Virtual Library
Start Your Search
Matthew Smeltzer
Author of
-
+
MA03 - New and Revisited Prognostic Factors in Early Stage Lung Cancer (ID 119)
- Event: WCLC 2020
- Type: Mini Oral
- Track: Early Stage/Localized Disease
- Presentations: 1
- Moderators:
- Coordinates: 1/29/2021, 15:30 - 16:30, Scientific Program Auditorium
-
+
MA03.05 - Hierarchy of Adverse-Prognosis Quality Markers after Curative-Intent Resection of Non-Small Cell Lung Cancer (NSCLC) (ID 2377)
15:30 - 16:30 | Author(s): Matthew Smeltzer
- Abstract
- Presentation
Introduction
Curative-intent resection for non-small cell lung cancer (NSCLC) often fails. Poor oncologic quality resection is associated with worse-than-expected survival. There is no universal benchmark of good quality resection, but the IASLC proposed a new definition of complete resection (IASLC-CR) including the quality of lymph node (LN) evaluation, besides margin status. Unquantified surgical quality deficits can induce unexpected imbalances, potentially confounding clinical trials involving patients with early-stage NSCLC, such as adjuvant therapy and biomarker trials. We evaluated the hierarchy of adverse surgical quality markers based on their individual and combined survival impact, using the IASLC-CR as a comparator.
Methods
We analyzed the population-based Mid-South Quality of Surgical Resection Cohort from 2009-2020. IASLC-CR requires: negative margins, systematic or lobe-specific nodal dissection, no extracapsular nodal invasion and non-involvement of the highest mediastinal LN. We evaluated hazard ratios (HR) and 95% confidence intervals (CI) from Proportional Hazards models.
Results
Of 3438 resections, 1131 (33%) had IASLC-CR. Among the 2307 not meeting IASLC-CR, 153 (6.6%) were R1/R2, 208 (9%) had no LN sampled (pNX), 130 (5.6%) underwent wedge resection, 285 (12.4%) had no mediastinal LN sampled (mediastinal NX), 906 (39.3%) had no subcarinal (station 7) LN sampled, and 271 (11.7%) had no hilar (station 10) LN sampled.
The hierarchical HR are shown in Table 1. Compared to IASLC-CR, HRs were: 2.60 (CI: 2.07-3.25) for R1/R2; 1.80 (CI: 1.39-2.32) for wedge resections; 1.65 (CI: 1.34-2.04) for pNX; 1.37 (CI: 1.12-1.67) for mediastinal NX; 1.37 (CI: 1.11-1.69) when station 10 LN was not examined; 1.24 (CI: 1.07-1.43) when station 7 LN was not examined. Patients with combinations of R1/R2 resections and other quality deficits had particularly poor survival. However, patients with combinations of sublobar resection and suboptimal nodal evaluation (pNX, mediastinal NX or missed key nodal stations- 7 or 10), or combinations of missed key nodal stations, such as mediastinal NX and missed station 10 (HR: 1.65, CI: 1.30-2.10) or missed stations 7 and 10 (HR: 1.59, CI: 1.38-1.84) also had significantly worse-than-expected survival.
