Procedure-Specific Risk Prediction for Recurrence in Patients Undergoing Lobectomy or Sublobar Resection for Small (≤2 cm) Lung Adenocarcinoma: An International Cohort Analysis

Procedure-Specific Risk Prediction for Recurrence in Patients Undergoing Lobectomy or Sublobar Resection for Small (≤2 cm) Lung Adenocarcinoma: An International Cohort Analysis

ORIGINAL ARTICLE Procedure-Specific Risk Prediction for Recurrence in Patients Undergoing Lobectomy or Sublobar Resection for Small (2 cm) Lung Adeno...

1MB Sizes 0 Downloads 40 Views

ORIGINAL ARTICLE

Procedure-Specific Risk Prediction for Recurrence in Patients Undergoing Lobectomy or Sublobar Resection for Small (2 cm) Lung Adenocarcinoma: An International Cohort Analysis Sarina Bains, MD,a Takashi Eguchi, MD,a,b Arne Warth, MD,c,d Yi-Chen Yeh, MD,e Jun-ichi Nitadori, MD,f Kaitlin M. Woo, MS,g Teh-Ying Chou, MD, PhD,e,h Hendrik Dienemann, MD,d,i Thomas Muley, PhD,d,j Jun Nakajima, MD,f Aya Shinozaki-Ushiku, MD,k Yu-Chung Wu, MD,l Shaohua Lu, MD,m,n Kyuichi Kadota, MD, PhD,m,o David R. Jones, MD,a William D. Travis, MD,m Kay See Tan, PhD,g Prasad S. Adusumilli, MDa,p,* a

Thoracic Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, New York, New York Division of Thoracic Surgery, Department of Surgery, Shinshu University, Matsumoto, Japan c Institute of Pathology, Heidelberg University, Heidelberg, Germany d Translational Lung Research Center Heidelberg, German Center for Lung Research, Heidelberg, Germany e Department of Pathology and Laboratory Medicine, Taipei Veterans General Hospital, Taipei, Taiwan f Department of Thoracic Surgery, University of Tokyo, Tokyo, Japan g Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, New York h Institute of Clinical Medicine, School of Medicine, National Yang-Ming University, Taipei, Taiwan i Department of Thoracic Surgery, Thoraxklinik at Heidelberg University, Heidelberg, Germany j Translational Research Unit, Thoraxklinik at Heidelberg University, Heidelberg, Germany k Department of Pathology, University of Tokyo, Tokyo, Japan l Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan m Department of Pathology, Memorial Sloan Kettering Cancer Center, New York, New York n Department of Pathology, Zhongshan Hospital, Fudan University, Shanghai, People’s Republic of China o Department of Diagnostic Pathology, Kagawa University, Kagawa, Japan p Center for Cell Engineering, Memorial Sloan Kettering Cancer Center, New York, New York b

Received 2 July 2018; revised 23 August 2018; accepted 16 September 2018 Available online - 22 September 2018

ABSTRACT Introduction: This work was performed to develop and validate procedure-specific risk prediction for recurrence following resection for early-stage lung adenocarcinoma (ADC) and investigate risk prediction utility in identifying patients who may benefit from adjuvant chemotherapy (ACT). Methods: In patients who underwent resection for small (2 cm) lung ADC (lobectomy, 557; sublobar resection, 352), an association between clinicopathologic variables and risk of recurrence was assessed by a competing risks approach. Procedure-specific risk prediction was developed based on multivariable regression for recurrence. External validation was conducted using cohorts (N ¼ 708) from Japan, Taiwan, and Germany. The accuracy of risk prediction was measured using a concordance index. We applied the lobectomy risk prediction approach to a propensity score–matched cohort of patients with stage II-III disease

(n ¼ 316, after matching) with or without ACT and compared lung cancer–specific survival between groups among low- or high-risk scores. Results: Micropapillary pattern, solid pattern, lymphovascular invasion, and necrosis were involved in the risk

*Corresponding author. Drs. Bains and Eguchi contributed equally to this work. Disclosure: Dr. Muley has received a grant from Roche Diagnostics. Dr. Travis has consulted for Genentech. The remaining authors declare no conflict of interest. Address for correspondence: Prasad S. Adusumilli, MD, Thoracic Service, Department of Surgery, Mesothelioma Program, Memorial Sloan Kettering Cancer Center, 1275 York Avenue, New York, New York 10065. E-mail: [email protected] ª 2018 International Association for the Study of Lung Cancer. Published by Elsevier Inc. All rights reserved. ISSN: 1556-0864 https://doi.org/10.1016/j.jtho.2018.09.008

Journal of Thoracic Oncology

Vol.

-

No.

-: ---

2 Bains et al

prediction following lobectomy, and micropapillary pattern, spread through air spaces, lymphovascular invasion, and necrosis following sublobar resection. Both internal and external validation showed good discrimination (concordance index in lobectomy and sublobar resection: internal, 0.77 and 0.75, respectively; and external, 0.73 and 0.79, respectively). In the stage II-III propensity score–matched cohort, among high-risk patients, ACT significantly reduced the risk of lung cancer–specific death (subhazard ratio 0.43, p ¼ 0.001), but not among low-risk patients. Conclusions: Procedure-specific risk prediction for patients with resected small lung ADC can be used to better prognosticate and stratify patients for further interventions.  2018 International Association for the Study of Lung Cancer. Published by Elsevier Inc. All rights reserved. Keywords: Adjuvant chemotherapy; Competing risks analysis; Lung cancer–specific death; Recurrence; Sublobar resection

Introduction Lung adenocarcinoma (ADC) is the most common histologic subtype of NSCLCs; 25% of which are diagnosed at stage IA.1 Following the results of the National Lung Screening Trial, the identification of early-stage lung ADC is expected to increase.2 The standard of care for early-stage lung ADC is curative-intent anatomic surgical resection by lobectomy; however, sublobar resection is appropriate for selected patients.3 Despite ongoing concerns about the adequacy of sublobar resection for cure, the use of sublobar resection is increasing.4-7 Development of a procedure-specific risk prediction model following sublobar resection or lobectomy that considers widely variable patient background and the recurrence risk will be useful to predict prognosis.4,5,8 Such a risk model can further help to stratify patients for prospective investigation of potential postoperative interventions (e.g., completion lobectomy and/or adjuvant therapies following sublobar resection or adjuvant therapies following lobectomy). Based on our literature review, among 10 risk prediction algorithms that have been described to predict prognosis following lung resection for all stages of NSCLC (Supplementary Fig. 1 and Supplementary Table 1), 6 have not been externally validated.9-18 Of the 4 that have, only two studies externally validated with an international cohort.9,10 No study addressed competing risks, which can bias prognostic assessment especially in earlystage disease.8 No study included prognostic histologic subtypes.19 A recent study by Liang et al.10 was independently validated using the National Cancer Database

Journal of Thoracic Oncology

Vol.

-

No.

