Genome-wide copy number analyses of samples from LACE-Bio project identify novel prognostic and predictive markers in early stage non-small cell lung cancer

Abstract

Background

Adjuvant chemotherapy (ACT) provides modest benefit in resected non-small cell lung cancer (NSCLC) patients. Genome-wide studies have identified gene copy number aberrations (CNA), but their prognostic implication is unknown.

Methods

DNA from 1,013 FFPE tumor samples from three pivotal multicenter randomized trials (ACT vs. control) in the LACE-Bio consortium (median follow-up: 5.2 years) was successfully extracted, profiled using a molecular inversion probe SNP assay, normalized relative to a pool of normal tissues and segmented. Minimally recurrent regions were identified. P values were adjusted to control the false discovery rate (Q values).

Results

A total of 976 samples successfully profiled, 414 (42%) adenocarcinoma (ADC), 430 (44%) squamous cell carcinoma (SCC) and 132 (14%) other NSCLC; 710 (73%) males. We identified 431 recurrent regions, with on average 51 gains and 43 losses; 253 regions (59%) were ≤3 Mb. Most frequent gains (up to 48%) were on chr1, 3q, 5p, 6p, 8q, 22q; most frequent losses (up to 40%) on chr3p, 8p, 9p. CNA frequency of 195 regions was significantly different (Q≤0.05) between ADC and SCC. Fourteen regions (7p11–12, 9p21, 18q12, and 19p11–13) were associated with disease-free survival (DFS) (univariate P≤0.005, Q<0.142), with poorer DFS for losses of regions including CDKN2A/B [hazard ratio (HR) for 2-fold lower CN: 1.5 (95% CI: 1.2–1.9), P<0.001, Q=0.020] and STK11 [HR =2.4 (1.3–4.3), P=0.005, Q=0.15]. Chromosomal instability was associated with poorer DFS (HR =1.5, P=0.015), OS (HR =1.2, P=0.189) and lung-cancer specific survival (HR =1.7, P=0.003).

Conclusions

These large-scale genome-wide analyses of gene CNA provide new candidate prognostic markers for stage I–III NSCLC.

Introduction

Lung cancer is the leading cause of cancer death worldwide. Non-small cell lung cancer (NSCLC), accounting for 85% of all lung cancers, has a 5-year survival of 59% for early resectable disease, but only 15% for cancers in advances stages (1). However, great differences within individual stages suggest the existence of unknown tumor factors. In the era of personalized medicine, the assessment of prognostic factors is crucial for individual treatment decision making. The activation of oncogenes (i.e., EGFR and KRAS) and the inhibition of tumor-suppressors (TP53) drive tumor progression. While targeting some of these genes is a promising therapeutic strategy in adenocarcinoma (ADC), most lung cancers lack proven (targetable) driver genes and identification of additional ones is critical. Recent developments of genome-wide profiling have identified new genes, but the studies reported to date are underpowered or lack a control arm. Bass et al. (2) profiled 40 esophageal squamous cell carcinomas (SCC) (29 primary and 11 cell lines) and 47 primary lung SCC for DNA copy number (CN) change. They reported that SOX2 (chr.3q26.33) was significantly amplified and that it was a lineage-survival oncogene by knockdown experiments in cell lines. However, the small sample size hindered assessment of the prognostic value of CN aberrations (CNA). The Cancer Genome Atlas (TCGA) recruited 10,000 samples from 33 cancer types and profiled alterations from genomic DNA, RNA, and protein. However, due to the inclusion criteria (≥70% tumor cellularity), advanced stages were underrepresented. Furthermore, the samples used in these studies were snap-frozen tissues whereas most of the samples in clinical settings are formalin-fixed and paraffin-embedded (FFPE). Thus, identifying prognostic markers from FFPE samples may be clinically relevant.

The Lung Adjuvant Cisplatin Evaluation (LACE-Bio) project comprises FFPE samples from four LACE adjuvant chemotherapy (ACT) trials and evaluated the prognostic and predictive role of biomarkers including ERCC1 (3), tumor infiltrating lymphocytes (TILs) (4), mucin (5), beta-tubulin (6), KRAS (7), EGFR (8) and TP53 (9). Importantly, 1,013 samples from three trials were profiled for their DNA CNAs. Since the trials were randomized and controlled, the data were fit for evaluating markers associated with the magnitude of ACT benefit.

Methods

Patients and samples

The LACE-Bio2 consortium includes patients from four pivotal trials comparing platinum-based ACT to observation after complete resection of stage I–III NSCLC (10-15). Of these, 1,013 patients from three trials had FFPE samples available, whereas samples in one trial (15) were exhausted. All individual trials including tissue collection for future research were approved by institutional review boards at each participating site.

DNA isolation and profiling

DNA was successfully extracted from 976 FFPE samples using the AllPrep DNA/RNA FFPE Kit (Quagen, Germantown, MD, USA), and profiled using the OncoScan CNV Plus Assay (ThermoFisher, Carlsbad, California, USA), a molecular inversion probe SNP assay (16). The platform algorithm delivered the median of the absolute values of all pairwise differences (MAPD) (17,18) as quality metrics; 777 samples with MAPD ≤0.3 were classified as optimal quality.

Statistical analyses

The data were normalized relative to a pool of reference normal samples and segmented using circular binary segmentation (19,20). Minimal recurrent regions were identified via the CGHregions algorithm (21). Tumor clonal composition number was estimated by using the OncoClone composition program (22). The primary endpoint was disease-free survival (DFS). Secondary endpoints were overall survival (OS) and lung-cancer specific survival (LCSS). CNAs were correlated to endpoints using Cox models stratified by trial and adjusted for treatment and clinicopathological factors. The regression models estimated the hazard ratio HR_gain for a 2-fold higher CN, with HR_loss = 1/HR_gain the relative hazard for a 2-fold lower CN. The predictive role of CNAs was estimated by further adding a treatment-by-CN interaction to the models. We performed univariate (by region) and two multivariate analyses (stepwise selection and penalized regression) (23,24). Q values were used to correct P values for multiple comparisons (25).

Preplanned sensitivity analyses included: histologic subgroups (ADC vs. SCC), optimal quality subgroup. CN differences between histologies were assessed by t-tests, with P values corrected via step-down multiple testing procedures (26,27). We compared results to those from our reanalysis of the TCGA (28,29) using exactly the same method. Known tumor suppressors and oncogenes were obtained from literature (30).

The association of the number of breakpoints (BPs), quantifying chromosomal instability, with clinicopathological factors was tested in univariate analyses, then in multivariate log-linear models. The association of chromosomal instability and of clonality with outcomes and treatment effect was studied in Cox models.

Full details of statistical methods are provided in the supplementary material.

Results

Three samples (Figure S1) were partially processed; 1 failed linkage to the clinical database; the inferred gender of 32 patients was incorrect; 1 sample was duplicate. In total, 976 samples were analyzed: 414 (42%) ADC, 430 (44%) SCC, 132 (14%) other NSCLC; 485 were in the control and 491 in the ACT groups (Table 1).

Figure S1.

Flowchart. FFPE, formalin fixation and paraffin embedding; CALBG, Cancer and Leukemia Group B trial 9633 (8); IALT, International Adjuvant Lung Trial (4,5); JBR.10, National Cancer Institute of Canada intergroup (6,7); CAN, copy number aberration.

Table 1. Demographic characteristics of patients with OncoScan analysis results.

Characteristics	Control group (N=485) (No., %)	Chemotherapy group (N=491) (No., %)	Total (N=976) (No., %)
Trial
CALGB	66 [14]	58 [12]	124 [13]
IALT	258 [53]	266 [54]	524 [54]
JBR 10	161 [33]	167 [34]	328 [34]
Age
≤55	137 [28]	137 [28]	274 [28]
55–64	202 [42]	205 [42]	407 [42]
≥65	146 [30]	149 [30]	295 [30]
Sex
Female	139 [29]	127 [26]	266 [27]
Male	346 [71]	364 [74]	710 [73]
PS
0	248 [51]	252 [51]	500 [51]
1–2	235 [49]	238 [49]	473 [49]
Histology
Adenocarcinoma	207 [43]	207 [42]	414 [42]
Squamous cell carcinoma	218 [45]	212 [43]	430 [44]
Other	60 [12]	72 [15]	132 [14]
T
T1	57 [12]	64 [13]	121 [12]
T2	372 [77]	365 [75]	737 [76]
T3/T4	54 [11]	60 [12]	114 [12]
N
N0	250 [52]	251 [51]	501 [52]
N1	167 [35]	175 [36]	342 [35]
N2	66 [14]	63 [13]	129 [13]
Surgery
Lobectomy/other	344 [71]	333 [68]	677 [69]
Pneumonectomy	141 [29]	157 [32]	298 [31]

The 217,611 array probes were grouped into 431 common-CN regions; 253 regions (59%) were ≤3 Mb, 340 (79%) were ≤10 Mb (Figure 1); 166 regions had a loss (177 a gain) in ≥10% patients. On average, patients had 94 CNAs (standard deviation 69), 51 gains and 43 losses.

Figure 1.

The landscape of copy number aberrations in all 976 LACE-Bio patients available for OncoScan assay analysis.

The most frequent CN gains (Table S1) were in 1q21–23, 3q22–26, 5p13–15, 6p24, 8q21–24, 22q11, containing genes TERT, PIK3CA, MECOM, CCNL1 among others. The most frequent CN losses were in chromosomes 3p21.31, 8p23, and 9p21.3, containing CDKN2A/B. These results remained consistent in the optimal quality samples subset (N=777; Figure S2, Table S2).

Table S1. Most frequent copy number aberrations in all the samples (N=976).