Table 1 Independent effect
N (%)
Univariate HR (CI)
R1/R2
153 (11.9)
2.60 (2.07 - 3.25)
Wedge resection
130 (10.3)
1.80 (1.39 - 2.32)
Non-examination of lymph nodes (pNX)
208 (15.5)
1.65 (1.34 - 2.04)
Non-examination of mediastinal nodes (mNx)
285 (20.1)
1.37 (1.12 - 1.67)
Non-examination of station 10
271 (19.3)
1.37 (1.11 - 1.69)
Non-examination of station 7
906 (44.5)
1.24 (1.07 - 1.43)
Combined effect
R2/R1+ Mediastinal NX
15 (1.3)
4.38 (2.51 - 7.63)
R2/R1+ Missing station 7
80 (6.6)
3.15 (2.40 - 4.14)
Wedge resection + Mediastinal NX
18 (1.6)
2.98 (1.71 - 5.19)
pNX+ Segmentectomy
24 (2.1)
2.31 (1.43 - 3.71)
R2/R1+ Segmentectomy
5 (0.4)
2.16 (0.69 - 6.75)
R2/R1+ Missing station 10
57 (4.8)
1.97 (1.39 - 2.79)
R2/R1+ pNX
15 (1.3)
1.96 (1.01 - 3.80)
R2/R1+ Wedge resection
18 (1.6)
1.93 (1.03 - 3.62)
Wedge resection + Missing station 7
257 (18.5)
1.78 (1.47 - 2.17)
Segmentectomy + Missing station 10
69 (5.8)
1.76 (1.26 - 2.46)
Wedge resection + Missing station 10
225 (16.6)
1.70 (1.40 - 2.08)
pNX + Wedge resection
157 (12.2)
1.70 (1.34 - 2.15)
Mediastinal NX + Missing station 10
162 (12.5)
1.65 (1.30 - 2.10)
Missing station 7 + Missing station 10
845 (42.8)
1.59 (1.38 - 1.84)
Segmentectomy + Missing station 7
81 (6.7)
1.57 (1.14 - 2.17)
Segmentectomy + Mediastinal NX
11 (1.0)
1.48 (0.61 - 3.57)
Conclusion
The proposed IASLC definition of complete resection is a robust reference with which to quantify the adverse survival implications of potentially avoidable markers of oncologically unsound resection. Station-specific information is important in ensuring balanced comparison of NSCLC resection populations, such as those enrolled in clinical trials of novel adjuvant therapies and validation of biomarkers of residual postoperative surgical risk.
Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.
-
+
MA09 - Prognosis and Staging (ID 187)
- Event: WCLC 2020
- Type: Mini Oral
- Track: Staging
- Presentations: 3
- Moderators:
- Coordinates: 1/31/2021, 09:15 - 10:15, Scientific Program Auditorium
-
+
MA09.03 - Comparative Effectiveness of a Lymph Node Collection Kit Versus ‘Heightened Awareness’ on Lung Cancer Surgery Quality and Outcomes (ID 3726)
09:15 - 10:15 | Author(s): Matthew Smeltzer
- Abstract
Introduction
Poor-quality pathologic nodal staging is widespread and adversely impacts survival, stimulating policy-level efforts to overcome it. The American College of Surgeons Commission on Cancer (CoC) is heightening awareness by benchmarking the proportion of resections with examination of the hilar and at least 3 named mediastinal nodal stations at CoC-accredited hospitals. The IASLC’s proposed revised ‘Residual Disease’ (R-factor) classification includes lymph node evaluation quality- R0 resections which fail lymph node quality standards are recategorized as ‘R-Uncertain’. We compared the relative impact of heightened awareness of nodal staging quality versus a lymph node collection kit on nodal staging quality, adjuvant therapy eligibility, postoperative mortality, and survival.
Methods
We conducted a non-randomized stepped-wedge implementation study of a lymph node collection kit for curative-intent NSCLC resections in 12 institutions in four contiguous US Hospital Referral Regions. We categorized resections from 2009-2020 into three groups: pre-implementation baseline; post-implementation kit; and post-implementation non-kit (‘heightened awareness’) cases.
We used Chi-squared (Fisher for small sample size), Kruskall-wallis, Kaplan-Meier, and Cox regression analyses to examine differences in demographic and clinical outcomes across the three cohorts. Differences in proportion or means, with 95% confidence intervals, estimated effect size of the kit and heightened awareness interventions.
Results
Of 3715 resections: 39% (n=1437) were baseline; 40% (n=1502) kit; and 21% (n=776) non-kit, heightened awareness cases. Various patient-level demographic and clinical characteristics had similar distributions across all three cohorts.
Cohort proportions with unexamined lymph nodes (pathologic NX) were 11% (baseline) versus 0% (kit) versus 10% (heightened awareness), p<0.001; unexamined mediastinal lymph nodes were 27% versus 1% versus 22%, p<0.001. Proportions attaining the CoC benchmark were 23% versus 79% versus 35%, p<0.001; and proportions achieving IASLC-defined Complete Resection were 10% versus 57% versus 21%, p<0.001. In addition to improved surgical quality, duration of surgery was significantly shorter, post-operative complications were less frequent, and survival rate was significantly higher for kit cases compared to baseline and heightened awareness cases (Table 1).
