent resections from 2009-2016 in all 11 hospitals in 4 US Dartmouth Referral Regions. We evaluated patient demographics, disease characteristics, pre-operative evaluation, treatment, and perioperative complications to identify risk factors for 30-, 60-, 90-, and 120-day mortality using logistic regression models.
Results:
The 2,258 patients’ median age was 67, 48% were female; 78% were White, 21% Black. The 30-, 60-, 90-, and 120-day post-operative mortality rates were 4%, 6%, 8%, and 9%. After adjusting for all other factors, American Society of Anesthesiologists score (ASA) (p=0.0405), prior lung cancer (p=0.0406), and Charlson comorbidity score (p=0.0163) were associated with 30-day mortality. Adjusted models for 120-day mortality indicate associations with age (p=0.0001), tumor size (p=0.0012), intra-operative blood loss (p=0.0150), hospital (p=0.0065), ASA (p=0.0035), prior lung cancer (p=0.0466), and Charlson score (p=0.0064) (Table 1). Patients >75 years old had 1.5 times the odds of 120-day mortality compared with those <49. A Charlson score >=3 (vs. 0) conferred 2.7 times the odds of 120-day mortality. On average, each 1 cm increase in tumor size increased the odds of 120-day mortality by 12%. Patients with all three risk factors (age >75, Charlson score >=3, tumor >4cm) had 26.5% 120-day mortality. Although 17.5% of pneumonectomy patients died within 120 days, extent or duration of surgery were not significant after adjusting for other factors.N (total=2258) 30-Day Mortality 120-Day Mortality % % Age < 49 101 3 7.9 50-64 730 2.6 4.3 65-74 937 4.7 9.9 75+ 490 6.1 13.1 p=0.1954 p=0.0001 Tumor Size(mean) 2258 3.6 3.9 p=0.1834 p=0.0012 Surgery Type Lobectomy/Wedge 1696 3.5 7.8 Pneumonectomy 143 9.1 17.5 Bilobectomy 126 6.4 11.9 Segmentectomy/Wedge 293 5.5 7.9 p=0.4359 p=0.6029 Previous Lung Cancer No 2166 4 8.3 Yes 92 10.9 17.4 p=0.0406 p=0.0466 Charlson Comorbidity 0 455 1.8 4.2 1-2 1132 3.8 8.2 ≥3 671 6.7 12.5 p=0.0163 p=0.0064 Blood loss(surgical) 0-500cc 2048 4 7.8 501-1000cc 136 6.6 16.9 >1000cc 74 8.1 18.9 p=0.4842 p=0.015
Conclusion:
Age, ASA, Charlson score, and tumor size are important risk factors for post-operative mortality. Inter-hospital disparity suggests an opportunity for institution-level corrective interventions. Patients with the combination of age >75, Charlson score >=3, and advanced T-category had a high rate of post-operative mortality.
-
+
P1.08-027 - Evolution of Survival in a Regional Population-Based US Lung Cancer Resection Cohort (ID 6122)
14:30 - 14:30 | Author(s): M.A. Ray
- Abstract
Background:
Quality variances in surgical resection and pathology examination practice translate into survival disparity in patients with early stage lung cancer after curative-intent resection. We evaluated the survival patients from two eras in a US regional cohort.
Methods:
All curative-intent lung cancer resections in 11 US hospitals in 4 contiguous Dartmouth Hospital Referral Regions were analyzed for stage-stratified survival before and after an ongoing regional quality improvement campaign started in 2009. Overall and stage-stratified survival of patients with surgery in the 2004-2009 (pre-era) v 2010-2015 (post-era) were compared using the log-rank test and Cox proportional hazards models.
Results:
Of the total cohort of 3246 patients, 40.6% were in the earlier era, 59.4% in the later era. Demographic characteristics were similar between cohorts (Table 1). Preoperative PET/CT, brain MRI scans, bronchoscopy, and adjuvant therapy were more frequently used in the later era. Patients in the early era had an unadjusted hazard ratio (HR) of 1.22 (p=0.0006). After controlling for stage, tumor size, neoadjuvant therapy, comorbidity score, grade, extent of surgery, patients in the pre-era had a HR of 1.49 (p<0.0001). Figure 1Figure 2
Conclusion:
Survival has improved since introduction of a regional quality improvement campaign in a high lung cancer mortality region of the US.