Chr	Region ID	Loss Freq	Gain Freq	Start	End	Mb	Genes	cytoBands
1	6	15.8%	44.7%	110 231 909	110 240 929	9.0E−3		p13.3
	13	0.5%	31.8%	145 394 955	148 544 968	3.0E+0	BCL9, TXNIP	q21.1–2
	15	0.5%	32.8%	149 742 045	161 515 326	1.0E+1	MLLT11, NTRK1, PRCC, TPM3, PYHIN1, EFNA1, MUC1, PLEKHO1, AIM2	q21.2–q23.3
	16	3.8%	30.6%	161 591 477	161 607 441	2.0E−2		q23.3
	17	1.7%	31.7%	161 609 660	161 622 701	1.0E−2		q23.3
3	48	30.0%	0.2%	46 804 388	46 831 840	3.0E−2		p21.31
	50	31.0%	1.6%	75 444 906	75 554 646	1.0E−1		p12.3
	64	3.5%	31.2%	134 402 484	143 774 592	9.0E+0	XRN1	q22.2–q24
	65	3.0%	36.9%	143 794 370	151 504 070	8.0E+0	WWTR1	q24, q25.1
	66	3.6%	38.2%	151 520 944	151 546 041	3.0E−2		q25.1
	67	2.8%	39.8%	151 563 527	162 500 864	1.0E+1	CCNL1, GMPS	q25.1–q26.1
	68	12.7%	40.0%	162 540 700	162 602 984	6.0E−2		q26.1
	69	2.6%	41.5%	162 640 497	162 702 814	6.0E−2		q26.1
	70	2.2%	41.8%	162 719 684	165 245 914	3.0E+0		q26.1
	71	2.3%	42.6%	165 270 444	165 296 562	3.0E−2		q26.1
	72	1.8%	44.8%	165 314 375	169 905 944	5.0E+0	MECOM	q26.1–2
	73	2.2%	45.4%	169 918 311	175 861 931	6.0E+0	PRKCI, PRKCI	q26.2–32
	74	2.6%	45.1%	175 889 230	175 905 626	2.0E−2		q26.32
	75	2.4%	45.5%	175 920 884	187 866 388	1.0E+1	PIK3CA, DCUN1D1, BCL6	q26.32–q27.3
	76	3.2%	43.1%	187 870 778	189 361 993	1.0E+0		q27.3–q28
	77	3.5%	43.0%	189 365 570	189 367 551	2.0E−3		q28
	78	3.1%	42.7%	189 370 963	195 341 037	6.0E+0		q28–q29
	79	3.3%	42.0%	195 419 229	197 852 564	2.0E+0		q29
5	101	0.3%	47.6%	38 139	685 504	6.0E−1	SDHA	p15.33
	102	2.4%	47.1%	718 972	766 213	5.0E−2		p15.33
	103	0.2%	45.7%	776 473	8 685 711	8.0E+0	TERT	p15.33–31
	104	1.5%	45.6%	8 704 021	8 737 812	3.0E−2		p15.31
	105	0.3%	45.9%	8 753 733	17 516 734	9.0E+0		p15.31–p15.1
	106	1.5%	45.9%	17 602 685	17 634 942	3.0E−2		p15.1
	107	0.3%	44.0%	17 648 614	32 164 826	1.0E+1		p15.1–p14.3
	108	0.2%	43.4%	32 168 437	45 893 362	1.0E+1	DAB2	p13.3–p12
	109	1.2%	40.3%	45 895 885	45 915 513	2.0E−2		p12
	110	1.5%	39.0%	45 939 674	46 381 782	4.0E−1		p12–p11
6	122	1.5%	40.3%	11 488 926	11 492 749	4.0E−3		p24.2
8	159	30.4%	2.6%	172 417	2 232 383	2.0E+0		p23.3–2
	160	31.8%	2.5%	2 254 703	2 260 986	6.0E−3		p23.2
	177	33.2%	14.0%	39 274 995	39 383 000	1.0E−1		p11.22
	199	0.7%	31.7%	91 055 345	114 039 680	2.0E+1	RUNX1T1, TP53INP1	q21.3–q23.3
	200	2.6%	31.1%	114 041 368	114 044 217	3.0E−3		q23.3
	201	0.9%	33.2%	114 052 153	123 551 840	9.0E+0	EXT1	q23.3–q24.13
	202	0.7%	38.2%	123 567 563	130 055 981	6.0E+0	MYC, MTSS1	q24.13–21
	203	1.0%	35.6%	130 070 130	137 656 246	8.0E+0	WISP1	q24.21–23
	204	4.3%	32.6%	137 693 433	137 855 026	2.0E−1		q24.23
	205	1.0%	33.7%	137 862 600	146 114 526	8.0E+0	MAFA, MAFA	q24.3–23
9	207	31.9%	4.0%	204 738	12 433 357	1.0E+1	JAK2, KANK1, PTPRD	p24.3–p23
	208	32.3%	4.0%	12 445 364	13 036 438	6.0E−1		p23
	209	32.8%	3.9%	13 059 473	16 048 844	3.0E+0		p23–p22.3
	210	31.4%	3.9%	16 060 347	20 876 513	5.0E+0	MLLT3	p22.3–p21.3
	211	34.8%	3.4%	20 890 669	21 179 174	3.0E−1		p21.3
	212	35.9%	3.3%	21 194 379	21 778 976	6.0E−1		p21.3
	213	38.7%	3.2%	21 785 018	21 845 577	6.0E−2		p21.3
	214	40.2%	3.0%	21 853 221	22 176 560	3.0E−1	CDKN2A, CDKN2B	p21.3
	215	36.3%	3.3%	22 202 151	23 953 634	2.0E+0		p21.3
	216	34.7%	4.1%	23 971 815	24 725 697	8.0E−1		p21.3
	217	35.0%	3.9%	24 741 204	24 750 179	9.0E−3		p21.3
	218	32.8%	4.5%	24 769 948	25 268 867	5.0E−1		p21.3
	219	30.5%	4.6%	25 294 701	27 670 083	2.0E+0		p21.3–2
	220	30.1%	4.8%	27 678 194	27 700 539	2.0E−2		p21.2
22	425	21.3%	45.5%	24 346 428	24 390 318	4.0E−2		q11.23

Figure S2.

Copy number aberrations in optimal quality (MAPD ≤0.3) samples only (N=777).

Table S2. Most frequent copy number aberrations in the optimal quality samples only (N=777).

Chr	Region ID	Loss Freq	Gain Freq	Start	End	Mb	Genes	cytoBands
1	13	1%	34%	145 394 955	148 544 968	3.0E+0	BCL9, TXNIP	q21.1–2
1	15	0%	35%	149 742 045	161 515 326	1.0E+1	MLLT11, NTRK1, PRCC, TPM3, PYHIN1, EFNA1, MUC1, PLEKHO1, AIM2	q21.2–q23.3
1	16	1%	32%	161 591 477	161 607 441	2.0E−2		q23.3
1	17	1%	34%	161 609 660	161 622 701	1.0E−2		q23.3
1	18	0%	32%	161 641 596	196 703 707	4.0E+1	FCGR2B, PBX1, TPR, LHX4, CDC73	q23.3–q31.3
3	65	3%	32%	143 794 370	151 504 070	8.0E+0	WWTR1	q24–q25.1
3	66	3%	33%	151 520 944	151 546 041	3.0E−2		q25.1
3	67	3%	34%	151 563 527	162 500 864	1.0E+1	CCNL1, GMPS	q25.1–q26.1
3	69	2%	34%	162 640 497	162 702 814	6.0E−2		q26.1
3	70	2%	34%	162 719 684	165 245 914	3.0E+0		q26.1
3	71	2%	34%	165 270 444	165 296 562	3.0E−2		q26.1
3	72	2%	35%	165 314 375	169 905 944	5.0E+0	MECOM	q26.1–2
3	73	2%	34%	169 918 311	175 861 931	6.0E+0	PRKCI, PRKCI	q26.2–32
3	74	2%	33%	175 889 230	175 905 626	2.0E−2		q26.32
3	75	2%	33%	175 920 884	187 866 388	1.0E+1	PIK3CA, DCUN1D1, BCL6	q26.32–q27.3
3	76	3%	33%	187 870 778	189 361 993	1.0E+0		q27.3–q28
3	77	3%	32%	189 365 570	189 367 551	2.0E−3		q28
3	78	3%	34%	189 370 963	195 341 037	6.0E+0		q28–q29
3	79	3%	34%	195 419 229	197 852 564	2.0E+0		q29
5	101	0%	47%	38 139	685 504	6.0E−1	SDHA	p15.33
5	102	1%	45%	718 972	766 213	5.0E−2		p15.33
5	103	0%	46%	776 473	8 685 711	8.0E+0	TERT	p15.33–31
5	104	0%	45%	8 704 021	8 737 812	3.0E−2		p15.31
5	105	0%	45%	8 753 733	17 516 734	9.0E+0		p15.31–p15.1
5	106	0%	45%	17 602 685	17 634 942	3.0E−2		p15.1
5	107	0%	44%	17 648 614	32 164 826	1.0E+1		p15.1–p13.3
5	108	0%	42%	32 168 437	45 893 362	1.0E+1	DAB2	p13.3–p12
5	109	1%	41%	45 895 885	45 915 513	2.0E−2		p12
5	110	1%	39%	45 939 674	46 381 782	4.0E−1		p12–p11
8	199	1%	32%	91 055 345	114 039 680	2.0E+1	RUNX1T1, TP53INP1	q21.3–q23.3
8	200	1%	32%	114 041 368	114 044 217	3.0E−3		q23.3
8	201	1%	34%	114 052 153	123 551 840	9.0E+0	EXT1	q23.3–q24.13
8	202	1%	37%	123 567 563	130 055 981	6.0E+0	MYC, MTSS1	q24.13–21
8	203	1%	36%	130 070 130	137 656 246	8.0E+0	WISP1	q24.21–23
8	204	2%	33%	137 693 433	137 855 026	2.0E−1		q24.23
8	205	1%	34%	137 862 600	146 114 526	8.0E+0	MAFA	q24.3–23
9	209	30%	4%	13 059 473	16 048 844	3.0E+0		p23–p22.3
9	213	30%	3%	21 785 018	21 845 577	6.0E−2		p21.3
9	214	31%	3%	21 853 221	22 176 560	3.0E−1	CDKN2A, CDKN2B	p21.3
9	215	30%	3%	22 202 151	23 953 634	2.0E+0		p21.3

The CN profile was heterogeneous across histology and results were confirmed in our reanalysis of the TCGA data (Figure S3). The frequency of 195 regions (49% were ≤3 Mb and 71% ≤10 Mb; Table S3) was significantly different between ADC and SCC (Q≤0.05). The most significant differences were: more gains in 3q (including genes PIK3CA, MECOM, CCNL1), 22q (NF2, PDGFB) and 12p (KRAS) in SCC; more losses in 3p (RASSF1), 4 (PTTG2, NKX2-1), and 5q in SCC.

Figure S3.

Copy number aberrations in adenocarcinomas (A and D) and squamous cell carcinomas (B and E) in the LACE-Bio (A,B,C) and the Cancer Genome Atlas (TCGA) data (D,E,F).

Table S3. Regions [195] with significantly (Q ≤0.05) different copy number aberration frequency between adenocarcinomas and squamous cell carcinomas in all the samples (N=976).