Pairwise effect size estimates further highlight the differences: both kit and heightened awareness were associated with significant improvement over baseline quality metrics and survival (Table 1).
Conclusion
Heightened awareness and surgery with a lymph node collection kit significantly improved surgical quality and outcomes, but the kit was more effective. We propose to use these effect size estimates to determine sample size and statistical power for a prospective, institutional cluster randomized comparative effectiveness clinical trial of both interventions.
-
+
MA09.08 - Improving Overall Survival in Non-Small Cell Lung Cancer by Reducing R-Uncertain Resections With a Lymph Node Specimen Collection Kit (ID 1409)
09:15 - 10:15 | Presenting Author(s): Matthew Smeltzer
- Abstract
Introduction
The IASLC proposed a re-definition of residual disease (R-factor) classification, from complete (R0), microscopic incomplete (R1) and grossly incomplete (R2) to R0, ‘R-uncertain’, R1 and R2. The majority of R-uncertainty is caused by poor lymph node (LN) evaluation. The adverse prognostic implication of R-uncertainty has been independently validated. We previously demonstrated longer survival after surgical resection with a LN specimen collection kit, and now evaluate R-factor redistribution as the mechanism of its survival benefit.
Methods
The population-based Mid-South Quality of Surgical Resection cohort includes >95% of lung cancer resections within 4 hospital referral regions in the Mid-South USA from 2009-2019. We prospectively introduced a LN kit within this cohort using a staggered implementation design. We hypothesized that by improving nodal evaluation quality, the kit would increase the proportion of R0 and decrease R-uncertain resections; and would minimize the adverse impact of R-uncertainty by reducing extreme cases of R-uncertainty such as non-examination of LN (pNX) or mediastinal LN (mNX). We used multivariable logistic and proportional hazards regression, estimating adjusted odds ratios (aOR) and adjusted hazard ratios (aHR) with 95% confidence intervals (CI) while controlling for age, sex, histology, pT-category, pM-category, and comorbidities.
Results
Of 3,505 resections, 46% were female, 78% Caucasian/20% African-American, with median age 68 years; the LN kit was used in 39%. Overall, resections were 34% R0, 60% R-uncertain, and 6% R1/2. The R0 percentage increased from 9% in 2009 to 56% in 2019 (p<0.0001).
Kit cases were 66% R0 and 29% R-uncertain, compared to 14% R0 and 79% R-uncertain in non-kit cases (Table 1, p<0.0001). Compared to non-kit resections, kit resections had 12.6 times the adjusted odds of R0 vs. R-uncertain (Table 1, p<0.0001).
Of 2,100 R-uncertain resections, kit cases had lower percentages of pNX, 1% vs. 14% (p<0.0001) and mNX, 8% vs. 35% (p<0.0001). With the kit, more R-uncertain cases had station 7 (43% vs. 22%, p<0.0001) and station 10 (67% vs. 45%, p<0.0001) sampled.
The aHR for kit cases versus non-kit cases was 0.75 ([CI 0.66-0.85], p<0.0001); aHR for R0 versus R-uncertain was 0.78 ([CI 0.69-0.88], p<0.0001). In 2,100 subjects with R-uncertain resections, kit cases had an aHR of 0.79 versus non-kit cases ([CI 0.64-0.99], p=0.0384); however, in 1,199 R0 resections the survival difference between kit and non-kit cases was not significant (aHR: 0.85 [0.68-1.07], p=0.17).
R-Factor
Kit Used
aOR (95% CI)
N
No
Yes
R0
305 (14%)
894 (66%)
1.00 (--)
1199
R-Uncertain
1703 (79%)
397 (29%)
12.6 (10.6-14.9)
2100
R1/R2
141 (7%)
65 (5%)
6.4 (4.6-8.8)
206
Conclusion
The LN kit increases overall survival by increasing R0 and reducing R-uncertain rates; and also by diminishing extreme R-uncertainty (pNX and mNX).