-

(United States); however, the TNM staging system still had a superior predictive capability.20 In an effort to expand the prognostication of lung ADC beyond the use of the TNM staging system, a multidisciplinary group comprising experts from the International Association for the Study of Lung Cancer (IASLC), American Thoracic Society (ATS), and European Respiratory Society (ERS) introduced a classification of lung ADC in 2011,19 which has been validated in independent international cohorts and was incorporated into the 2015 WHO classification.21-24 We and others reported the prognostic impact of micropapillary (MIP) and solid (SOL) subtypes in stage I lung ADC and were the first to describe tumor spread through air spaces (STAS), which has since been validated by others. 22,23,25-31 These prognostic pathologic variables including MIP, SOL, and STAS can be associated with each other.29 To our knowledge, no group has investigated how these variables interact and influence outcomes. To incorporate multiple, relevant clinicopathologic factors for a given patient, we specifically chose a risk-based prediction model.32 In this study, we developed two procedure-specific scores for predicting recurrence after curative-intent lobectomy or sublobar resection. Given the significant competing risks in elderly patients with early-stage lung cancer, we used a competing risks analysis.8 We validated our model with an external data set consisting of three cohorts of patients from Japan, Taiwan, and Germany. The current guidelines do not recommend platinumbased adjuvant chemotherapy (ACT) for patients with stage IA NSCLC.33-35 However, if we could reliably identify stage IA patients with a higher risk of recurrence equal to stage II-III NSCLC patients, we may be able to predict the benefit from ACT. To test this potential utility of this approach, we would ideally use a cohort of patients with stage IA lung ADC. However, because only a minor fraction of stage IA lung ADC (<5%) undergo ACT both at our and other centers, we used a propensityscore matched cohort of patients with resected stage II-III lung ADC with or without ACT to investigate the survival benefit from ACT in those patients classified as high risk using the proposed lobectomy risk prediction score.

Methods Study Cohorts for Development and Validation of Risk Prediction for Recurrence This retrospective study was approved by the Institutional Review Board at Memorial Sloan Kettering Cancer Center (MSK). The MSK Thoracic Service’s prospectively maintained lung cancer database was reviewed to identify consecutive patients who had been

---

2018

surgically treated for pathologic stage I small (2 cm) lung ADC between January 1, 1995, and December 31, 2011. At MSK, lobectomy was considered a standard surgical treatment for patients with stage I NSCLC. Sublobar resection was considered for patients with poor pulmonary reserve or other major comorbidities that relatively contraindicate lobectomy or for patients with a small (2 cm) peripheral nodule with 50% or more ground-glass appearance on computed tomography. Pathologic stage was based on the eighth edition of the American Joint Committee on Cancer Staging Manual.36 Exclusion criteria included induction therapy, multiple nodules, positive surgical margin (R1 or R2), other lung cancer diagnosis in the past 2 years, other disease progression, and inadequate tissue available to review. In total, 909 patients at MSK met the inclusion criteria (Supplementary Fig. 2A). The development (MSK) cohort of patients was used to create the two procedure-specific risk scores. To externally validate the risk prediction, we used a combined cohort of three independent international sets of patients (Japan, Taiwan, and Germany) (Supplementary Fig. 2B). The University of Tokyo (Japan) cohort consisted of 250 consecutive patients treated between September 1, 1998, and December 31, 2012. The Taipei Veterans General Hospital (Taiwan) cohort consisted of 259 consecutive patients treated between April 1, 1996, and December 31, 2010. All tumor slides were reviewed by Y.Y. The Heidelberg University (Germany) cohort consisted of 199 consecutive patients treated between April 1, 2002, and December 31, 2014. All tumor slides were reviewed by A.W. Inclusion and exclusion criteria were the same as those for the primary cohort. Histologic evaluation was performed by experienced pathologists in each institution (A.S. at the University of Tokyo, Y.Y. at Taipei Veterans General Hospital, and A.W. at Heidelberg University) following the same method as the primary cohort, as described later in this Methods section.

Stage II-III Cohort The MSK Thoracic Service’s prospectively maintained database was reviewed to identify consecutive patients who had been surgically treated for pathologic stage II and III lung ADC between January 1, 2000, and December 31, 2013. Exclusion criteria included induction therapy, multiple nodules, positive surgical margin (R1 or R2), other lung cancer diagnosis in the past 2 years, concurrent other disease progression, wedge resection, and no available tumor slides to review. To evaluate benefit from platinum-based (cisplatin or carboplatin plus other drug) ACT, we also excluded patients with unknown ACT status, unknown regimens, nonplatinum-based ACT, adjuvant EGFR–tyrosine kinase inhibitor therapy, and less than 2 cycles of platinum-

Risk Prediction for Recurrence

3

based ACT. Because perioperative mortality and morbidity can affect treatment decisions as well as outcomes, we excluded patients who died within 90 days of surgery and patients with planned but cancelled ACT due to postoperative morbidity or recurrence. In total, 589 patients with stage II-III lung ADC met our inclusion criteria (Supplementary Fig. 1C). In this study, ACT was defined as any additional intravenously administered chemotherapy after the primary surgery, within 3 months after surgery, without recurrence of the resected primary tumor. We evaluated the regimen of ACT, the date of the first dose, and the number of performed cycles of ACT. In patients who did not receive ACT, we evaluated the specific reason for this.

Histologic Evaluation Tumor slides were reviewed by at least two experienced thoracic pathologists (K.K., L.S., and W.D.T.) who were blinded to patient clinical outcomes. The percentage of each histologic pattern was recorded in 5% increments, and tumors were classified by the predominant subtype (in accordance with the IASLC/ ATS/ERS and 2015 WHO classifications): ADC in situ; minimally invasive adenocarcinoma; lepidic-, acinar-, papillary-, MIP-, or SOL-predominant invasive adenocarcinoma; invasive mucinous ADC; and colloid ADC.19,24 Tumor STAS was defined as isolated tumor cells within air spaces surrounding the main tumor.29 The presence of visceral pleural, lymphovascular invasion (LVI), and necrosis was also investigated.

Statistical Methods To develop procedure-specific risk prediction, patients who underwent lobectomy were analyzed separately from those who underwent sublobar resection. The outcome of interest was recurrence. Patients were monitored from date of surgery until recurrence or death, whichever came first. The probability of recurrence was estimated as the cumulative incidence of recurrence (CIR) from the time of surgery and analyzed using a competing risks approach, considering deaths without recurrence as competing events.37 Differences in CIR between groups were assessed using Gray’s method and univariable Fine and Gray’s tests. Definitions of the clinicopathologic factors included in the univariable analyses are presented in the Supplementary Materials. Factors that yielded p < 0.1 in univariable analyses were considered as candidates in the multivariable models. Technical information on imputation of missing data by predictive mean matching and model building using the adaptive lasso procedure is presented in the Supplementary Materials. The procedure-specific risk prediction was developed using recurrence probability estimates derived from the