Chr	Region ID	Loss Freq			Gain Freq		Q	Start	End	Mb	Genes	cytoBands
		ADC	SCC		ADC	SCC
1	1	7%	18%		1%	1%	0.001	754 192	12 833 428	1.0E+0	SKI, PARK7, CHD5, ERRFI1, TP73	p36.21–33
	3	5%	15%		2%	1%	0.001	13 181 849	72 758 707	6.0E+0	EPS15, FGR, JUN, LCK, PAX7, STIL, TAL1, NBL1, EPHB2, MUTYH, NBL1, ARNT	p36.21–p31.1
	5	8%	17%		2%	1%	0.001	72 814 783	110 222 219	4.0E+0	BCL10	p31.1–p13.3
	7	10%	21%		1%	0%	0.001	110 246 359	110 761 020	5.0E−2		p13.3
	8	11%	18%		2%	1%	0.001	110 777 105	120 508 803	1.0E+0	NRAS, RBM15, RAP1A	p13.3–p12
	11	1%	6%		26%	13%	0.001	144 852 910	145 095 477	2.0E−2		q21.1
	12	1%	3%		31%	16%	0.001	145 115 883	145 382 341	3.0E−2		q21.1
	13	0%	1%		43%	22%	0.001	145 394 955	148 544 968	3.0E−1	BCL9, TXNIP	q21.1–2
	15	0%	1%		45%	22%	0.001	149 742 045	161 515 326	1.0E+0	MLLT11, NTRK1, PRCC, TPM3, PYHIN1, EFNA1, MUC1, PLEKHO1, AIM2	q21.2–q23.3
	16	3%	5%		41%	21%	0.027	161 591 477	161 607 441	2.0E−3		q23.3
	17	1%	2%		42%	22%	0.001	161 609 660	161 622 701	1.0E−3		q23.3
	18	0%	0%		37%	22%	0.001	161 641 596	196 703 707	4.0E+0	FCGR2B, PBX1, TPR, LHX4, CDC73	q23.3–q31.3
	20	1%	1%		37%	20%	0.001	196 823 613	196 882 344	6.0E−3		q31.3
	21	1%	1%		37%	19%	0.001	196 922 021	248 687 952	5.0E+0	FH, PHLDA3, LIN9, LGR6, RASSF5	q31.3–q44
	22	2%	3%		37%	20%	0.001	248 773 062	249 212 878	4.0E−2		q44
2	23	1%	1%		2%	7%	0.001	21 494	34 689 435	3.0E+0	ALK, MYCN, NCOA1, RHOB	p25.3–p22.3
	25	1%	0%		2%	10%	0.001	34 741 001	89 572 881	5.0E+0	REL, MSH2	p22.3–p11.2
	32	4%	14%		3%	5%	0.006	141 786 613	141 882 709	1.0E−2		q22.1
	33	4%	17%		4%	5%	0.001	141 893 894	142 075 788	2.0E−2		q22.1
	42	1%	11%		3%	2%	0.001	206 472 683	220 020 335	1.0E+0	IDH1	q33.3–q35
	43	3%	15%		3%	1%	0.001	220 035 105	220 042 675	8.0E−4		q35
	44	3%	13%		3%	1%	0.001	220 056 954	242 834 648	2.0E+0	PAX3, DIS3L2	q35–q37.3
3	46	15%	36%		1%	1%	0.001	63 411	30 625 123	3.0E+0	RAF1, RARB, VHL	p26.3–p24.1
	47	16%	38%		1%	0%	0.001	30 638 028	46 778 842	2.0E+0	MLH1, LIMD1, DLEC1	p24.1–p21.31
	48	18%	41%		1%	0%	0.004	46 804 388	46 831 840	3.0E−3		p21.31
	49	17%	41%		1%	1%	0.001	46 852 679	75 394 787	3.0E+0	RHOA, NCKIPSD, TCTA, USP4, NPRL2, RASSF1, TUSC2, FHIT, NAT6, PBRM1, PRKCD	p21.31–p12.3
	50	18%	43%		2%	1%	0.001	75 444 906	75 554 646	1.0E−2		p12.3
	51	16%	40%		1%	1%	0.001	75 815 879	78 927 132	3.0E−1		p12.3
	52	16%	40%		1%	2%	0.001	78 930 451	78 937 737	7.0E−4		p12.3
	53	17%	39%		2%	2%	0.001	78 939 727	84 046 628	5.0E−1		p12.3–1
	54	15%	35%		3%	6%	0.001	84 068 424	89 392 778	5.0E−1		p12.1–p11.1
	56	13%	33%		5%	7%	0.001	89 423 343	90 418 473	1.0E−1		p11.1
	58	9%	5%		9%	34%	0.001	93 530 364	100 327 532	7.0E−1		q11.1–q12.2
	59	8%	4%		10%	36%	0.001	100 351 896	111 493 739	1.0E+0	TFG	q12.2–q13.13
	60	7%	4%		11%	39%	0.001	111 521 268	111 555 200	3.0E−3		q13.2
	61	7%	3%		10%	40%	0.001	111 567 238	129 762 859	2.0E+0		q13.2–q22.1
	62	6%	2%		12%	43%	0.001	129 784 388	129 810 022	3.0E−3		q22.1
	63	6%	2%		12%	46%	0.001	129 823 705	134 398 090	5.0E−1		q22.1–2
	64	5%	1%		14%	50%	0.001	134 402 484	143 774 592	9.0E−1	XRN1	q22.2–q24
	65	5%	0%		14%	61%	0.001	143 794 370	151 504 070	8.0E−1	WWTR1	q24–q25.1
	66	5%	1%		15%	64%	0.001	151 520 944	151 546 041	3.0E−3		q25.1
	67	5%	0%		15%	67%	0.001	151 563 527	162 500 864	1.0E+0	CCNL1, GMPS	q25.1–q26.1
	68	15%	11%		21%	61%	0.001	162 540 700	162 602 984	6.0E−3		q26.1
	69	4%	1%		18%	68%	0.001	162 640 497	162 702 814	6.0E−3		q26.1
	70	4%	0%		17%	69%	0.001	162 719 684	165 245 914	3.0E−1		q26.1
	71	4%	0%		18%	70%	0.001	165 270 444	165 296 562	3.0E−3		q26.1
	72	3%	0%		18%	74%	0.001	165 314 375	169 905 944	5.0E−1	MECOM	q26.1–2
	73	4%	0%		17%	77%	0.001	169 918 311	175 861 931	6.0E−1	PRKCI, PRKCI	q26.2–32
	74	4%	1%		16%	77%	0.001	175 889 230	175 905 626	2.0E−3		q26.32
	75	4%	0%		16%	78%	0.001	175 920 884	187 866 388	1.0E+0	PIK3CA, DCUN1D1, BCL6	q26.32–q27.3
	76	5%	1%		15%	74%	0.001	187 870 778	189 361 993	1.0E−1		q27.3–q28
	77	5%	1%		15%	74%	0.001	189 365 570	189 367 551	2.0E−4		q28
	78	5%	0%		15%	73%	0.001	189 370 963	195 341 037	6.0E−1		q28–q29
	79	6%	0%		15%	72%	0.001	195 419 229	197 852 564	2.0E−1		q29
4	80	5%	21%		3%	1%	0.001	69 404	9 153 037	9.0E−1	WHSC1	p16.3–1
	82	5%	22%		3%	0%	0.001	9 586 764	34 772 494	3.0E+0		p16.1–p15.1
	84	3%	20%		3%	1%	0.001	34 847 676	48 061 771	1.0E+0	PTTG2	p15.1–p12
	85	3%	16%		5%	4%	0.001	48 083 885	49 085 053	1.0E−1		p12–p11
	86	3%	16%		4%	3%	0.001	49 085 414	49 092 454	7.0E−4		p11
	92	9%	23%		2%	3%	0.001	87 465 741	88 195 494	7.0E−2	AFF1	q21.3–q22.1
	93	5%	19%		2%	4%	0.001	88 208 266	90 739 539	3.0E−1		q22.1
	95	6%	21%		1%	2%	0.001	90 784 528	122 271 282	3.0E+0	TET2	q22.1–q27
	96	7%	25%		1%	2%	0.001	122 282 972	122 288 144	5.0E−4		q27
	97	6%	25%		2%	2%	0.001	122 299 078	134 264 058	1.0E+0	IL2	q27–q28.3
	98	6%	26%		2%	2%	0.001	134 271 747	134 295 157	2.0E−3		q28.3
	99	6%	27%		3%	2%	0.001	134 304 705	166 810 338	3.0E+0		q28.3–q32.3
	100	12%	30%		1%	1%	0.001	166 830 580	190 915 650	2.0E+0	HPGD	q32.3–q35.2
5	111	5%	17%		27%	22%	0.005	49 441 966	49 562 291	1.0E−2		q11.1
	112	10%	26%		16%	9%	0.001	49 597 497	49 608 094	1.0E−3		q11.1
	113	10%	29%		13%	7%	0.001	49 640 141	51 484 497	2.0E−1		q11.1–2
	114	13%	35%		8%	2%	0.001	51 505 665	60 219 800	9.0E−1	PLK2, PDE4D	q11.2–q12.1
	115	15%	36%		6%	1%	0.001	60 241 946	68 828 372	9.0E−1		q12.1–q13.2
	116	17%	36%		5%	1%	0.001	70 306 678	112 950 805	4.0E+0	FER, RASA1, APC, MCC	q13.2–q22.2
	118	18%	36%		5%	1%	0.001	112 997 656	139 372 404	3.0E+0	AFF4, IRF1	q22.2–q31.2
	119	15%	31%		11%	6%	0.010	139 379 707	139 400 093	2.0E−3		q31.2
	120	16%	35%		5%	1%	0.001	139 411 703	180 698 312	4.0E+0	CSF1R, ARHGAP26, NPM1, PDGFRB, NSD1, PTTG1	q31.2–q35.3
6	121	1%	7%		9%	4%	0.001	204 909	11 474 632	1.0E+0		p25.3–p24.2
	123	0%	6%		10%	5%	0.001	11 521 599	31 276 175	2.0E+0	DEK	p24.2–p21.33
	125	1%	6%		10%	2%	0.001	31 297 365	32 528 026	1.0E−1		p21.33–32
	126	1%	7%		11%	2%	0.001	32 561 716	32 577 756	2.0E−3		p21.32
	127	1%	5%		9%	2%	0.001	32 581 816	42 552 548	1.0E+0	PIM1	p21. 32–p21.1
	128	0%	3%		10%	7%	0.026	42 572 859	51 038 424	8.0E−1	VEGFA	p21.1–p12.3
	133	6%	2%		4%	5%	0.017	64 281 705	65 202 867	9.0E−2		q12
	134	11%	3%		2%	3%	0.001	65 286 066	78 962 125	1.0E+0		q12–q14.1
	136	16%	6%		1%	1%	0.001	79 042 157	103 728 158	2.0E+0	UFL1	q14.1–q16.3
	138	14%	7%		1%	2%	0.001	103 766 477	170 913 051	7.0E+0	FOXO3, FYN, MAS1, MLLT4, MYB, ROS1, SASH1, FRK, RPS6KA2, LATS1	q16.3–q27
7	152	2%	1%		14%	20%	0.001	88 259 445	100 958 270	1.0E+0		q21.13–q22.1
8	163	25%	36%		4%	2%	0.023	16 025 118	25 073 138	9.0E−1	PCM1, LZTS1, DMTN, NKX3-1, MTUS1	p22–p21.2
	169	19%	20%		8%	20%	0.047	36 025 847	36 621 588	6.0E−2		p12–p11.23
	170	18%	15%		10%	25%	0.001	36 633 318	37 651 477	1.0E−1		p11.23
	171	18%	13%		11%	28%	0.001	37 667 018	37 767 527	1.0E−2		p11.23
	172	17%	11%		11%	30%	0.001	37 786 457	38 127 768	3.0E−2	PPAPDC1B	p11.23
	173	17%	10%		11%	31%	0.001	38 137 530	38 139 729	2.0E−4	WHSC1L1	p11.23
	174	17%	9%		10%	32%	0.001	38 143 357	38 528 508	4.0E−2	WHSC1L1	p11.23–22
	175	17%	10%		10%	30%	0.001	38 552 757	39 217 074	7.0E−2		p11.22
	178	15%	11%		11%	27%	0.001	39 412 457	39 969 006	6.0E−2		p11.22–21
	179	15%	12%		11%	23%	0.001	39 976 970	41 058 677	1.0E−1		p11.21
	180	14%	11%		12%	22%	0.001	41 077 098	41 261 544	2.0E−2		p11.21
	181	13%	10%		13%	22%	0.003	41 280 152	42 559 586	1.0E−1	KAT6A	p11.21
	182	12%	10%		13%	20%	0.012	42 574 931	43 157 099	6.0E−2		p11.21–1
9	211	28%	42%		3%	3%	0.003	20 890 669	21 179 174	3.0E−2		p21.3
	212	29%	43%		3%	3%	0.001	21 194 379	21 778 976	6.0E−2		p21.3
	213	32%	46%		3%	3%	0.001	21 785 018	21 845 577	6.0E−3		p21.3
	214	33%	48%		3%	2%	0.001	21 853 221	22 176 560	3.0E−2	CDKN2A, CDKN2B	p21.3
	215	30%	44%		3%	3%	0.012	22 202 151	23 953 634	2.0E−1		p21.3
	216	28%	41%		4%	4%	0.017	23 971 815	24 725 697	8.0E−2		p21.3
	232	21%	16%		1%	4%	0.001	71 035 938	91 434 530	2.0E+0		q21.11–q22.1
	233	19%	14%		2%	5%	0.003	91 440 590	96 341 196	5.0E−1	FAM120A	q22.1–31
	234	18%	14%		1%	4%	0.005	96 382 906	141 054 761	4.0E+0	ABL1, SET, TAL2, NR4A3, NUP214, GFI1B, DAB2IP, DEC1, PTCH1, TSC1	q22.31–q34.3
10	235	3%	8%		5%	3%	0.001	126 070	35 317 317	4.0E+0	BMI1, NET1, MAP3K8, BMI1, MLLT10, ZMYND11	p15.3–p11.21
	237	3%	6%		6%	3%	0.001	35 351 249	42 608 180	7.0E−1		p11.21–q11.21
	238	4%	9%		4%	5%	0.005	42 614 561	46 177 093	4.0E−1	RET, RASSF4	q11.21–22
	239	7%	14%		8%	9%	0.024	46 965 151	47 127 279	2.0E−2		q11.22
	241	6%	12%		4%	3%	0.001	48 302 618	51 594 462	3.0E−1	NCOA4	q11.22–23
	242	7%	15%		2%	2%	0.001	51 785 728	63 841 130	1.0E+0	CCDC6	q11.23–q21.2
	243	7%	16%		3%	2%	0.001	63 861 600	68 064 594	4.0E−1		q21.2–3
	244	8%	17%		2%	2%	0.001	68 077 764	68 114 481	4.0E−3		q21.3
	245	8%	17%		3%	2%	0.001	68 128 614	83 887 644	2.0E+0		q21.3–q23.1
	246	8%	20%		3%	1%	0.001	83 894 966	91 400 384	8.0E−1	PTEN	q23.1–q23.31
	247	9%	18%		3%	1%	0.001	91 422 054	131 496 457	4.0E+0	FRAT1, DMBT1, FGFR2, TLX1, MXI1, NFKB2, WDR11, C10orf90, DMBT1, PDCD4, SUFU, NEURL1	q23.31–q26.3
	249	10%	19%		3%	2%	0.001	131 540 611	135 434 303	4.0E−1		q26.3
11	250	3%	12%		0%	0%	0.001	192 764	27 362 359	3.0E+0	RRAS2, CSNK2A3, HRAS, LMO1, AKIP1, CARS, CDKN1C, HTATIP2, PRKCDBP	p15.5–p14.1
	258	4%	9%		4%	3%	0.015	65 186 348	65 271 832	9.0E−3		q13.1
	259	1%	5%		7%	5%	0.004	65 283 143	67 656 861	2.0E−1	BRMS1	q13.1–2
	264	3%	8%		7%	5%	0.001	71 620 699	77 051 820	5.0E−1		q13.4–5
	265	4%	10%		8%	3%	0.001	77 055 862	89 473 402	1.0E+0	PICALM	q13.5–q14.3
	266	8%	12%		5%	2%	0.007	89 654 897	118 621 410	3.0E+0	YAP1, DDX6, KMT2A, POU2AF1, ZBTB16, PDGFD, CADM1, ATM, ARHGAP20	q14.3–q23.3
12	268	6%	1%		11%	22%	0.001	189 400	866 208	7.0E−2		p13.33
	270	7%	1%		9%	24%	0.001	876 288	9 623 841	9.0E−1	FGF6, ING4	p13.33–31
	273	8%	1%		9%	24%	0.001	9 739 885	11 501 856	2.0E−1	STYK1	p13.31–2
	274	14%	7%		8%	22%	0.001	11 515 281	11 543 338	3.0E−3		p13.2
	275	8%	1%		7%	23%	0.001	11 555 467	17 613 438	6.0E−1	CDKN1B, ETV6, CREBL2	p12.3–p13.2
	276	5%	1%		12%	23%	0.001	17 635 521	31 241 345	1.0E+0	KRAS	p12.3–p11.21
	277	4%	1%		10%	20%	0.001	31 296 219	33 982 162	3.0E−1		p11.21–1
	278	4%	2%		9%	18%	0.001	33 994 599	34 749 065	8.0E−2		p11.1
	279	3%	2%		9%	15%	0.026	34 755 947	34 828 211	7.0E−3		p11.1
	285	3%	1%		5%	6%	0.004	72 362 265	93 623 290	2.0E+0	BTG1, ZDHHC17	q21.1–q22
13	294	13%	23%		4%	3%	0.018	81 036 784	103 258 600	2.0E+0		q31.1–q33.1
	295	13%	24%		4%	2%	0.001	103 273 465	103 286 814	1.0E−3		q33.1
14	303	3%	11%		17%	8%	0.001	28 292 800	34 485 112	6.0E−1		q12–q13.1
	304	2%	10%		22%	9%	0.001	34 490 900	35 596 092	1.0E−1		q13.1–2
	305	2%	10%		25%	11%	0.001	35 605 465	38 959 413	3.0E−1	NKX2-1, MBIP	q13.2–q21.1
	306	2%	10%		21%	10%	0.001	38 976 380	39 121 670	1.0E−2		q21.1
	307	2%	10%		20%	9%	0.001	39 186 659	39 246 264	6.0E−3		q21.1
	308	3%	10%		18%	8%	0.001	39 311 307	41 601 018	2.0E−1	PNN	q21.1
	310	3%	10%		16%	8%	0.001	41 673 714	43 083 557	1.0E−1		q21.1
	311	4%	11%		12%	7%	0.022	43 093 389	61 786 976	2.0E+0		q21.1–q23.1
	312	4%	12%		10%	7%	0.014	61 799 920	61 917 178	1.0E−2		q23.1
15	327	13%	8%		1%	2%	0.001	20 161 372	34 671 061	1.0E+0		q11.1–q14
	329	14%	7%		1%	2%	0.001	34 870 223	38 622 707	4.0E−1		q14
	330	14%	7%		1%	1%	0.001	38 623 948	43 995 380	5.0E−1	ZFYVE19, BUB1B	q14–q15.3
	331	10%	4%		2%	3%	0.001	44 016 417	76 738 959	3.0E+0	PML, ARID3B, PML	q15.3–q24.3
	332	8%	3%		2%	5%	0.001	76 752 698	93 407 788	2.0E+0	AKAP13, FES, FES, ST20, IDH2	q24.3–q26.1
	333	7%	2%		3%	7%	0.001	93 429 646	102 397 317	9.0E−1		q26.1–3
16	334	4%	10%		4%	0%	0.001	83 887	6 988 411	7.0E−1	TSC2, AXIN1	p13.3
	335	4%	9%		13%	4%	0.008	6 999 231	7 013 483	1.0E−3		p13.3
	336	4%	7%		7%	1%	0.001	7 023 927	28 609 205	2.0E+0	MYH11, TNFRSF17, LITAF, PALB2	p13.3–p11.2
	338	2%	5%		6%	2%	0.002	28 628 225	32 137 965	4.0E−1	FUS, PYCARD	p11.2
17	346	16%	24%		1%	1%	0.004	400 959	18 928 388	2.0E+0	CRK, ELAC2, GAS7, USP6, TP53, KCTD11, DPH1, FLCN, HIC1, XAF1	p13.3–p11.2
	348	8%	17%		3%	3%	0.001	21 690 667	22 217 883	5.0E−2		p11.2–1
	352	2%	4%		12%	7%	0.001	34 815 264	36 854 507	2.0E−1		q12
	353	1%	6%		10%	8%	0.001	36 861 302	45 005 703	8.0E−1	ERBB2, MLLT6, RARA, WNT3, BRCA1	q12–q21.32
	360	1%	7%		15%	11%	0.001	80 185 188	80 263 427	8.0E−3		q25.3
18	361	7%	5%		4%	12%	0.001	12 842	14 240 269	1.0E+0	YES1, EPB41L3	p11.32–21
	362	7%	5%		4%	10%	0.001	14 270 974	15 377 471	1.0E−1		p11.21
	363	8%	5%		5%	9%	0.026	18 554 999	21 648 788	3.0E−1		q11.1–2
	364	11%	6%		4%	10%	0.001	21 659 508	24 123 575	2.0E−1	ZNF521, SS18	q11.2
	365	14%	7%		3%	10%	0.001	24 143 454	27 670 629	4.0E−1		q11.2–q12.1
	366	15%	7%		4%	11%	0.001	27 678 287	29 104 698	1.0E−1		q12.1
	367	15%	8%		3%	9%	0.001	29 119 357	29 715 321	6.0E−2		q12.1
	368	16%	9%		3%	9%	0.001	29 736 017	29 737 077	1.0E−4		q12.1
	369	14%	8%		3%	9%	0.001	29 754 749	29 779 205	2.0E−3		q12.1
	370	14%	8%		3%	9%	0.001	29 790 889	30 339 291	5.0E−2		q12.1
	371	15%	9%		2%	7%	0.001	30 358 394	33 590 529	3.0E−1		q12.1–2
19	378	11%	6%		0%	0%	0.001	1 335 531	9 051 725	8.0E−1	MLLT1, SH3GL1, TCF3, VAV1	p13.3–2
	380	11%	4%		1%	3%	0.001	9 059 232	20 499 493	1.0E+0	LYL1, RAB8A, ELL, CDKN2D	p13.2–p12
	382	11%	4%		2%	5%	0.001	20 723 899	20 758 368	3.0E−3		p12
	383	10%	5%		2%	3%	0.001	20 769 956	24 505 466	4.0E−1		p12–p11
20	399	6%	3%		10%	21%	0.001	69 094	13 595 807	1.0E+0	RASSF2	p13–p12.1
	400	5%	3%		11%	20%	0.004	13 618 382	14 780 319	1.0E−1		p12.1
	402	5%	4%		12%	20%	0.008	14 827 680	15 557 228	7.0E−2		p12.1
	403	5%	2%		12%	22%	0.001	15 560 791	23 693 161	8.0E−1		p12.1–p11.21
	404	3%	3%		13%	22%	0.024	23 700 872	25 672 987	2.0E−1		p11.21–p11.1
21	419	10%	18%		1%	1%	0.014	9 648 315	10 964 139	1.0E−1		p11.2–1
	420	8%	16%		1%	1%	0.002	14 344 537	34 787 312	2.0E+0	OLIG2, TCP10L	q11.2–q22.11
	421	8%	19%		1%	1%	0.001	34 796 886	48 097 610	1.0E+0	ERG, ETS2, RUNX1, SIK1	q22.11–q22.3
22	422	6%	5%		2%	9%	0.001	16 054 713	19 009 167	3.0E−1		q11.1–21
	423	5%	2%		3%	14%	0.001	19 026 877	21 462 601	2.0E−1		q11.21
	424	5%	3%		3%	11%	0.001	21 804 610	24 338 651	3.0E−1	BCR, SMARCB1	q11.21–23
	426	6%	6%		12%	18%	0.026	24 394 088	24 396 598	3.0E−4		q11.23
	427	7%	3%		3%	11%	0.001	24 398 768	25 917 803	2.0E−1		q11.23–q12.1
	428	9%	3%		1%	11%	0.001	25 942 595	36 907 098	1.0E+0	EWSR1, PATZ1, RASL10A, CHEK2, MN1, NF2	q12.1–3
	429	8%	3%		2%	10%	0.001	36 919 447	39 343 292	2.0E−1		q12.3–q13.1
	430	9%	4%		1%	11%	0.001	39 363 830	42 517 758	3.0E−1	PDGFB, MKL1	q13.1–2
	431	9%	5%		1%	9%	0.001	42 518 382	51 213 826	9.0E−1	PIM3, PRR5	q13.2–33