-
+
MA09.10 - Location of Lymph Nodes Missed after Invasive Mediastinal Staging in a Well-Staged Non-Small Cell Lung Cancer Cohort (ID 2318)
09:15 - 10:15 | Author(s): Matthew Smeltzer
- Abstract
Introduction
Nodal stage is prognostic and predictive of adjuvant therapy benefit in potentially curable non-small cell lung cancer (NSCLC). Using a cohort of patients who met National Comprehensive Cancer Network (NCCN) pathologic nodal staging quality (examination of ≥1 N1 and ≥3 mediastinal nodal stations) as reference, we evaluated the thoroughness of invasive nodal staging to determine the frequency of non-examination of specific hilar/mediastinal nodes, and missed node metastasis. We also compared survival between NCCN-quality and non-NCCN quality-attaining resections on the basis of missed nodal metastasis on invasive staging. We hypothesized that the non-NCCN group without missed lymph node (LN) metastasis are at higher risk for missed pathologic nodal involvement, impairing their survival; and conversely, there will be less survival difference in patients with known missed LN metastasis.
Methods
We evaluated all curative-intent resections from 2009-2019 in 12 hospitals; compared patients meeting and not meeting NCCN criteria and those with and without invasive staging using Chi-squared and Wilcoxon tests. We used Log-rank tests and Cox regression to test differences in survival (hazards).
Results
Of 3,787 resections, 45% (n=1692) met NCCN criteria, of which 510 (30%) had invasive staging; versus 429 of 2095 (20%) patients not meeting NCCN criteria. Of invasively staged patients, 66% had mediastinoscopy, 46% had endobronchial ultrasound. Patients who met NCCN quality criteria were more often female (48% v 43%), white (80% v 76%), but had similar clinical stage distribution as those who did not meet NCCN criteria.In more than 50% of patients, stations 3, 2, 6, 5 and 10L were never examined (Table 1). The greatest discrepancy between NCCN and non-NCCN cohorts was in non-examination of stations 9, 7, and 8 (Table 1). Stations 10,7,8, 4, and 5 most frequently had missed nodal metastasis.
Invasively staged patients had higher hazards of death (Log-rank p<0.001, HR: 1.24, 95% CI [1.096-1.393]). NCCN quality resections had better survival (Log-rank p<0.001, HR: 0.77, 95% CI [0.70, 0.86]). Invasively-staged NCCN-quality resections without missed node metastasis had better survival (Log-rank p=0.0293, HR:0.745, 95% CI[0.571, 0.972] ); there was no survival difference between NCCN and non-NCCN resections among those without missed metastasis (Log-rank p=0.4585, HR: 0.87, 95% CI [0.605, 1.254]).
Conclusion
Stations 3, 5 and 6, which are relatively inaccessible to pre-surgical biopsy, are infrequently examined at surgery. Stations 2, 4L and 10L which are accessible are also infrequently examined. Poor pathologic nodal examination remains harmful even in patients with negative invasive clinical staging tests.
-
+
P50 - Small Cell Lung Cancer/NET - Real World Outcomes (ID 232)
- Event: WCLC 2020
- Type: Posters
- Track: Small Cell Lung Cancer/NET
- Presentations: 1
- Moderators:
- Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall
-
+
P50.11 - Long-Term Survival after Surgical Resection of Carcinoid Tumors in a Population-Based Cohort (ID 3258)
00:00 - 00:00 | Author(s): Matthew Smeltzer
- Abstract
Introduction
Pulmonary carcinoid (PC) tumors constitute approximately 2% of all lung malignancies. Because of their rarity, management of these tumors remains controversial, although surgical management generally follows similar recommendations as non-small cell lung cancer. We sought to define characteristics of typical carcinoids (TCs) and atypical carcinoids (ACs) and identify factors that predict overall survival (OS) after curative-intent surgical resection.
Methods
We examined patients from a population-based cohort who underwent lung resection for PCs from 12 hospitals across 7 different healthcare systems from 2009 to 2019. Patients were staged according to the 8th edition of the tumor, node, and metastasis (TNM) staging system. TCs, ACs, and PCs with unspecified histology were compared using Chi-square/Fisher’s exact, or Kruskal-Wallis tests. 5-year OS was analyzed using Kaplan-Meier curves and survival between groups was compared with the log-rank test. Univariate and multivariable Proportional Hazards models were used to determine predictors of OS.