4 Bains et al

final multivariable competing risks models. The predictions of recurrence in both models were projected at 3 years and 5 years. The median follow-up duration for the development and combined external cohorts were computed based on reverse Kaplan-Meier approach. In both models, categorical covariates were included as dummy variables, and nonlinearity of continuous variables was assessed using restricted cubic splines.38 Cuberoot transformation was applied to MIP and SOL percentage, given the right-skewed distribution data with a high proportion of zeros. On further investigation of multicollinearity, pleural invasion was excluded from the variable-selection procedure due to its high correlation with LVI. Development of risk prediction was performed with blinding of the external-validation cohort data. The predictive performance of the procedure-specific risk prediction was assessed by examining discrimination (concordance index [C-index]), calibration (calibration plots), and overall accuracy (Brier score). We also generated decision curves to assess the net benefit of risk prediction assisted decisions. Additional information on the performance measures of the risk prediction is presented in the Supplementary Materials. Internal validations were performed with 1000 bootstrap resamples. To complete external validation, we applied the procedure-specific risk prediction to data from Japan, Taiwan, and Germany. The external-validation cohort comprised 708 patients (lobectomy, 551; sublobar resection, 157). Because of the limited sample size and the incidence of recurrence, the external cohorts were analyzed as a combined cohort. As an exploratory analysis, overall survival (OS) for patients with high and low risk of recurrence was estimated using the KaplanMeier approach and compared using log-rank tests (additional information in the Supplementary Materials). To investigate the potential utility of risk prediction to benefit from ACT in high-risk patients, we developed a propensity score–matched cohort of patients with stage II-III disease who were treated by lobectomy with or without ACT and applied the lobectomy risk prediction score to the matched cohort. The propensity score– matching procedure selects matched pairs with similar baseline probability of being in either the ACT or the noACT group.39,40 For matching, we used age, sex, smoking status, comorbidities (chronic obstructive pulmonary disease, cardiovascular disease, etc.), serum creatinine level, pulmonary function (forced expiratory volume in 1 second, and diffusion capacity of the lung for carbon monoxide), maximum standard uptake value in (18) F-fluorodeoxyglucose–positron-emission tomography, type of resection, surgical approach, pathologic tumor size, p-N status, p-stage, pleural invasion, LVI, necrosis, tumor STAS, and morphologic grade. To account for the multiple imputations of the missing data, separate

Journal of Thoracic Oncology

Vol.

-

No.

-

logistic regression models were generated from each imputed data set (additional information in the Supplementary Materials). Propensity score–matched pairs were identified without replacement using a 1:1 nearest neighbor matching with caliper width equal to 0.286. The caliper width was determined by the recommendation from Austin (0.2 of the standard deviation of the logit of the propensity scores).41 Balance of covariates between the groups was assessed by the absolute standardized mean difference (ASMD) before and after the matching procedure. ASMD < 0.1 indicates balance in the covariate between the two groups.42 We investigated the association between receipt of ACT and various outcomes using the matched cohort: primary outcome, lung cancer–specific cumulative incidence of death (CID); secondary outcomes, CIR, and OS. The relationship was characterized by the patient-level score calculated from the lobectomy risk prediction. We categorized as high or low risk based on the predetermined 5-year CIR 18% (lobectomy score 95) that was a previously reported average 5-year CIR in stage I lung ADC.28 Hence, we categorized each patient as “high score” (score is at or above 95) or “low score.” OS was estimated using the Kaplan-Meier approach and compared between the two groups on the basis of the log-rank test, stratified by pathologic stage. Statistical analyses were conducted using R 3.3.1 (R Development Core Team, Austria, Vienna), including the “survival,” “cmprsk,” “crrp,” “ClevClinicQHS,” “QHScrnomo,” “rms,” and “pec” packages, downloaded in January 2017.

Results Patient Characteristics The development (MSK) cohort comprised 909 patients (lobectomy, 557; sublobar resection, 352). The external-validation cohort (N ¼ 708) included patients from Japan (n ¼ 250; lobectomy/sublobar resection, 165/85), Taiwan (n ¼ 259; lobectomy/sublobar resection, 211/48), and Germany (n ¼ 199; lobectomy/sublobar resection, 175/24). Patient clinicopathologic characteristics and outcomes for each cohort, as well as a comparison between the development and externalvalidation cohorts, are shown in Table 1. Among patients who underwent lobectomy, the development cohort was older and had higher proportion of women, larger invasive size tumor, LVI, MIP (5%), and lower proportion of low-grade tumors than the externalvalidation cohorts. The development sublobar resection cohort had higher proportion of women, larger invasive size tumor, LVI, necrosis, MIP (5%), SOL (5%), and lower proportion of low-grade tumors than the externalvalidation cohorts. In patients who underwent lobectomy, the number of distant recurrence events was

Sublobar resection External cohort

External cohort

MSK n ¼ 557

Combined N ¼ 551

Japan n ¼ 165

Taiwan n ¼ 211

Germany n ¼ 175

246 (44) 311 (56)

313 (57) 238 (43)

83 (50) 82 (50)

130 (62) 81 (38)

100 (57) 75 (43)

340 (61) 217 (39)

283 (51) 268 (49)

83 (50) 82 (50)

114 (54) 97 (46)

86 (49) 89 (51)

93 (17) 464 (83) 0

151 (27) 257 (47) 143

84 (51) 81 (49) 0

46 (22) 33 (16) 132

21 (12) 143 (82) 11

498 (89) 59 (11)

417 (76) 134 (24)

152 (92) 13 (8)

147 (70) 64 (30)

118 (67) 57 (33)

88 (16) 469 (84)

110 (20) 441 (80)

40 (24) 125 (76)

43 (20) 168 (80)

27 (15) 148 (85)

198 (36) 359 (64)

270 (49) 281 (51)

109 (66) 56 (34)

104 (49) 107 (51)

57 (33) 118 (67)

338 (61) 219 (39)

397 (72) 154 (28)

130 (79) 35 (21)

180 (85) 31 (15)

87 (50) 88 (50)

498 (89) 59 (11)

417 (76) 134 (24)

152 (92) 13 (8)

147 (70) 64 (30)

118 (67) 57 (33)

444 (80) 108 (19) 5

457 (83) 94 (17) 0

145 (88) 20 (12)

160 (76) 51 (24)

152 (87) 23 (13)

80 (14) 352 (63) 125 (22)

169 (31) 264 (48) 118 (21)

83 (50) 55 (33) 27 (16)

71 (34) 95 (45) 45 (21)

15 (9) 114 (65) 46 (26)

p MSK vs. Comba

MSK N ¼ 352

Combined n ¼ 157

Japan n ¼ 85

Taiwan n ¼ 48

Germany n ¼ 24

120 (34) 232 (66)

59 (38) 98 (62)

34 (40) 51 (60)

18 (38) 30 (63)

7 (29) 17 (71)

224 (64) 128 (36)

61 (39) 96 (61)

37 (44) 48 (56)

17 (35) 31 (65)

7 (29) 17 (71)

52 (15) 300 (85) 0

43 (27) 79 (50) 35

37 (44) 48 (56) 0

5 (10) 9 (19) 34

1 (4) 22 (92) 1

296 (84) 56 (16)

132 (84) 25 (16)

79 (93) 6 (7)

34 (71) 14 (29)

19 (79) 5 (21)

114 (32) 238 (68)

63 (40) 94 (60)

43 (51) 42 (49)

13 (27) 35 (73)

7 (29) 17 (71)

191 (54) 161 (46)

107 (68) 50 (32)

71 (84) 14 (16)

26 (54) 22 (46)

10 (42) 14 (58)

229 (65) 123 (35)

130 (83) 27 (17)

75 (88) 10 (12)

43 (90) 5 (10)

12 (50) 12 (50)

296 (84) 56 (16)

132 (84) 25 (16)

79 (93) 6 (7)

34 (71) 14 (29)

19 (79) 5 (21)

274 (78) 76 (22) 2

143 (91) 14 (9) 0

79 (93) 6 (7)

45 (94) 3 (6)

19 (79) 5 (21)

86 (24) 189 (54) 77 (22)

85 (54) 43 (27) 29 (18)

62 (73) 12 (14) 11 (13)

21 (44) 16 (33) 11 (23)

2 (8) 15 (63) 7 (29)