Copy-number aberrations associated with prognosis

The median follow-up for DFS (510 events) was 5.3 years. In univariate analyses (Table 2), 14 focal regions (11 ≤3 Mb, 14 ≤10 Mb) in loci 7p11–12, 9p21, 18q12, 19p11–13 were prognostic (P≤0.005) with Q≤0.142. Losses associated with shorter DFS were in: 8 regions in 9p21.3 (loss frequency: 31–40%, including CDKN2A/B), with HR_loss =1.5 (95% CI: 1.2–1.9) (P<0.001, Q=0.02); one region in 19p13 [STK11, 11%, HR_loss =2.4 (1.3–4.3), P=0.005, Q=0.15]; one in 18q12.1 [12%, HR_loss =1.6 (1.2–2.3), P=0.004, Q=0.12]. Other seemingly deleterious losses were found in 19p11–13 (MLLT1, SH3GL1, TCF3, VAV1). Gains in 7p11–12 (frequency: 17%) were associated with shorter DFS [HR_gain =2.0 (1.2–3.2), P=0.005, Q=0.14]. Two of these regions (7p12.3 and 9p21.3) remained significant in multivariate analyses (Table S4), which also suggested a benefit [HR_loss =0.32 (0.16–0.61), P<0.001] for losses in a region in 1p31–36 (9.8%), including EPS15, FGR, JUN, LCK, PAX7, STIL, TAL1, NBL1, EPHB2, MUTYH, ARNT. Penalized regression confirmed the prognostic role of the region in 9p21.3, plus another one containing CDKN2A/B (Table S5).

Table 2. Genomic regions with prognostic effect of copy number aberrations (CNA).