Results
We identified 148 patients. 76 were TCs, 31 ACs, and 41 had unspecified histology. Median age was 65 years (range 22-83) and the majority were female (71%). 20% were African American; 46% were nonsmokers. Pathologic stages were 0 (3%), I (74%), II (16%), and III (8%). Surgical procedures were 8% wedge resection, 76% lobectomy, 5% bilobectomy, 4% pneumonectomy, 5% segementectomy, and 2% lobectomy+wedge. Median follow-up time was 42 months. Patients with ACs were significantly older (p=0.0004) and more likely to be later pathologic stage (p=0.0175) than TCs/unspecifieds but did not differ in sex, race, smoking status, surgical technique, extent of resection, number of lymph nodes sampled, or 30-day postoperative mortality. The combined 5-year OS rate was 84% (95% CI: 74%-91%); 49% for atypical, 91% typical (p= 0.0003). Patients with pathologic N1/N2 nodal metastases had a significantly decreased OS compared to those without nodal disease (p=0.0048). Multivariable models showed that atypical histology (p=0.0055), male sex (p=0.0093), and pathologic N1/N2 nodal metastases (p=0.0281) remained significant negative predictors of OS after adjusting for age at surgery.
Demographic/Clinical Characteristics
Category
N (%)
Typical
76 (51)
Atypical
31 (21)
Not Specified
41 (28)
Total
N=148
p-value
Age
median (range)
63.5 (34-83)
69 (44-78)
61 (22-82)
65 (22-83)
0.0004
Sex
Male
20 (26)
12 (39)
11 (27)
43 (29)
0.4113
Female
56 (74)
19 (61)
30 (73)
105 (71)
Race
Caucasian
61 (80)
22 (71)
35 (85)
118 (80)
0.3179
Black or African-American
15 (20)
9 (29)
6 (15)
30 (20)
Smoking Status
Active
16 (21)
7 (23)
4 (10)
27 (18)
0.152
Former
19 (25)
14 (45)
12 (29)
45 (30)
Never
36 (47)
10 (32)
22 (54)
68 (46)
Not Reported
5 (7)
0 (0)
3 (7)
8 (5)
Number of patients with any lymph node sampled during surgery
4 (5)
3 (10)
1 (2)
8 (5)
0.4436
Total number of lymph nodes sampled
median (range)
3.5 (1-4)
7 (2-13)
1 (1)
3.5 (1-13)
0.2484
Aggregate pathologic Stage (8th)
Occult/Stage 0
2 (3)
2 (6)
0 (0)
4 (3)
0.0175
Stage I
61 (80)
15 (48)
33 (80)
109 (74)
Stage II
9 (12)
8 (26)
6 (15)
23 (16)
Stage III
4 (5)
6 (19)
2 (5)
12 (8)
Pathologic T stage (8th)
pTX/pT0
1 (1)
1 (3)
0 (0)
2 (1)
0.0182
pT1
60 (79)
17 (55)
29 (71)
106 (72)
pT2
14 (18)
10 (32)
8 (20)
32 (22)
pT3
0 (0)
3 (10)
4 (10)
7 (5)
pT4
1 (1)
0 (0)
0 (0)
1 (1)
Pathologic N stage (8th)
pNX
3 (4)
3 (10)
1 (2)
7 (5)
0.1459
pN0
61 (80)
18 (58)
35 (85)
114 (77)
pN1
9 (12)
6 (19)
4 (10)
19 (13)
pN2
3 (4)
4 (13)
1 (2)
8 (5)
Pathologic M stage (8th)
pM0
76 (100)
31 (100)
41 (100)
.
30-day postoperative mortality
No
73 (96)
29 (94)
41 (100)
143 (97)
0.3375
Yes
3 (4)
2 (7)
0 (0)
5 (3)
Conclusion
Our results are consistent with previous reports showing that TCs and ACs have different disease characteristics, with ACs having a worse prognosis following resection. However, it is unknown to what extent other prognostic factors are involved. We found that presence of nodal metastasis (N1/N2) and sex were additional predictors of OS for patients with carcinoid tumors. Future studies should assess the involvement of pathologic stage and lymph node involvement which could lead to an improvement in prognostication and patient management.