<0.001

p MSK vs. Comba 0.7

<0.001

0.001

<0.001

<0.001

<0.001

1

<0.001

0.11

<0.001

0.003

<0.001

<0.001

<0.001

1

<0.001

0.3

<0.001

<0.001

(continued)

Risk Prediction for Recurrence

Variables Age at surgery 65 >65 Sex Female Male Smoking history Never Former /current Unknown p-stage IA IB Tumor size 1 cm >1 cm Invasive tumor size 1 cm >1 cm LVI Absent Present Pleural invasion Absent Present Necrosis Absent Present Unkonwn Histologic gradeb Low Intermediate High

2018

Lobectomy

---

Table 1. Clinicopathologic and Demographic Characteristics for each Cohort and Comparison Between the Memorial Sloan Kettering Cancer Center (Development) Cohort and the Combined External Cohort

5

6 Bains et al

Table 1. Continued Lobectomy

Sublobar resection External cohort

MSK n ¼ 557

Combined N ¼ 551

Japan n ¼ 165

Taiwan n ¼ 211

Germany n ¼ 175

287 (52) 270 (48)

365 (66) 186 (34)

147 (89) 18 (11)

114 (54) 97 (46)

104 (59) 71 (41)

353 (63) 204 (37)

367 (67) 184 (33)

113 (68) 52 (32)

156 (74) 55 (26)

98 (56) 77 (44)

366 (66) 191 (34)

350 (64) 201 (36)

121 (73) 44 (27)

133 (63) 78 (37)

56 13 43

71 19 52

14 3 11

124 39 85 11 (81, 13) 88 (85, 91) 80 (76, 83)

74 38 36 13 (10, 17) 89 (86, 93) 61 (57, 63)

15 7 8 9 (5, 15) 92 (88, 97) 63 (54, 68)

p MSK vs. Comba

MSK N ¼ 352

Combined n ¼ 157

Japan n ¼ 85

Taiwan n ¼ 48

Germany n ¼ 24

197 (56) 155 (44)

121 (77) 36 (23)

79 (93) 6 (7)

27 (56) 21 (44)

15 (63) 9 (38)

220 (63) 132 (38)

118 (75) 39 (25)

69 (81) 16 (19)

37 (77) 11 (23)

12 (50) 12 (50)

96 (55) 79 (45)

226 (64) 126 (36)

113 (72) 44 (28)

73 (86) 12 (14)

27 (56) 21 (44)

13 (54) 11 (46)

30 8 22

27 8 19

73 43 30

23 15 8

10 6 4

10 7 3

3 2 1

31 15 16 13 (9, 18) 91 (86, 95) 78 (59, 86)

28 16 12 18 (12, 26) 85 (78, 92) 51 (44, 57)

129 54 75 21 (17, 26) 72 (67, 77) 70 (67, 76)

27 11 16 15 (9, 21) 84 (77, 90) 56 (50, 63)

11 5 6 12 (6, 20) 90 (83, 97) 56 (43, 67)

10 5 5 18 (10, 31) 81 (70, 94) 63 (49, 79)

6 1 5 19 (6, 45) 61 (39, 97) 35 (20, 47)

<0.001

<0.001

0.3

0.006

0.5

0.025e 0.085f

p MSK vs. Comba

0.1

0.3e 0.004f

Data are number (%). a Comparison between MSK versus combined external cohorts. b Histologic grade based on predominant subtypes (low grade, lepidic; intermediate, acinar or papillary; high, micropapillary, solid, invasive mucinous, or colloid). c Only locoregional recurrence without distant recurrence. d Data are shown as percentage (95% confidence interval). e Gray’s test. f Log-rank test. g Estimated median follow-up time (95% confidence interval). CIR, cumulative incidence of recurrence; LVI, lymphovascular invasion; MIP, micropapillary; MSK, Memorial Sloan Kettering Cancer Center; OS, overall survival; SOL, solid; STAS, spread through air spaces.

Journal of Thoracic Oncology

MIP Absent (<5%) Present (5%) SOL Absent (<5%) Present (5%) STAS Absent Present Outcomes No. of recurrence Any Locoregional c Distant No. of death Any Recurrence (+) Recurrence (-) 5-yr CIR (%)d 5-yr OS (%)d Median follow-up (months)g

External cohort

Vol. -

No. -

---

2018

higher than that of locoregional recurrence events in both the development and the external-validation cohorts (number of events for locoregional versus distant: development, 13 versus 43; external-validation, 19 versus 52). In contrast, in patients who underwent sublobar resection, the number of distant recurrence events was lower than that of locoregional recurrence events in both the development and the externalvalidation cohorts (number of events for locoregional versus distant: development, 43 versus 30; externalvalidation, 15 versus 8). Estimated 5-year CIR and OS are available in Table 1. The CIR and OS curves for each cohort are shown in the Supplementary Figure 3.

Lobectomy Risk Prediction Score In the multivariable model for recurrence following lobectomy, four risk factors (MIP percentage, SOL percentage, LVI, and necrosis) were significantly associated with hazard of recurrence lobectomy (Table 2). The resulting lobectomy risk prediction (Fig. 1) and scoring systems (Supplementary Table 2) were generated using the model estimates. In internal validation, the optimismcorrected C-index was 0.77 (95% confidence interval [CI]: 0.72–0.83), indicating good discrimination. In external validation, the C-index was 0.73 (95% CI: 0.67– 0.79), indicating good discrimination (Supplementary Table 3). The C-indices of the lobectomy risk prediction were higher than those of the TNM classification (risk prediction score versus TNM: internal 0.74 versus 0.58; and external 0.73 versus 0.68) (Supplementary Table 4). The calibration plots at 5 years for the development (MSK) and external-validation cohorts indicate moderate calibration (Fig. 2). Brier scores (Supplementary Table 3) and decision curves (Supplementary Fig. 4) are presented in the Supplementary materials.

Sublobar-Resection Risk Prediction Score In the multivariable model for recurrence following sublobar resection, four risk factors (MIP percentage, STAS, LVI, and necrosis) were significantly associated with hazard of recurrence after sublobar resection (Table 2). The resulting sublobar-resection risk prediction (Fig. 1) and scoring system (Supplementary Table 2) were generated using the model estimates. In internal validation, the optimism-corrected C-index was 0.75 (95% CI: 0.72–0.81), indicating good discrimination. In external validation, the C-index was 0.79 (95% CI: 0.73– 0.87), indicating good discrimination (Supplementary Table 3). The C-indices of the sublobar resection risk prediction score were higher than those of the TNM classification (risk prediction score versus TNM: internal 0.75 versus 0.60; and external 0.79 versus 0.78)

Risk Prediction for Recurrence

7

(Supplementary Table 4). The calibration plots at 5 years for the development and external-validation cohorts indicate moderate calibration (Fig. 2). Brier scores (Supplementary Table 3) and decision curves (Supplementary Fig. 4) are presented in the Supplementary materials.

Distribution of Patients in Each Cohort by Type of Procedure and Risk of Recurrence The distribution of patients who underwent sublobar resection was highest in the MSK cohort (39%), followed by the Japan (34%), Taiwan (19%), and Germany (12%) cohorts (Supplementary Fig. 5A). The distribution of patients categorized by probability of 5-year CIR estimated by procedure-specific risk prediction in each cohort is shown in Supplementary Figure 5. The recurrence risk–based patient distribution was different between cohorts especially in patients who were categorized into the lowest risk category. The Japan and Taiwan cohorts were composed more heavily of patients in the lowest risk category (lobectomy, 59% and 47%, respectively; and sublobar resection, 76% and 52%, respectively) than the composition of the MSK and Germany cohorts (lobectomy, 29% and 26%, respectively; and sublobar resection, 37% and 29%, respectively).