Region ID	Chr	cytoBands	Mb	CNA frequency			Disease-free survival					Overall survival					Lung-cancer specific survival				Genes
				Loss	Gain		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q
142	7	p12.3–p11.2	8.0E+0	0.7%	17%		0.51 (0.31–0.82)	2.0 (1.2–3.2)	0.005	0.142
185	8	p11.1–q11.1	4.0E+0	7.3%	17.4%												2.0 (1.2–3.1)	0.51 (0.32–0.82)	0.005	0.130
211	9	p21.3	3.0E−1	34.8%	3.4%		1.7 (1.3–2.3)	0.57 (0.44–0.76)	<0.001	0.020		1.8 (1.4–2.5)	0.55 (0.41–0.74)	<0.001	0.029		1.8 (1.4–2.4)	0.55 (0.41–0.74)	<0.001	0.019
212	9	p21.3	6.0E−1	35.9%	3.3%		1.6 (1.2–2.0)	0.64 (0.49–0.84)	0.001	0.076		1.7 (1.3–2.2)	0.59 (0.44–0.79)	<0.001	0.044		1.6 (1.2–2.1)	0.62 (0.47–0.83)	0.001	0.062
213	9	p21.3	6.0E−2	38.7%	3.2%		1.5 (1.2–1.9)	0.67 (0.53–0.86)	0.001	0.076		1.5 (1.2–2.0)	0.66 (0.51–0.86)	0.002	0.079		1.5 (1.2–2.0)	0.65 (0.50–0.85)	0.001	0.062
214	9	p21.3	3.0E−1	40.2%	3.0%		1.5 (1.2–1.9)	0.66 (0.53–0.81)	<0.001	0.020		1.5 (1.2–1.9)	0.67 (0.54–0.84)	<0.001	0.049		1.6 (1.2–2.0)	0.64 (0.51–0.80)	<0.001	0.021	CDKN2A, CDKN2B
215	9	p21.3	2.0E+0	36.3%	3.3%		1.6 (1.2–2.1)	0.62 (0.48–0.81)	<0.001	0.034		1.7 (1.3–2.2)	0.59 (0.45–0.79)	<0.001	0.044		1.7 (1.3–2.2)	0.61 (0.46–0.80)	<0.001	0.029
217	9	p21.3	9.0E−3	35.0%	3.9%		1.7 (1.3–2.2)	0.60 (0.46–0.79)	<0.001	0.026		1.5 (1.2–2.1)	0.65 (0.48–0.87)	0.004	0.084		1.7 (1.3–2.3)	0.58 (0.44–0.78)	<0.001	0.026
218	9	p21.3	5.0E−1	32.8%	4.5%		1.6 (1.2–2.2)	0.61 (0.45–0.82)	0.001	0.076		1.7 (1.2–2.3)	0.60 (0.43–0.83)	0.002	0.079		1.7 (1.2–2.4)	0.59 (0.42–0.82)	0.002	0.064
219	9	p21.3–2	2.0E+0	30.5%	4.6%		1.8 (1.2–2.6)	0.57 (0.39–0.82)	0.003	0.106		1.8 (1.2–2.7)	0.56 (0.37–0.83)	0.004	0.084		1.9 (1.2–2.8)	0.54 (0.36–0.80)	0.002	0.085
220	9	p21.2	2.0E−2	30.1%	4.8%												1.6 (1.2–2.1)	0.64 (0.47–0.86)	0.004	0.11
222	9	p21.1	2.0E+0	27.8%	5.2%							1.9 (1.2–2.8)	0.54 (0.36–0.81)	0.003	0.084
223	9	p21.1	7.0E−1	26.0%	6.4%							1.8 (1.2–2.8)	0.54 (0.36–0.82)	0.003	0.084
312	14	q23.1	1.0E−1	8.5%	8.9%							2.2 (1.3–3.6)	0.46 (0.28–0.76)	0.002	0.079
366	18	q12.1	1.0E+0	10.8%	8.6%							1.9 (1.2–3.1)	0.52 (0.32–0.82)	0.005	0.105
367	18	q12.1	6.0E−1	10.9%	7.1%							2.0 (1.2–3.3)	0.49 (0.30–0.80)	0.005	0.092
368	18	q12.1	1.0E−3	12.1%	6.6%		1.6 (1.2–2.3)	0.61 (0.44–0.85)	0.004	0.119		1.7 (1.2–2.5)	0.58 (0.41–0.82)	0.002	0.079		1.7 (1.2–2.4)	0.59 (0.42–0.85)	0.004	0.11
376	19	p13.3	1.0E+0	10.7%	0.4%		2.4 (1.3–4.3)	0.42 (0.23–0.77)	0.005	0.142											FSTL3, STK11
378	19	p13.3–2	8.0E+0	9.2%	0.3%		2.6 (1.4–4.8)	0.38 (0.21–0.72)	0.003	0.106		2.9 (1.5–5.6)	0.34 (0.18–0.66)	0.001	0.078		3.4 (1.7–6.6)	0.29 (0.15–0.58)	<0.001	0.029	MLLT1, SH3GL1, TCF3, VAV1
379	19	p13.2	3.0E−3	10.3%	0.9%		2.2 (1.3–3.8)	0.45 (0.26–0.76)	0.003	0.106		2.3 (1.3–4.2)	0.43 (0.24–0.77)	0.004	0.092		2.6 (1.4–4.6)	0.39 (0.22–0.69)	0.001	0.062
380	19	p13.2–p12	1.0E+1	8.0%	2.2%							2.7 (1.4–5.1)	0.38 (0.20–0.73)	0.004	0.084						LYL1, RAB8A, ELL, CDKN2D
383	19	p12–p11	4.0E+0	8.0%	2.5%		2.4 (1.4–4.2)	0.42 (0.24–0.74)	0.003	0.106		2.7 (1.5–5.0)	0.36 (0.2–0.66)	<0.001	0.066		2.5 (1.3–4.5)	0.41 (0.22–0.75)	0.004	0.11

The univariate hazard ratio (HR) for loss shows the relative risk of a patient with a 2-fold lower CN, for example one copy as compared to two copies. The HR for gain shows the relative risk of a patient with a 2-fold higher CN, for example four copies as compared to two copies. Of note, HR for gain is 1/HR for loss. CI, confidence interval; Chr, chromosome. *, hazard ratio for a 2-fold lower copy number; **, hazard ratio for a 2-fold higher copy number.

Table S4. Prognostic effect of the copy number of genomic regions. Multivariate results.

Chr	Region ID	CNA frequency			Disease-free survival				Overall survival				Lung-cancer specific survival			Mb	Genes	cytoBands
		Losses	Gains		HR for loss* (95% CI)	HR for gain** (95% CI)	P		HR for loss* (95% CI)	HR for gain** (95% CI)	P		HR for loss* (95% CI)	HR for gain** (95% CI)	P
1	3	9.8%	1.6%		0.32 (0.16–0.61)	3.2 (1.6–6.1)	<0.001						0.21 (0.10–0.43)	4.8 (2.3–10)	<0.001	6.0E+1	EPS15, FGR, JUN, LCK, PAX7, STIL, TAL1, NBL1, EPHB2, MUTYH, ARNT	p36.21–p31.1
3	72	1.8%	44.8%						1.8 (1.3–2.6)	0.55 (0.38–0.79)	0.001					5.0E+0	MECOM	q26.1–2
6	122	1.5%	40.3%										0.78 (0.66–0.91)	1.3 (1.1–1.5)	0.002	4.0E−3		p24.2
7	142	0.7%	17.3%		0.46 (0.29–0.75)	2.1 (1.3–3.5)	0.002									8.0E+0		p12.3–p11.2
8	159	30.4%	2.6%										0.53 (0.34–0.83)	1.9 (1.2–3.0)	0.005	2.0E+0		p23.3–2
9	211	34.8%	3.4%		1.9 (1.4–2.5)	0.54 (0.41–0.72)	<0.001		2.1 (1.5–2.8)	0.49 (0.36–0.66)	<0.001		2.1 (1.5–3.0)	0.47 (0.34–0.65)	<0.001	3.0E−1		p21.3
19	378	9.2%	0.3%										3.7 (1.7–7.7)	0.27 (0.13–0.58)	<0.001	8.0E+0	MLLT1, SH3GL1, TCF3, VAV1	p13.3–2
20	409	0.8%	20.2%										2.3 (1.2–4.2)	0.44 (0.24–0.81)	0.009	1.0E−1		q11.21

*, hazard ratio for a 2-fold lower copy number; **, hazard ratio for a 2-fold higher copy number.

Table S5. Prognostic effect of the copy number of genomic regions. Multivariate results obtained via penalized regression.

Chr	Region ID	CNA frequency			Disease-free survival				Overall survival				Lung-cancer specific survival			Mb	Genes	cytoBands
		Losses	Gains		HR for loss* (95% CI)	HR for gain** (95% CI)	P		HR for loss* (95% CI)	HR for gain** (95% CI)	P		HR for loss* (95% CI)	HR for gain** (95% CI)	P
1	3	9.8%	1.6%										0.92 (0.83–1.0)	1.1 (0.98–1.2)	0.129	6.0E+1	EPS15, FGR, JUN, LCK, PAX7, STIL, TAL1, NBL1, EPHB2, MUTYH, NBL1, ARNT	p36.21–p31.1
1	14	15.6%	25.9%										1.0 (0.92–1.1)	0.99 (0.90–1.1)	0.837	2.0E−1		q21.2
3	71	2.3%	42.6%										1.0 (0.90–1.1)	0.99 (0.88–1.1)	0.919	3.0E−2		q26.1
6	122	1.5%	40.3%										0.97 (0.87–1.1)	1.0 (0.93–1.1)	0.554	4.0E−3		p24.2
7	142	0.7%	17.3%										0.97 (0.87–1.1)	1.0 (0.93–1.1)	0.531	8.0E+0		p12.3–p11.2
8	166	23.5%	7.4%										0.98 (0.89–1.1)	1.0 (0.91–1.1)	0.771	1.0E+0		p12
8	185	7.3%	17.4%										1.0 (0.94–1.1)	0.96 (0.87–1.1)	0.460	4.0E+0		p11.1–q11.1
9	211	34.8%	3.4%		1.0 (0.91–1.2)	0.98 (0.86–1.1)	0.686						1.1 (0.92–1.2)	0.94 (0.81–1.1)	0.427	3.0E−1		p21.3
9	214	40.2%	3.0%		1.0 (0.88–1.1)	1.0 (0.88–1.1)	0.984						1.0 (0.89–1.2)	0.98 (0.85–1.1)	0.724	3.0E−1	CDKN2A, CDKN2B	p21.3
9	217	35.0%	3.9%										1.0 (0.90–1.2)	0.98 (0.85–1.1)	0.717	9.0E−3		p21.3
12	270	3.9%	16.4%										0.98 (0.88–1.1)	1.0 (0.92–1.1)	0.705	9.0E+0	FGF6, ING4	p13.33–31
14	312	8.5%	8.9%										1.0 (0.92–1.1)	0.99 (0.90–1.1)	0.844	1.0E−1		q23.1
18	368	12.1%	6.6%										1.1 (0.95–1.2)	0.95 (0.86–1.1)	0.321	1.0E−3		q12.1
19	378	9.2%	0.3%										1.1 (0.92–1.2)	0.95 (0.84–1.1)	0.443	8.0E+0	MLLT1, SH3GL1, TCF3, VAV1	p13.3–2
19	379	10.3%	0.9%										1.0 (0.90–1.1)	0.99 (0.87–1.1)	0.821	3.0E−3		p13.2
19	383	8.0%	2.5%										1.0 (0.91–1.1)	0.99 (0.88–1.1)	0.802	4.0E+0		p12–p11
20	410	1.1%	20.8%										1.0 (0.9–1.1)	0.99 (0.89–1.1)	0.895	2.0E+0	HCK	q11.21–22

Results from a model adjuster by treatment arm, patient age, sex, performance status (PS), histology, T, and N stage. *, hazard ratio for a 2-fold lower copy number; **, hazard ratio for a 2-fold higher copy number.