-
+
P2.08 - Poster Session with Presenters Present (ID 491)
- Event: WCLC 2016
- Type: Poster Presenters Present
- Track: Patient Support and Advocacy Groups
- Presentations: 1
- Moderators:
- Coordinates: 12/06/2016, 14:30 - 15:45, Hall B (Poster Area)
-
+
P2.08-010 - The Reach and Adoption of a Multidisciplinary Thoracic Oncology Program within a U.S. Community Healthcare System (ID 6021)
14:30 - 14:30 | Author(s): M.A. Ray
- Abstract
Background:
The Mid-South region of the US is the center of lung cancer incidence, and has a high proportion of underserved persons. We developed a multidisciplinary (MD) program for lung cancer care, involving weekly physician conferences, and a clinic in which conference-recommended treatment plans are discussed and implemented with patients. This study evaluates the reach and adoption of this program within a community healthcare system.
Methods:
We evaluated the reach of MD care by comparing demographic characteristics of participating patients, to the larger metropolitan and regional population of patients in a community healthcare system. Patients were referred to the program through their treating physician. Adoption was evaluated by assessing the number of physicians within each specialty who have referred patients to the MD program.
Results:
550 patients were presented at MD conference. and 265 were seen at the MD clinic from 2014-2015. MD patients were younger, more likely to be female (p<0.01), and African-American (p<0.01). In patients <65 years old, the MD clinic had >twice the percentage of uninsured patients (p<0.01). In patients >65 years old the MD clinic had a higher percentage of commercially insured patients (p<0.01) (Table1). 71 physicians referred patients to the MD conference; the greatest concentration of specialties were hematology/oncology (31%) and internal medicine (21%). 39 physicians referred patients into the MD clinic, with the greatest frequency from internal (30%) and pulmonary medicine (25%). When comparing the number of active physicians by specialty, to those who referred patients into the MD conference, hematology/oncology had the greatest amount (40%), followed by pulmonary medicine (31%). For clinic referrals it was pulmonary medicine (24%), followed by hematology/oncology (13%). Figure 1
Conclusion:
A MD model has been implemented that can effectively reach underserved populations within the region. MD care was adopted primarily by oncologists, pulmonologists, and internists. Further efforts should be taken to expand physician adoption.
-
+
P3.01 - Poster Session with Presenters Present (ID 469)
- Event: WCLC 2016
- Type: Poster Presenters Present
- Track: Biology/Pathology
- Presentations: 1
- Moderators:
- Coordinates: 12/07/2016, 14:30 - 15:45, Hall B (Poster Area)
-
+
P3.01-031 - Impact of a Novel Lung Gross Dissection Method on Intrapulmonary Lymph Node Yield (ID 6119)
14:30 - 14:30 | Author(s): M.A. Ray
- Abstract
Background:
Incomplete retrieval of intrapulmonary lymph nodes and missed nodal metastasis are associated with worse-than-expected survival after (NSCLC) resection. We tested the nodal yield from a novel gross dissection method.
Methods:
multi-institutional prospective cohort study of intrapulmonary (stations 11-14) lymph node yield from lobectomy/greater NSCLC resection specimens from 11 US hospitals from 2009-2016. A novel gross dissection protocol was used in 2 hospital pathology departments from June 2012 onwards. Intrapulmonary lymph node yields from all lobectomy or greater resections before and after the new protocol in the intervention hospitals were compared to yields from 9 non-intervention hospitals over the same time-span, using Wilcoxon-Mann-Whitney. From February 2015, some randomly selected discarded remnant lung specimens in the intervention hospitals were re-dissected for inadvertently discarded lymph nodes as a quality control measure.
Results:
Intrapulmonary lymph node yields in the 2 groups of hospitals was similar at baseline, followed by a significant increase in the intervention hospitals with the novel dissection protocol (Table 1). Subsequently, in 112 specimens re-dissected for independent quality control after application of the novel dissection protocol, discarded lymph nodes were found in 30 (27%), down from 90% historically; discarded lymph nodes with metastasis in 6 (5%), down from 29% historically; and missed N1 nodal metastasis was found in 1 of 67 (1.5%) pN0 patients, down from 12% historically. The median number of missed intrapulmonary lymph nodes was 0 (down from 6 historically), the mean (standard deviation) was 0.88 (2.58). The gross dissection protocol required a median of 15 minutes (range 10 – 24). Figure 1
Conclusion:
A novel gross dissection protocol significantly improves the thoroughness of intrapulmonary lymph node retrieval and can be successfully implemented in community-level pathology departments, providing a pathway for quality improvement in pathologic nodal staging of resected NSCLC.