Exploratory OS Analysis by High/Low Risk of Recurrence OS curves in the development and external validation cohorts, by high/low risk of recurrence (based on risk prediction score), are shown in Supplementary Figure 6. In the development cohort, patients with a high risk of recurrence had worse OS than those with lower risk, for both procedures (5-year OS [95% CI], low risk versus high risk: lobectomy, versus 91% [89–94] versus 79% [73–86], p < 0.001; and sublobar resection, 81% [76–87] versus 58% [50–68], p < 0.001). Similarly in the external-validation cohort, patients with a high risk of recurrence had worse OS than those with lower risk, for both procedures (5-year OS [95% CI], low risk versus high risk: lobectomy, 90% [87– 94] versus 88% [82–94], p ¼ 0.008; and sublobar resection, 90% [84–96] versus 62% [46–73], p < 0.001).

Lobectomy Risk Prediction Score for Predicting Benefit From ACT Using Stage II-III Propensity Score-Matched Cohort The 1:1 matching for ACT versus no-ACT resulted in 158 matched pairs (n ¼ 316), with balanced covariates between ACT groups (Table 3). Analysis of lung cancer– specific CID curves revealed a significant survival benefit for ACT only in the high score group (p ¼ 0.003) (Fig. 3A). A similar pattern was observed for recurrence and OS (Figs. 3B and C). In a multivariable competing

Lobectomy

Sublobar resection

Univariable Analysis

Final Multivariable Model

SHR

95% CI

p

Age >65 years ref. 65 years Age (per 1 year increase) Male sex ref. Female Smoker ref. Never p-stage IB ref. IA Tumor size >1 cm ref. 1 cm Tumor size (per 1-cm increase) Invasive tumor size >1 cm ref. 1 cm Invasive tumor size (per 1-cm increase) Lymphovascular invasion ref. Absent Pleural invasion ref. Absent Necrosis present ref. Absent Histologic grade High Intermediate ref. Low MIP percentage (linear, per 10% increase) MIP percentage (nonlinear, 10% vs. 0%)a SOL percentage (nonlinear, 15% vs. 0%)a STAS present ref. Absent

10% 12% N/A 11% 10% 12% 4% 26% 9% 11% 8% N/A

0.78

(0.46 to 1.32)

0.4

0.99 1.03

(0.97 to 1.02) (0.60 to 1.76)

0.6 0.9

3.75

(1.17 to 11.97)

0.026

2.90

(1.56 to 5.41)

0.001

1.16

(0.55 to 2.45)

0.7

1.42

(0.69 to 2.92)

0.3

2.13

(1.13 to 4.04)

0.020

2.39

(1.31 to 4.38)

0.005

20% 5% 26% 9% 24% 7%

4.27

(2.40 to 7.59)

<0.001

2.90

(1.56 to 5.41)

0.001

3.49

(2.06 to 5.93)

<0.001

21% 9% 4% N/A

6.37 2.36

(1.94 to 20.88) (0.73 to 7.65)

0.002 0.15

1.32

(1.13 to 1.53)

<0.001

13% 6% N/A

N/A

2.89

(2.04 to 4.09)

0.002

16% 8%

2.14

(1.27 to 3.60)

0.004

SHR

2.14

1.75

1.21

2.12

95% CI

(1.11 to 4.14)

(1.00 to 3.04)

(1.04 to 1.43)

(1.39 to 3.22)

p

0.024

0.050

5-yr CIR

SHR

95% CI

p

21% 22% N/A 25% 19% 22% 17% 37% 18% 23% 17% N/A

0.98

(0.61 to 1.57)

0.9

1.00 1.40

(0.98 to 1.03) (0.88 to 2.22)

0.7 0.15

1.26

(0.64 to 2.49)

0.5

2.38

(1.39 to 4.05)

0.001

1.40

(0.83 to 2.37)

0.2

1.42

(0.85 to 2.36)

0.177

29% 15% N/A

2.16

(1.34 to 3.47)

0.001

2.87

(1.90 to 4.34)

<0.001

38% 12% 37% 18% 35% 18%

3.86

(2.40 to 6.21)

<0.001

2.38

(1.39 to 4.05)

0.001

2.59

(1.59 to 4.20)

<0.001

25% 27% 4%

6.32 6.04

(2.16 to 18.45) (2.20 to 16.61)

0.001 <0.001

N/A

3.10

(2.17 to 4.42)

0.001

N/A

1.63

(1.23 to 2.16)

<0.001

42% 10%

5.22

(3.14 to 8.68)

<0.001

95% CI

p

2.02

(1.16 to 3.51)

0.013

1.72

(1.06 to 2.82)

0.031

1.77

(1.17 to 2.67)

0.007

2.48

(1.26 to 4.89)

0.009

0.021

<0.001

-

The two values are the third and first quartiles, respectively, of the variable distribution. CI, confidence interval; CIR, cumulative incidence of recurrence; N/A, not applicable; MIP, micropapillary; SHR, subhazard ratio; SOL, solid; STAS, spread through air spaces.

a

SHR

Vol.

5-yr CIR

Final Multivariable Model

Journal of Thoracic Oncology

Variables

Univariable Analysis

8 Bains et al

Table 2. Results of the Univariable and Multivariable Competing Risks Analyses for any Recurrence, by Lobectomy and Sublobar Resection

No. -

---

2018

Risk Prediction for Recurrence

9

Figure 1. Risk prediction for patients who underwent lobectomy (A) or sublobar resection (B)These algorithms present a method to calculate the 3-year or 5-year cumulative incidence of recurrence (CIR) after resection on the basis of a patient’s combination of characteristics. To calculate the probability of recurrence, locate the patient’s micropapillary level (%) and draw a straight line up to the “Scores” axis to derive the score associated with the level. Repeat for the other three covariates (solid, lymphovascular invasion [LVI], and necrosis). Add the scores for all covariates and locate the sum on the “Total Scores” axis. Draw a vertical line down from “Total Scores” to the last two axes to obtain the corresponding 3-year and 5-year CIR following lobectomy. STAS, spread through air spaces.

risks regression analysis for lung cancer–specific death, a significant interaction was observed between high/low risk score and ACT (p ¼ 0.026) after adjustment for stage and STAS. Specifically, for high-risk patients, ACT significantly reduced the risk of lung cancer–specific death (subhazard ratio: 0.43, 95% CI: 0.26–0.72, p ¼ 0.001) (Supplementary Table 5). No significant benefit of ACT was observed among low-risk patients. In the propensity score–matched stage II-III cohort, 14% of patients did not undergo lobectomy (included were pneumonectomy, bilobectomy, and segmentectomy). Survival analyses for only patients who underwent lobectomy (N ¼ 272) are provided in Supplementary Figure 7. Similar to the results for all matched patients

(Fig. 3), a significant survival benefit for ACT was observed only in the high-score group.