The median follow-up for OS (451 events) was 5.3 years. The above-mentioned CN losses in 9p21, 18q12, 19p13 were also prognostic of shorter OS (P≤0.005, Q≤0.092; Table 2), together with 5 additional regions in 9p21.1, 18q12.1, and 19p12–13 (ELL). One further focal region on 14q23.1 (8.5% of losses, 89% of gains) was prognostic for OS (P=0.002, Q=0.079), with HR_loss =2.2 (1.3–3.6), corresponding to HR_gain =0.46 (0.28–0.76). The prognostic role of a region in 9p12.3 was confirmed in multivariate analyses (Table S4), together with the possible benefit for gains in 3q26 [MECOM, 45%, HR_gain =0.55 (0.38–0.79), P=0.001]. Penalized regression (Table S5) did not select any region for OS.

The median follow-up for LCSS (427 events) was 5.0 years. Results were similar to DFS, with the addition of one region in chr8, for which gains (17%) were associated with longer LCSS [HR_gain =0.51 (0.32–0.82), P=0.005, Q=0.13]. In multivariate analyses (Table S4), two of the three regions associated with DFS (chr3 and 9) were also associated with LCSS, in addition to regions in 6p24.2 [HR_gain =1.3 (1.1–1.5), P=0.002], 8p23 [HR_loss =0.53 (0.34–0.83), P=0.005], 19p13 [MLLT1, SH3GL1, TCF3, VAV1; HR_loss =3.7 (1.7–7.7), P<0.001], and 20q11.21 [HR_gain =0.44 (0.24–0.81), P=0.009]. Penalized regression (Table S5) selected 17 prognostic regions for LCSS on chr1 (EPS15, FGR, JUN, LCK, PAX7, STIL, TAL1, NBL1, EPHB2, MUTYH, NBL1, ARNT), chr9 (CDKN2A/B), chr12 (FGF6, ING4), chr19 (MLLT1, SH3GL1, TCF3, VAV1), and chr20 (HCK).

Copy-number aberrations associated with the effect of ACT

The average ACT effect on DFS estimated within the 976 patients with CN data was HR_ACT =0.85 (0.71–1.0) (P=0.06). Univariate analyses (Table 3) identified five regions in 14q32.33 as potentially predictive of better response to ACT (P<0.05), but with very high Q values. The effect of CNAs in these regions was similar. CN loss in one region in 14q32.33 had HR_loss for interaction of 0.42 (0.22–0.83) (P=0.012, Q=0.010), corresponding to HR_gain for interaction of 2.4 (1.2–4.6). This means that, given a treatment effect (ACT vs. control) of HR_[ACT|CN=2] =0.85 for a patient with CN=2, such an effect is stronger for a patient with CN=1 (HR_[ACT|CN=1] =0.42×0.85=0.36) and reversed with CN=4 (HR_[ACT|CN=4] =2.4×0.85=2.0). The predictive role of this region was the only confirmed in multivariate analyses (Table S6), with HR_loss for interaction of 0.39 (0.20–0.79) (P=0.009).

Table 3. Predictive effect of the copy number aberration (CNA) at various genomic regions for the magnitude of the effect of adjuvant chemotherapy. Univariate results.

Region ID	Chr	cytoBands	Mb	CNA Frequency			Disease-free survival					Overall survival					Lung-cancer specific survival				Genes
				Losses	Gains		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q
160	8	p23.2	6.0E−3	31.8%	2.5%							0.73 (0.57–0.95)	1.4 (1.1–1.8)	0.019	0.76
235	10	p15.3–p11.21	4.0E+1	5.9%	4.0%							0.27 (0.09–0.82)	3.6 (1.2–11)	0.021	0.76						BMI1, NET1, MAP3K8, BMI1, MLLT10, ZMYND11
237	10	p11.21–q11.21	7.0E+0	4.6%	4.5%							0.28 (0.09–0.83)	3.6 (1.2–11)	0.022	0.76
238	10	q11.21–22	4.0E+0	7.2%	4.6%							0.26 (0.08–0.8)	3.9 (1.3–12)	0.019	0.76						RET, RASSF4
318	14	q32.33	2.0E−2	8.9%	10.0%		0.40 (0.19–0.88)	2.5 (1.1–5.4)	0.022	>0.99							0.33 (0.14–0.78)	3.0 (1.3–7.1)	0.011	0.90
319	14	q32.33	1.0E−1	10.8%	11.4%		0.42 (0.22–0.83)	2.4 (1.2–4.6)	0.012	0.99							0.38 (0.18–0.79)	2.6 (1.3–5.4)	0.009	0.90
324	14	q32.33	5.0E−2	9.1%	9.0%		0.37 (0.15–0.87)	2.7 (1.2–6.5)	0.022	0.99
325	14	q32.33	3.0E−2	16.3%	8.3%		0.68 (0.51–0.92)	1.5 (1.1–2.0)	0.012	0.99							0.66 (0.48–0.91)	1.5 (1.1–2.1)	0.012	0.90
326	14	q32.33	4.0E−1	8.7%	9.2%		0.33 (0.14–0.77)	3.1 (1.3–7.2)	0.011	0.99
333	15	q26.1–3	9.0E+0	4.7%	4.5%							0.21 (0.06–0.71)	4.8 (1.4–16)	0.012	0.76
408	20	q11.21	5.0E−1	0.8%	20.8%												0.18 (0.06–0.52)	5.6 (1.9–16)	0.002	0.57
409	20	q11.21	1.0E−1	0.8%	20.2%												0.17 (0.05–0.54)	5.9 (1.9–19)	0.003	0.57

*, hazard ratio for a 2-fold lower copy number; **, hazard ratio for a 2-fold higher copy number.

Table S6. Predictive effect of the copy number of genomic regions. Multivariate results.

Chr	Region ID	CNA frequency			Disease-free survival				Overall survival				Lung-cancer specific survival			Mb	cytoBands
		Losses	Gains		HR for loss* (95% CI)	HR for gain** (95% CI)	P		HR for loss* (95% CI)	HR for gain** (95% CI)	P		HR for loss* (95% CI)	HR for gain** (95% CI)	P
8	159	30.4%	2.6%						0.42 (0.19–0.93)	2.4 (1.1–5.2)	0.032					2.0E+0	p23.3–2
14	319	10.8%	11.4%		0.39 (0.20–0.79)	2.5 (1.3–5.1)	0.009		0.35 (0.17–0.74)	2.8 (1.4–6.0)	0.006		0.37 (0.17–0.82)	2.7 (1.2–5.9)	0.015	1.0E−1	q32.33
20	409	0.8%	20.2%										0.11 (0.03–0.39)	8.8 (2.6–30)	<0.001	1.0E−1	q11.21

Results from a model adjuster by treatment arm, patient age, sex, performance status (PS), histology, T, and N stage. The multivariate model has been obtained via stepwise selection (αin =0.10 and αout =0.01). Only regions with P<0.005 are shown. *, hazard ratio for a 2-fold lower copy number; **, hazard ratio for a 2-fold higher copy number.

The average effect of ACT on OS was HR_ACT=0.95 (0.79–1.1) (P=0.58). At a raw P<0.05, 5 regions were possibly associated to the ACT effect for OS, but with very high Q values (Table 3). One region in 8p23.2 showed a treatment effect enhanced for the 31.8% of patients with a CN loss [HR_loss for interaction 0.73 (0.57–0.95), P=0.019, Q=0.76), meaning that the HR for a patient with CN=1 was HR_[ACT|CN=1] =0.73×0.95=0.69. An adjacent region in 8p23 was selected in multivariate analyses (Table S6), with a similar effect [HR_loss for interaction 0.42 (0.19–0.93), P=0.032). In univariate analyses, 3 regions in chr10 (BMI1, NET1, MAP3K8, BMI1, MLLT10, ZMYND11, RET, RASSF4) with 5–7% losses and 4–5% gains showed predictive effects with HR_loss for interaction 0.26–0.28 and CN_gain for interaction 3.6–3.9. One region in 15q26 (losses: 4.7%, gains: 4.5%) had HR_loss for interaction 0.21 (0.06–0.71) (P=0.012, Q=0.76) corresponding to HR_gain for interaction 4.8 (1.4–16). In multivariate analyses (Table S6) one region in 14q32.33 was predictive (P=0.006), with HR_loss for interaction 0.35 (0.17–0.74) and HR_gain for interaction 2.8 (1.4–6.0).

The average effect of ACT on LCSS was HR_ACT =0.83 (0.68–1.0) (P=0.05). Three of the above-mentioned regions in 14q32 predictive of ACT effect for DFS were also predictive for LCSS (Table 3). Two additional regions in 20q11.21 (gain frequency: 20%) had possibly significant interaction with ACT, with HR_gain for interaction 5.6 (1.9–16) (P=0.002, Q=0.57) and 5.9 (1.9–19) (P=0.003, Q=0.57), respectively. Two of them (14q32 and 20q11) were confirmed in multivariate analyses (Table S6).

Penalized regression did not select any predictive region for either endpoint.

Sensitivity analyses

The results within the optimal quality sample subgroup (Tables S7-S10) were consistent with those of the whole population. Table S11 shows the genomic regions for which the prognostic effect was significantly different between ADC and SCC (interaction P<0.005). CN gains in two regions in 1q23–31 (FCGR2B, PBX1, TPR, LHX4, CDC73) were associated to shorter DFS in ADC [HR =2.8 (1.3–5.8) and 2.3 (1.1–4.7)] and longer DFS in SCC [HR =0.44 (0.18–1.1) and 0.53 (0.27–1.0)]. One of these regions showed similar results for LCSS. Similar results were observed for 3 regions in 7p11 (also for LCSS and including EGFR), one in 7q11, one in 11p14 (also for OS), and one in 20q11, with increased risk in ADC and reduced risk in SCC for CN gains. Of note, only one region (chr11p14) had quite low interaction Q-value and only for OS (Q=0.056). Conversely, CN gains in 3 further regions [1p13, 4p12–15 (PTTG2), 4q27] were associated to longer OS in ADC [HR =0.50 (0.19–1.3), 0.20 (0.07–0.57), and 0.51 (0.28–0.92), respectively] than in SCC [HR =2.4 (1.0–5.6), 2.1 (0.9–4.8), and 1.6 (0.97–2.6), respectively].

Table S7. Prognostic effect of the copy number of genomic regions. Univariate results in optimal quality samples only.