Discussion We developed a clinical tool to predict recurrence in patients with resected small lung ADC. Our risk prediction algorithms are distinct from previously designed algorithms for the following reasons: (1) we performed a multivariable analysis including recently recognized prognostic markers such as MIP, SOL, and STAS; (2) we created surgical procedure–specific risk prediction for lobectomy and sublobar resection; (3) we selected recurrence as an outcome and performed competing risks analyses; (4) we performed external validation

10 Bains et al

Journal of Thoracic Oncology

Vol.

-

No.

-

Figure 2. Internal and external calibration for 5-year cumulative incidence of recurrence (CIR) derived from the risk prediction for patients who underwent lobectomy (A,B) or sublobar resection (C,D). The calibration of the risk score for lobectomy (A, development cohort; B, validation cohort) and sublobar resection (B, development cohort; D, validation cohort) is shown. The horizontal axis is the algorithm prediction of recurrence at 5 years. The vertical axis is the corresponding observed 5-year CIR, based on the cumulative incidence function with the competing risks approach. The dashed line is the reference line on which an ideal algorithm would lie. The solid line indicates the performance of the current risk prediction (dots represent the average predicted 5-year CIR). X’s indicate the optimism-corrected estimate of the risk predictive algorithm performance based on 500 bootstrap resamples. The vertical bars represent 95% confidence intervals. MSK, Memorial Sloan Kettering Cancer Center.

using a data set comprised of three separate international cohorts, confirming the applicability of the scores to patients with varying ethnic backgrounds, histologic subtypes, and clinical presentations; and (5) we applied the lobectomy risk prediction algorithm to a propensity score–matched cohort of patients — in which all clinical, radiological, surgical, and pathologic variables known to contribute to patient lung cancer and noncancer survival were matched — and showed a survival benefit for platinum-based ACT only in high-risk patients (according to lobectomy risk prediction score), and not

in low-risk patients, which suggests our lobectomy risk prediction score has the potential to predict benefit from ACT. The use of cohorts with information available on prognostic histologic variables enabled a comprehensive analysis and facilitated the development of scores with high accuracy. The C-index was higher than those for previously designed lung cancer risk scores (Supplementary Table 4). One reason for these high C-indices is that we included several recently validated — and IASLC- and WHO-recommended — histopathologic

---

2018

Risk Prediction for Recurrence

11

Table 3. Clinicopathologic Demographic Characteristics in the Stage II-III Cohort and Difference Between Adjuvant Chemotherapy Versus no Adjuvant Chemotherapy Before/After Propensity-Score Matching Before Matching Variables Background Age, years Sex female male Smoking never former current COPD CVD Other comorbidityb Cr (mg/dL) FEV1 (%) DLCO (%) Radiology SUVmax Surgery Resection type Pneumonectomy Bilobectomy Lobectomy Segmentectomy Approach MIS Thoracotomyc Pathology Tumor size (cm) pN 0 1 2 p-Stage II III LVI Pleural invasion 0 1 or 2 3 Necrosis STAS Histologic graded Low Intermediate High

After Matching

Adjuvant (-) n ¼ 312

Adjuvant (þ) n ¼ 277

ASMDa

Adjuvant (-) n ¼ 158

Adjuvant (þ) n ¼ 158

ASMDa

71 (63, 78)

66 (59, 72)

0.527

68 (62, 74)

68 (63, 74)

0.046

188 (60) 124 (40)

181 (65) 96 (35)

0.105

99 (63) 59 (37)

103 (65) 55 (35)

0.053

49 (16) 222 (71) 41 (13) 54 (17) 58 (19) 97 (31) 1$0 (0.9, 1.2) 87 (74, 99) 76 (63, 90)

51 (18) 193 (70) 33 (12) 46 (17) 33 (12) 81 (29) 1.0 (0.8, 1.1) 88 (72, 101) 81 (69, 95)

0.076

24 (15) 113 (72) 21 (13) 24 (15) 26 (16) 49 (31) 1.0 (0.9, 1.2) 85 (70, 99) 78 (67, 90)

0.024

0.019 0.187 0.040 0.177 0.031 0.238

25 (16) 113 (72) 20 (13) 25 (16) 23 (15) 43 (27) 1.0 (0.8, 1.1) 87 (73, 100) 80 (64, 95)

0.017 0.052 0.084 0.089 0.079 0.020

6.8 (3.7, 11.0)

7.0 (3.5, 9.9)

0.084

6.2 (3.3, 10.8)

5.8 (3.0, 10.2)

0.002

15 (5) 8 (3) 271 (87) 18 (6)

11 (4) 6 (2) 245 (88) 15 (5)

0.053

9 (6) 3 (2) 136 (86) 10 (6)

9 (6) 4 (3) 136 (86) 9 (6)

0.050

57 (18) 260 (83)

86 (31) 191 (69)

0.342

33 (21) 125 (79)

29 (18) 129 (82)

0.064

4.1 (2.5, 5.0)

2.8 (2.0, 4.1)

0.433

2.7 (2.0, 4.5)

2.8 (1.9, 4.5)

0.044

131 (42) 108 (35) 73 (23)

37 (13) 119 (43) 121 (44)

0.700

35 (22) 68 (43) 55 (35)

33 (21) 71 (45) 54 (34)

0.041

196 (63) 116 (37) 182 (58)

139 (50) 138 (50) 220 (79)

0.257

88 (56) 70 (44) 108 (68)

91 (58) 67 (42) 112 (71)

0.038

230 (74) 73 (23) 9 (3) 112 (36) 144 (46)

180 (65) 91 (33) 6 (2) 55 (20) 183 (72)

112 (71) 41 (26) 5 (3) 45 (28) 104 (66)

112 (71) 43 (27) 3 (2) 42 (27) 108 (68)

14 (4) 153 (49) 145 (46)

6 (2) 147 (53) 124 (45)

4 (3) 81 (51) 73 (46)

5 (3) 79 (50) 74 (47)

0.468 0.213

0.377 0.412 0.142

0.055 0.083

0.043 0.054 0.043

Data are shown as number (%) or median (25th, 75th percentile). a ASMD (absolute standardized mean difference) < 0.1 indicates balance in the covariate between the two groups. Bold represents ASMD  0.1. b Comorbidities other than COPD or CVD included in Charlson comorbidity index evaluation. c Included conversion from MIS to thoracotomy. d Histologic grade based on predominant subtypes (low grade, lepidic; intermediate, acinar or papillary; high, micropapillary, solid, invasive mucinous, or colloid). ASMD, absolute standardized mean difference; COPD, chronic obstructive pulmonary disease; Cr, serum creatinine; CVD, cardiovascular disease; DLCO, diffusion capacity of the lung for carbon monoxide; FEV1, forced expiratory volume in 1 second; LVI, lymphovascular invasion; MIS, minimally invasive surgery; pN, pathological lymph node metastasis status; PS, propensity score; p-Stage, pathologic stage; STAS, spread through air spaces; SUVmax, maximum standardized uptake value.

12 Bains et al

Journal of Thoracic Oncology

Vol.

-

No.