Chr	Region ID	CNA frequency			Disease-free survival					Overall survival					Lung-cancer specific survival				Mb	Genes	cytoBands
		Losses	Gains		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q
3	71	2.1%	34.0%							1.8 (1.2–2.6)	0.56 (0.39–0.81)	0.002	0.096						3.0E−2		q26.1
	72	2.1%	34.7%		1.8 (1.2–2.6)	0.57 (0.39–0.83)	0.004	0.164		2.1 (1.4–3.1)	0.48 (0.32–0.73)	<0.001	0.096		1.9 (1.2–2.8)	0.54 (0.35–0.81)	0.003	0.123	5.0E+0	MECOM	q26.1–2
9	210	29.0%	3.6%		2.3 (1.4–3.7)	0.44 (0.27–0.72)	0.001	0.077		2.2 (1.3–3.8)	0.45 (0.27–0.76)	0.003	0.107		2.6 (1.5–4.3)	0.39 (0.23–0.66)	<0.001	0.051	5.0E+0	MLLT3	p22.3–p21.3
	211	34.8%	3.4%		1.9 (1.4–2.7)	0.52 (0.37–0.72)	<0.001	0.053		2.0 (1.4–2.9)	0.50 (0.34–0.72)	<0.001	0.075		2.1 (1.5–3)	0.48 (0.33–0.68)	<0.001	0.016	3.0E−1		p21.3
	212	35.9%	3.3%		1.6 (1.2–2.2)	0.62 (0.45–0.85)	0.003	0.147		1.7 (1.2–2.4)	0.57 (0.41–0.81)	0.001	0.096		1.7 (1.2–2.4)	0.59 (0.42–0.83)	0.002	0.100	6.0E−1		p21.3
	213	38.7%	3.2%		1.5 (1.1–2)	0.67 (0.5–0.88)	0.004	0.164							1.6 (1.2–2.1)	0.64 (0.47–0.86)	0.003	0.123	6.0E−2		p21.3
	214	40.2%	3.0%		1.5 (1.2–1.9)	0.66 (0.52–0.84)	<0.001	0.077							1.5 (1.2–2.0)	0.65 (0.50–0.84)	0.001	0.066	3.0E−1	CDKN2A, CDKN2B	p21.3
	215	36.3%	3.3%		1.7 (1.2–2.2)	0.61 (0.45–0.82)	0.001	0.077		1.7 (1.2–2.4)	0.59 (0.42–0.81)	0.001	0.096		1.7 (1.3–2.4)	0.58 (0.42–0.80)	<0.001	0.058	2.0E+0		p21.3
	217	35.0%	3.9%		1.7 (1.3–2.4)	0.58 (0.42–0.79)	<0.001	0.077							1.9 (1.4–2.7)	0.52 (0.38–0.73)	<0.001	0.027	9.0E−3		p21.3
	218	32.8%	4.5%		1.7 (1.2–2.3)	0.60 (0.43–0.83)	0.002	0.136		1.7 (1.2–2.4)	0.59 (0.41–0.85)	0.004	0.133		1.8 (1.3–2.6)	0.54 (0.38–0.77)	<0.001	0.051	5.0E−1		p21.3
	219	30.5%	4.6%		2.0 (1.3–3.0)	0.51 (0.34–0.76)	0.001	0.077		2.0 (1.3–3.1)	0.49 (0.32–0.77)	0.002	0.096		2.1 (1.4–3.3)	0.47 (0.3–0.73)	<0.001	0.051	2.0E+0		p21.3–2
	220	30.1%	4.8%												1.6 (1.2–2.2)	0.63 (0.46–0.86)	0.004	0.133	2.0E−2		p21.2
	222	27.8%	5.2%							2.1 (1.3–3.3)	0.48 (0.30–0.77)	0.002	0.096						2.0E+0		p21.1
	223	26.0%	6.4%		1.9 (1.2–2.9)	0.53 (0.35–0.82)	0.004	0.164		2.1 (1.3–3.3)	0.48 (0.3–0.77)	0.002	0.096						7.0E−1		p21.1
19	383	8.0%	2.5%		2.5 (1.3–4.6)	0.41 (0.22–0.76)	0.005	0.164		2.8 (1.4–5.3)	0.36 (0.19–0.7)	0.003	0.103						4.0E+0		p12–p11
	386	4.6%	7.3%												2.2 (1.3–3.9)	0.45 (0.26–0.79)	0.005	0.161	3.0E−1		q11

Results from a model adjuster by treatment arm, patient age, sex, performance status (PS), histology, T, and N stage. *, hazard ratio for a 2-fold lower copy number; **, hazard ratio for a 2-fold higher copy number.

Table S8. Prognostic effect of the copy number of genomic regions. Multivariate results in optimal quality samples only.

Chr	Region ID	CNA frequency			Disease-free survival				Overall survival					Lung-cancer specific survival			Mb	Genes	cytoBands
		Losses	Gains		HR for loss* (95% CI)	HR for gain** (95% CI)	P		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q		HR for loss* (95% CI)	HR for gain** (95% CI)	P
3	72	1.8%	44.8%		2.4 (1.6–3.5)	0.42 (0.28–0.63)	<0.001		2.8 (1.8–4.4)	0.35 (0.23–0.55)	<0.001			2.5 (1.6–3.8)	0.40 (0.26–0.62)	<0.001	5.0E+0	MECOM	q26.1–2
9	211	34.8%	3.4%		2.4 (1.7–3.3)	0.43 (0.30–0.60)	<0.001		2.8 (1.9–4.1)	0.36 (0.24–0.53)	<0.001						3.0E−1		p21.3
12	277	2.4%	14.9%											0.47 (0.27–0.81)	2.1 (1.2–3.7)	0.007	3.0E+0		p11.21–1
17	354	2.7%	10.3%		0.45 (0.23–0.86)	2.2 (1.2–4.3)	0.016		0.41 (0.21–0.82)	2.4 (1.2–4.9)	0.012						2.0E−2		q21.32
19	396	4.6%	10.5%						0.64 (0.46–0.88)	1.6 (1.1–2.2)	0.006						2.0E−2		q13.32

Results from a model adjuster by treatment arm, patient age, sex, performance status (PS), histology, T, and N stage. *, hazard ratio for a 2-fold lower copy number; **, hazard ratio for a 2-fold higher copy number.

Table S9. Predictive effect of the copy number of genomic regions. Univariate results in optimal quality samples only.

Chr	Region ID	CNA frequency			Disease-free survival					Overall survival					Lung-cancer specific survival				Mb	Genes	cytoBands
		Losses	Gains		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q		HR for loss* (95% CI)	HR for gain** (95% CI)	P	Q
5	102	0.6%	45.3%							0.45 (0.22–0.88)	2.2 (1.1–4.4)	0.021	0.705						5.0E−2		p15.33
9	232	17.2%	2.4%		4.0 (1.2–14)	0.25 (0.07–0.85)	0.026	0.979											2.0E+1		q21.11–q22.1
	233	15.6%	3.2%		4.1 (1.2–14)	0.25 (0.07–0.84)	0.025	0.979											5.0E+0	FAM120A	q22.1–31
10	238	7.2%	4.6%							0.22 (0.06–0.82)	4.6 (1.2–17)	0.024	0.705						4.0E+0	RET, RASSF4	q11.21–22
14	318	8.9%	10.0%		0.29 (0.11–0.75)	3.4 (1.3–8.8)	0.010	0.979							0.22 (0.08–0.62)	4.5 (1.6–13)	0.004	0.614	2.0E−2		q32.33
	319	10.8%	11.4%		0.37 (0.17–0.81)	2.7 (1.2–6.0)	0.013	0.979											1.0E−1		q32.33
	325	16.3%	8.3%		0.66 (0.45–0.95)	1.5 (1.0–2.2)	0.028	0.979											3.0E−2		q32.33
	299	13.5%	9.4%							0.33 (0.13–0.83)	3.0 (1.2–7.6)	0.019	0.705						5.0E−1		q11.2
	302	6.4%	9.5%							0.29 (0.10–0.84)	3.4 (1.2–9.8)	0.022	0.705						4.0E+0		q11.2–q12
17	360	2.1%	14.2%							0.23 (0.07–0.82)	4.3 (1.2–15)	0.023	0.705						8.0E−2		q25.3
20	408	0.8%	20.8%												0.19 (0.06–0.63)	5.3 (1.6–18)	0.007	0.614	5.0E−1		q11.21
	409	0.8%	20.2%												0.19 (0.06–0.65)	5.3 (1.5–18)	0.008	0.614	1.0E−1		q11.21
	411	3.0%	17.4%												0.18 (0.05–0.60)	5.7 (1.7–19)	0.006	0.614	7.0E+0	SRC, MAFB, RBL1, MAFB	q11.22–q12
	412	3.1%	17.5%												0.21 (0.07–0.68)	4.7 (1.5–15)	0.009	0.614	3.0E−2		q12

Results from a model adjuster by treatment arm, patient age, sex, performance status (PS), histology, T, and N stage. *, hazard ratio for a 2-fold lower copy number; **, hazard ratio for a 2-fold higher copy number.

Table S10. Predictive effect of the copy number of genomic regions. Multivariate results in optimal quality samples only.

Chr	Region ID	CNA frequency			Disease-free survival				Overall survival				Lung-cancer specific survival			Mb	cytoBands
		Losses	Gains		HR for loss* (95% CI)	HR for gain** (95% CI)	P		HR for loss* (95% CI)	HR for gain** (95% CI)	P		HR for loss* (95% CI)	HR for gain** (95% CI)	P
14	319	10.8%	11.4%						0.25 (0.10–0.62)	4.0 (1.6–10)	0.003					1.0E−1	q32.33
18	368	12.1%	6.6%		2.6 (1.1–6.1)	0.39 (0.16–0.91)	0.029									1.0E−3	q12.1
20	407	1.3%	18.4%										0.19 (0.05– 0.77)	5.2 (1.3–21)	0.020	5.0E−2	q11.21

Results from a model adjuster by treatment arm, patient age, sex, performance status (PS), histology, T, and N stage. *, hazard ratio for a 2-fold lower copy number; **, hazard ratio for a 2-fold higher copy number.

Table S11. Genomic regions with differential prognostic effect according to the histologic subtype.

Chr	Region ID	Disease free survival					Overall survival					Lung cancer specific survival				Mb	Genes	cytoBands
		HR for ADC (95% CI)	HR for SCC (95% CI)	P inter	Q inter		HR for ADC (95% CI)	HR for SCC (95% CI)	P inter	Q inter		HR for ADC (95% CI)	HR for SCC (95% CI)	P inter	Q inter
1	7						0.50 (0.19–1.3)	2.4 (1.0–5.6)	0.003	0.252						5.0E−1		p13.3
	16											1.6 (0.93–2.9)	0.68 (0.47–0.99)	0.004	0.213	2.0E−2		q23.3
	18	2.8 (1.3–5.8)	0.44 (0.18–1.1)	0.003	0.169							3.7 (1.7–8.1)	0.55 (0.20–1.5)	0.005	0.213	4.0E+1	FCGR2B, PBX1, TPR, LHX4, CDC73	q23.3–q31.3
	20	2.3 (1.1–4.7)	0.53 (0.27–1.0)	0.004	0.169											6.0E−2		q31.3
4	84						0.20 (0.07–0.57)	2.1 (0.9–4.8)	0.001	0.184						1.0E+1	PTTG2	p15.1–p12
	96						0.51 (0.28–0.92)	1.6 (0.97–2.6)	0.005	0.252						5.0E−3		q27
7	144	1.6 (1.0–2.4)	0.63 (0.40–0.97)	0.002	0.169							1.6 (1.0–2.5)	0.55 (0.33–0.94)	0.002	0.213	5.0E−1	EGFR	p11.2
	145	2.3 (1.3–4.0)	0.63 (0.33–1.2)	0.001	0.169							2.5 (1.4–4.6)	0.55 (0.26–1.1)	0.001	0.212	2.0E+0		p11.2
	146	2.6 (1.4–4.9)	0.59 (0.28–1.2)	<0.001	0.169							3.1 (1.6–5.9)	0.56 (0.24–1.3)	0.001	0.212	1.0E−2		p11.1
	147	2.5 (1.3–4.9)	0.62 (0.29–1.4)	0.004	0.169											8.0E−1		q11.1–21
11	251	1.3 (1.0–1.6)	0.78 (0.62–0.98)	0.002	0.169		1.4 (1.1–1.7)	0.72 (0.56–0.93)	<0.001	0.056						1.0E−2		p14.1
20	407	1.8 (0.79–4.3)	0.33 (0.15–0.74)	0.005	0.169											5.0E−2		q11.21

Results from a model adjuster by treatment arm, patient age, sex, performance status (PS), histology, T, and N stage. All the hazard ratios (HR) are for a 2-fold higher copy number.

Chromosomal instability

The number of BPs was heterogeneous across trials, higher for men and possibly for high performance status (Table S12). Patients with a very high number of BPs (≥314) had shorter DFS than patients with very few (≤109) [HR =1.5 (1.1–2.0), P=0.015). This result was weaker for OS [HR =1.2 (0.90–1.7), P=0.19), but stronger for LCSS [HR =1.7 (1.2–2.3), P=0.003). Flexible models (Figure S4) in all patients showed that the BP effect can be considered log-linear. Such an effect was HR =1.1 (0.99–1.2, P=0.084, Table 4) on DFS for a patient as compared to another having a two times fewer BPs; this log-linear effect was similar LCSS [HR =1.1 (1.0–1.3), P=0.036) and statistically not significant on OS [HR =1.0 (0.93–1.2), P=0.51). The treatment effect was independent of the number of BPs both when comparing extreme groups (HR range: 0.96 to 1.1, P range: 0.78 to 0.93) and in terms of log-linear effects (HR: 0.93 to 0.99, P: 0.53 to 0.93).