-

Figure 3. Lung cancer–specific cumulative incidence of death (CID) (A), cumulative incidence of recurrence (CIR) (B), and overall survival (C) in the propensity score–matched stage II-III cohort. A comparison between no adjuvant chemotherapy (ACT) and ACT by lobectomy algorithm-derived risk score. Lung cancer–specific CID curves in four groups are shown: the black solid line represents low algorithm-derived risk score without ACT; black dashed line, low score with ACT; red solid line, high score without ACT; and red dashed line, high score with ACT. In the comparison between high score/ACT and high score/no ACT (red dashed vs. solid lines), lung cancer–specific CID is significantly lower for ACT than for no ACT (p ¼ 0.003), whereas in the comparison between low score/ACT and low score/no ACT (black dashed vs. solid lines), no significant difference was observed between the two groups. High- or low-score categories were determined by whether the score was above or below the median. A similar pattern was observed for CIR and overall survival (OS): in high score, there was lower CIR and higher OS for ACT than for no ACT (p ¼ 0.005 and p ¼ 0.002, respectively), whereas in low score, there was no difference in CIR and OS between ACT and no ACT.

---

2018

factors that can influence outcomes following lobectomy or sublobar resection; other studies have included only tumor size as a prognostic variable. Prognostic stratification based on the new lung ADC classification has been reported to identify high-risk patients with SOL- or MIP-predominant tumors.22,23,28 In addition to the associated worse outcomes following surgery, Tsao et al.43 recently showed that high-grade (MIP and SOL) predominant subtypes can help predict patients’ benefit from platinum-based ACT. On the basis of our and others’ data, we focused on the continuous percentage of the high-grade subtypes (MIP and SOL) and the presence of STAS, LVI, and necrosis to develop scores to predict risk of recurrence.21-23,26,27,29,30 The subsequent analysis revealed that the proposed lobectomy risk score has the potential to predict benefit from ACT. As recurrence is strongly correlated with lung cancer–specific death (Supplementary Fig. 8 and Table 6), we used recurrence as our primary outcome. In early-stage disease, particularly among elderly patients, OS and recurrence-free survival are affected by competing events, which underlines the need to perform competing risks analysis.44,45 In this study, death without recurrence was treated as a competing event. Competing risks analysis addressed the occurrence of noncancer-related events when predicting cancerspecific outcomes to improve predictions. One limitation of our study is that the external cohorts had relatively fewer patients than the development cohort. Nevertheless, the inclusion of cohorts from Eastern and Western continents (with inherent variability in demographic factors and histologic subtypes), the proportions of patients who underwent lobectomy and sublobar resection, and the use of multiple pathologists to confirm pathologic interpretations add strength to our approach. There is significant variation in the prevalence of certain pathologic features — such as LVI, pleural invasion, and the presence of MIP — in different cohorts. This could be due to inherent differences in ethnic backgrounds but could also be secondary to interobserver variation among pathologists. Although all known histologic variables were used in the risk prediction development, due to lack of availability of imaging and molecular tumor analysis from patients who underwent resection in earlier period, we did not include them in our analysis. These algorithms do not provide assistance in selecting patients preoperatively for lobectomy versus sublobar resection. Future studies that improve the diagnostic accuracy of histologic subtypes during intraoperative frozen section may extend the use of our risk prediction scores intraoperatively. Another limitation is that we did not include pathologic lymph node evaluation in this analysis, which might have affected outcomes (Supplementary Methods and Table 7).

Risk Prediction for Recurrence

13

Both the frequency of ACT use and the reasons for not administering ACT for patients with stage II-III disease differed by year of surgery, especially before and after 2004 (Supplementary Table 8). Because this change in treatment strategy might have affected our results, we performed the same prognostic analysis in patients who underwent surgery between 2004 and 2013 (n ¼ 215; Supplementary Fig. 9): the difference in survival between the ACT and the no-ACT groups was similar to that of the entire cohort. In conclusion, we have developed and validated the two procedure-specific risk prediction algorithms. Given that 80% of patients with stage IA lung ADC survive for at least 5 years following surgical resection, it is imperative that postoperative interventions intended to improve outcomes identify patients at high risk of recurrence and that optimal treatments are investigated.1 Despite the success of curative-intent surgery, high rates of recurrence are seen in a subgroup of early-stage patients, particularly those with MIP, SOL, STAS, LVI, and necrosis. The two procedure-specific risk prediction algorithms can be used postoperatively to provide patients with more accurate prognostic information that is based on their individual clinicopathologic status and can also assist in stratifying patients for further interventions. In particular, the lobectomy risk prediction algorithm — with its demonstrated ability to predict benefit from platinum-based ACT in stage II-III ADC — has the potential to be used for patient selection in future trials of ACT for stage I lung ADC.

Acknowledgments The authors’ laboratory work is supported by grants from the National Institutes of Health (R01 CA217169, R01 CA236615, and P30 CA008748), the U.S. Department of Defense (CA170630, BC132124 and LC160212), the Joanne and John DallePezze Foundation, the Derfner Foundation, and the Mr. William H. Goodwin and Alice Goodwin, the Commonwealth Foundation for Cancer Research, and the Experimental Therapeutics Center of Memorial Sloan Kettering Cancer Center. The authors thank David B. Sewell and Alex Torres of the Memorial Sloan Kettering Thoracic Surgery Service for their editorial assistance.

Supplementary Data Note: To access the supplementary material accompanying this article, visit the online version of the Journal of Thoracic Oncology at www.jto.org and at https://doi. org/10.1016/j.jtho.2018.09.008.

References 1. Rami-Porta R, Bolejack V, Crowley J, et al. The IASLC Lung Cancer Staging Project: proposals for the revisions of the T descriptors in the forthcoming eighth edition of

14 Bains et al

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

the TNM classification for lung cancer. J Thorac Oncol. 2015;10:990–1003. Aberle DR, Adams AM, Berg CD, et al. Reduced lungcancer mortality with low-dose computed tomographic screening. N Engl J Med. 2011;365:395–409. Paul S, Altorki NK, Sheng S, et al. Thoracoscopic lobectomy is associated with lower morbidity than open lobectomy: a propensity-matched analysis from the STS database. J Thorac Cardiovasc Surg. 2010;139:366–378. Koike T, Yamato Y, Yoshiya K, et al. Intentional limited pulmonary resection for peripheral T1 N0 M0 small-sized lung cancer. J Thorac Cardiovasc Surg. 2003;125:924–928. Wolf AS, Richards WG, Jaklitsch MT, et al. Lobectomy versus sublobar resection for small (2 cm or less) nonsmall cell lung cancers. Ann Thorac Surg.y. 2011;92:1819–1823. discussion 1824-1815. Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus limited resection for T1 N0 non–small cell lung cancer. Lung Cancer Study Group. Ann Thorac Surg. 1995;60:615–622; discussion 622-613. Varlotto JM, Medford-Davis LN, Recht A, et al. Identification of stage I non–small cell lung cancer patients at high risk for local recurrence following sublobar resection. Chest. 2013;143:1365–1377. Eguchi T, Bains S, Lee MC, et al. Impact of increasing age on cause-specific mortality and morbidity in patients with stage I non-small-cell lung cancer: a competing risks analysis. J Clin Oncol. 2017;35:281–290. Wu J, Zhou L, Huang L, et al. Nomogram integrating gene expression signatures with clinicopathological features to predict survival in operable NSCLC: a pooled analysis of 2164 patients. J Exper Clin Cancer Res. 2017;36:4. Liang W, Zhang L, Jiang G, et al. Development and validation of a nomogram for predicting survival in patients with resected non-small-cell lung cancer. J Clin Oncol. 2015;33:861–869. Yang HC, Kim HR, Jheon S, et al. Recurrence risk-scoring model for stage I adenocarcinoma of the lung. Ann Surg Oncol. 2015;22:4089–4097. Song SH, Ahn JH, Lee HY, et al. Prognostic impact of nomogram based on whole tumour size, tumour disappearance ratio on CT and SUVmax on PET in lung adenocarcinoma. Eur Radiol. 2016;26:1538–1546. Wang Y, Wang R, Zheng D, et al. Predicting the recurrence risk factors and clinical outcomes of peripheral pulmonary adenocarcinoma 3 cm with wedge resection. J Cancer Res Clin Oncol. 2017;143:1043–1051. Huang Y, Liu Z, He L, et al. Radiomics signature: a potential biomarker for the prediction of disease-free survival in early-stage (I or II) non–small cell lung cancer. Radiology. 2016;281:947–957. Kent MS, Mandrekar SJ, Landreneau R, et al. A nomogram to predict recurrence and survival of highrisk patients undergoing sublobar resection for lung cancer: an analysis of a multicenter prospective study (ACOSOG Z4032). Ann Thorac Surg. 2016;102:239–246. Zhang J, Gold KA, Lin HY, et al. Relationship between tumor size and survival in non–small-cell lung cancer (NSCLC): an analysis of the surveillance, epidemiology, and end results (SEER) registry. J Thorac Oncol. 2015;10:682–690.

Journal of Thoracic Oncology

Vol.

-

No.

-

17. Lee HY, Lee SW, Lee KS, et al. Role of CT and PET imaging in predicting tumor recurrence and survival in patients with lung adenocarcinoma: a comparison with the International Association for the Study of Lung Cancer/ American Thoracic Society/European Respiratory Society classification of lung adenocarcinoma. J Thorac Oncol. 2015;10:1785–1794. 18. Zhang Y, Sun Y, Xiang J, et al. A clinicopathologic prediction model for postoperative recurrence in stage Ia non–small cell lung cancer. J Thorac Cardiovasc Surg. 2014;148:1193–1199. 19. Travis WD, Brambilla E, Noguchi M, et al. International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society international multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol. 2011;6:244–285. 20. Young KA, Efiong E, Dove JT, et al. External validation of a survival nomogram for non–small cell lung cancer using the national cancer database. Ann Surg Oncol. 2017;24: 1459–1464. 21. Yoshizawa A, Motoi N, Riely GJ, et al. Impact of proposed IASLC/ATS/ERS classification of lung adenocarcinoma: prognostic subgroups and implications for further revision of staging based on analysis of 514 stage I cases. Mod Pathol. 2011;24:653–664. 22. Warth A, Muley T, Meister M, et al. The novel histologic International Association for the Study of Lung Cancer/ American Thoracic Society/European Respiratory Society classification system of lung adenocarcinoma is a stageindependent predictor of survival. J Clin Oncol. 2012;30:1438–1446. 23. Hung JJ, Yeh YC, Jeng WJ, et al. Predictive value of the International Association for the Study of Lung Cancer/ American Thoracic Society/European Respiratory Society classification of lung adenocarcinoma in tumor recurrence and patient survival. J Clin Oncol. 2014;32:2357–2364. 24. Travis WD, Brambilla E, Nicholson AG, et al. The 2015 World Health Organization classification of lung tumors: impact of genetic, clinical and radiologic advances since the 2004 classification. J Thorac Oncol. 2015;10: 1243–1260. 25. Miyoshi T, Satoh Y, Okumura S, et al. Early-stage lung adenocarcinomas with a micropapillary pattern, a distinct pathologic marker for a significantly poor prognosis. Am J Surg Pathol. 2003;27:101–109. 26. Nitadori J, Bograd AJ, Kadota K, et al. Impact of micropapillary histologic subtype in selecting limited resection vs lobectomy for lung adenocarcinoma of 2cm or smaller. J Natl Cancer Inst. 2013;105:1212–1220. 27. Cha MJ, Lee HY, Lee KS, et al. Micropapillary and solid subtypes of invasive lung adenocarcinoma: clinical predictors of histopathology and outcome. J Thorac Cardiovasc Surg. 2014;147:921–928 e922. 28. Ujiie H, Kadota K, Chaft JE, et al. Solid predominant histologic subtype in resected stage I lung adenocarcinoma is an independent predictor of early, extrathoracic, multisite recurrence and of poor postrecurrence survival. J Clin Oncol. 2015;33:2877–2884. 29. Kadota K, Nitadori J, Sima CS, et al. Tumor spread through air spaces is an important pattern of invasion

---

30.

31.

32.

33.

34.

35.

36.

2018

and impacts the frequency and location of recurrences after limited resection for small stage I lung adenocarcinomas. J Thorac Oncol. 2015;10:806–814. Warth A, Muley T, Kossakowski CA, et al. Prognostic impact of intra-alveolar tumor spread in pulmonary adenocarcinoma. Am J Surg Pathol. 2015;39:793–801. Morales-Oyarvide V, Mino-Kenudson M. Tumor islands and spread through air spaces: distinct patterns of invasion in lung adenocarcinoma. Pathol Int. 2016;66:1–7. Iasonos A, Schrag D, Raj GV, et al. How to build and interpret a nomogram for cancer prognosis. J Clin Oncol. 2008;26:1364–1370. National Comprehensive Cancer Centers: NCCN clinical practice guidelines in oncology (NCCN Guidelines): Non–small cell lung cancer v4. 2016. http://www.nccn. org/professionals/physician_gls/pdf/nscl.pdf. Accessed September 3, 2018. Pisters KM, Evans WK, Azzoli CG, et al. Cancer Care Ontario and American Society of Clinical Oncology adjuvant chemotherapy and adjuvant radiation therapy for stages I-IIIA resectable non small-cell lung cancer guideline. J Clin Oncol. 2007;25:5506–5518. Kris MG, Gaspar LE, Chaft JE, et al. Adjuvant systemic therapy and adjuvant radiation therapy for stage I to IIIA completely resected non–small-cell lung cancers: American Society of Clinical Oncology/Cancer Care Ontario clinical practice guideline update. J Clin Oncol. 2017. Jco2017724401. Amin MBCancer. AJCo. AJCC Cancer Staging manual. New York, New York: Springer; 2016.

Risk Prediction for Recurrence

15

37. Dignam JJ, Zhang Q, Kocherginsky M. The use and interpretation of competing risks regression models. Clin Cancer Res. 2012;18:2301–2308. 38. Durrleman S, Simon R. Flexible regression models with cubic splines. Stat Med. 1989;8:551–561. 39. Rosenbaum PR, Rubin DB. Constructing a control group using multivariate matched sampling methods that incorporate the propensity score. Am Stat. 1985;39:33–38. 40. Austin PC. An Introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46:399–424. 41. Austin PC. Optimal caliper widths for propensity-score matching when estimating differences in means and differences in proportions in observational studies. Pharm Stat. 2011;10:150–161. 42. Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med. 2009;28:3083–3107. 43. Tsao MS, Marguet S, Le Teuff G, et al. Subtype classification of lung adenocarcinoma predicts benefit from adjuvant chemotherapy in patients undergoing complete resection. J Clin Oncol. 2015;33:3439–3446. 44. Carmona R, Zakeri K, Green G, et al. Improved method to stratify elderly patients with cancer at risk for competing events. J Clin Oncol. 2016;34:1270–1277. 45. Tan KS, Eguchi T, Adusumilli PS. Competing risks and cancer-specific mortality: why it matters. Oncotarget. 2018;9:7272–7273.