Table S12. Chromosomal instability.

Factor	nBP ratio	LCI	UCI	uP value	mP value
Trial
CALGB	1.0			0.001	<0.001
IALT	1.0	0.89	1.2
JBR 10	1.3	1.1	1.5
Age
≤55	0.96	0.86	1.1	0.631	0.327
55–64	1.0
≥65	1.0	0.93	1.2
Arm
Control	1.00			0.512	0.999
Chemotherapy	1.00	0.92	1.1
Sex
Woman	1.0			0.013	0.009
Men	1.1	0.99	1.2
PS
0	1.0			0.002	0.047
1–2	1.1	0.99	1.2
Surgery
Lobectomy/other	1.0			0.016	0.163
Pneumonectomy	1.0	0.94	1.2
Histology
Adenocarcinoma	1.0			<0.001	0.194
Squamous cell carcinoma	1.1	0.99	1.2
Other	1.1	0.93	1.2
T stage
T1	1.1	0.93	1.2	0.117	0.470
T2	1.0
T3/T4	1.1	0.92	1.2
N stage
N0	0.99	0.89	1.1	0.050	0.775
N1	1.0
N2	1.0	0.90	1.2

nBP ratio, the ratio between the expected number of breakpoints (BPs) as compared to the reference class; LCI and UCI, lower and upper bounds of the 95% confidence interval; uP value, P value in the univariate analyses (Kruskal-Wallis test); mP value, P value in the multivariate analysis (likelihood ratio test).

Figure S4.

Flexible model (splines) to account for the possibly non-linear effect of the number of breakpoints (BPs) on the patient outcomes in all patients (prognostic effects, left) and within each arm (predictive effect, right). The two vertical lines are the tertiles of the number of BPs.

Table 4. Association between chromosomal instability and patient outcomes.

Variables	HR (95% CI)	P value
Prognostic effect of the number of BPs
DFS
Comparison between extreme classes*	1.5 (1.1–2.0)	0.015
Comparison on all the sample range**	1.1 (0.99–1.2)	0.084
OS
Comparison between extreme classes*	1.2 (0.90–1.7)	0.19
Comparison on all the sample range**	1.0 (0.93–1.2)	0.51
LCSS
Comparison between extreme classes*	1.7 (1.2–2.3)	0.003
Comparison on all the sample range**	1.1 (1.0–1.3)	0.036
Predictive effect of the number of BPs
DFS
Comparison between extreme classes*	0.96 (0.54–1.7)	0.91
Comparison on all the sample range**	0.95 (0.54–1.7)	0.62
OS
Comparison between extreme classes*	0.97 (0.52–1.8)	0.93
Comparison on all the sample range**	0.93 (0.76–1.2)	0.53
LCSS
Comparison between extreme classes*	1.1 (0.57–2.1)	0.77
Comparison on all the sample range**	0.99 (0.80–1.2)	0.93

*, HR between the high-number-of-breakpoints group (≥314, N=200) and the low-number-of-breakpoints group (≤109, N=197); **, Log2-linear effect, i.e., the HR is the ratio of the risk of a patient with a given number of breakpoints (BPs) as compared to a patient with a 2-fold lower number of BPs. DFS, disease-free survival; OS, overall survival; LCSS, lung-cancer specific survival; HR, hazard ratio.

Clonality

Patients with 2+ clones (N=518) had shorter DFS and LCSS [HR =1.2 (1.0–1.4 and 1.0–1.3), Table S13] than patients with 0–1 clones (N=456). This result was statistically non-significant (P=0.054 and 0.051, respectively) notably for OS [HR =1.1 (0.88–1.3), P=0.48]. The treatment effect was not associated to clonality [P=0.63 (DFS), 0.47 (OS), 0.52 (LCSS)].

Table S13. Association between clonality and patient outcomes.

Variables	HR	LCI	UCI	P value
Prognostic effect of the number of BPs
Disease-free survival (DFS)
Stratified	1.2	0.99	1.4	0.063
Stratified + adjusted	1.2	1.0	1.4	0.054
Overall survival (OS)
Stratified	1.1	0.90	1.3	0.38
Stratified + adjusted	1.1	0.88	1.3	0.48
Lung-cancer specific survival (LCSS)
Stratified	1.2	0.99	1.45	0.068
Stratified + adjusted	1.2	1.0	1.5	0.051
Predictive effect of the number of BPs
Disease-free survival (DFS)
Stratified	1.2	0.81	1.6	0.42
Stratified + adjusted	1.1	0.76	1.6	0.63
Overall survival (OS)
Stratified	1.2	0.82	1.7	0.35
Stratified + adjusted	1.2	0.79	1.7	0.47
Lung-cancer specific survival (LCSS)
Stratified	1.3	0.86	1.9	0.25
Stratified + adjusted	1.1	0.77	1.7	0.52

HR, hazard ratio between the patients with 2 or more clones (N=518) and patients with 0 or 1 clones (N=456); LCI and UCI: lower and upper bounds of the 95% confidence interval; stratified, model stratified on the trial; adjusted, model adjusted on clinicopathological factors.

Discussion

Increased understanding of the genomic changes of NSCLC facilitates the identification of prognostic and predictive biomarkers and provides vital information for personalized therapy, potentially allowing tailored treatments for individual patients. We utilized NSCLC FFPE samples from the LACE-Bio project to profile DNA CNAs. The most frequent CN gains were found on 1p13, 1q21, 3q22–26, 5p13–15, 6p24, and 22q11, the most frequent losses on 3p21.31, 8p23, and 9p21.3. The more focal and less frequent losses might be due to harder identification of losses in tumors with stromal cell contamination. Telomerase reverse transcriptase (TERT), among the most frequently amplified genes, is the catalytic subunit of the enzyme telomerase; its overexpression has been associated with poor prognosis (31). Among the loss genes, cyclin-dependent kinase Inhibitor 2A (CDKN2A), reported to be deleted in many tumors including lung cancer (32), codes for two proteins, p16 (or p16^INK4a) and p14^arf, which act as tumor suppressors by regulating the cell cycle.

The different spectrum of CNAs between ADC and SCC has been reported previously (33,34). Genes such as PIK3CA (33) and PDGFB (35) were amplified in lung SCC. Cyclin L (CCNL1) has been identified as oncogene in head and neck cancer (36). Mutations in CHEK2 (37) and NF2 (38) have been reported to be associated with SCC. CN loss and promoter hypermethylation of RASSF1 was reported in SCCHN (39) and in early stage NSCLC (40). NKX2-1 amplification was significantly less frequent than in ADC (33).

Our analyses confirmed some of the prognostic genes reported in the literature, such as shorter survival with CN loss of CDKN2A/B (32). In the present study, CDKN2A/B CN loss occurred in 40% of the cases and was significantly associated with shorter DFS. CDKN2A/B CN loss was also prognostic in ADC. Copy number loss of the tumor suppressor STK11 (or LKB1) has been associated with increased risk of brain metastasis (41). We were not able to confirm this due to incomplete reporting of metastatic sites. NSCLC patients with STK11 exon 1 or 2 mutations have shorter survival (42). A recent meta-analysis (14 studies, 1915 patients with solid tumors) revealed that decreased expression of STK11 was a prognostic factor [HR =2.2 (1.5–3.2), P<0.001] (43). In the present study, STK11 CN loss was found in 11% of samples and was significantly associated with shorter DFS [HR =2.4 (1.3–4.3), P=0.005]. We also identified novel prognostic genes, such as FSTL3, which encodes a secreted glycoprotein, and transcriptional factors MLLT1, SH3GL1, and TCF3, and the guanine nucleotide exchange factor (GEF) gene VAV1. Its overexpression significantly increased the risk of death [HR =1.81 (1.39–2.36), P<0.001) (44). However, in the present study, the CN loss frequency of the region containing these genes was 9%. Additional studies on their prognostic value are warranted.

The LACE-bio study has the unique possibility to identify biomarkers that predict efficacy of ACT in NSCLC by comparison to observation arms. Three regions had significant differences in multivariate analyses between the two study arms, but they came with high false discovery rate (Table S6). Particularly, 8p23.3–2 losses were significantly associated with increased ACT efficacy for OS. The frequent gains of 20q11.21 strongly were associated with no benefit from ACT for LCSS. This deleterious effect from ACT was in strong contrast with the small group (0.8%) of patients with 20q11.21 loss where ACT lead to a notably high survival benefit. The 20q11.21 region is rich in genes that might have a potential role in cancer such as HCK (tyrosine kinase), BCL2L1 (apoptotic regulator), MAPRE1 and TPX2 (microtubule associated factors), DNMT3B (epigenetic modifier) and transcriptional regulators. It is even more striking to find the p53 and DNA damage-regulated gene named PDRG1 in 20q11.21. PDRG1 is an oncogene in lung cancer cell lines, is selectively regulated by DNA damaging agents such as UV, and promotes radioresistance (45,46). Whatsoever, its exact role in mediating resistance to ACT in NSCLC remains to be confirmed. Finally, the 14q32.33 region also had differential HR (loss predictive of ACT efficacy, gain predictive of inefficacy), but the proportion of patients with losses and gains were equally high (10.8% and 11.4% respectively), making interpretation more difficult in the context of prediction of ACT efficacy.

In exploratory analyses in the LACE-Bio2 samples, the prognosis of patients with very high chromosomal instability was significantly worse than for patients with very low, independently of the clinical factors. Chromosomal instability could likely be associated with the risk of relapses rather than to death. We found no association with the magnitude of the ACT effect.

The LACE-Bio data and tissue bank provided a valuable source for studying the prognostic and predictive role of the CN of genomic regions in stage I–III NSCLC. These large-scale genome-wide analyses were consistent with previous results and provide new candidate prognostic markers. Furthermore, as the data come from randomized controlled trials, we propose new markers which could predict the effect of ACT.

Detailed bioinformatics and statistical methods

Bioinformatics pre-processing

The CGH data were normalized relative to an internal pool of 390 reference normal tissues, segmented using circular binary segmentation (CBS) (17,18), and minimal recurrent regions were identified via the CGHregions algorithm (19). We planned to discard regions with <20 CNAs. The proportion of probes on the X-chromosome with called allelic imbalance allowed inferring patient gender. The inferred gender was compared to the actual gender and inconsistent samples were discarded.

Endpoints

The primary endpoint was:

• ❖ Disease-free survival (DFS), defined as the time from randomization to first recurrence (loco-regional or distant) or death from any cause.

Secondary endpoints were:

• ❖ Overall survival (OS), defined as the time from randomization to death from any cause, and;

• ❖ Lung-cancer specific survival (LCSS), defined as the time from randomization to death from lung cancer. Death without evidence of cancer relapse was treated as censoring for LCSS.

Statistical analyses

The called copy number (CN) of each region was correlated to survival endpoints via Cox models stratified by trial and adjusted for treatment arm, patient age, sex, performance status, histology, type of surgery, T, and N stage. The CN entered in the regression models as log₂(CN). Thus, the estimated hazard ratio (HR_gain) expresses the relative hazard for a 2-fold higher CN of a given region. Its reciprocal (HR_loss = 1/HR_gain) is the relative hazard for a 2-fold lower CN. To evaluate the predictive role of CNAs, a treatment-by-log₂(CN) interaction was further added. In both the prognostic and the predictive models, the Cox model was stratified by trial and adjusted for treatment arm and clinical variables.

We performed both univariate (each region separately) and multivariate analyses (several regions jointly). The P values were corrected to control the false discovery rate [Q values (20)]. Multivariate models were built by stepwise selection (αin =0.10 and αout =0.01) and using a penalized regression approach (21,22) with lasso penalty for prognostic analyses and adaptive lasso for predictive analyses.

Preplanned sensitivity analyses were

The analyses were repeated, in addition to the entire study population, within the following subgroups:

• ❖ Histological subtypes (ADC vs. SCC);

• ❖ Optimal quality subgroup (MAPD ≤0.3).

The significance of the CN differences between histologic subtypes was assessed by t-tests; the P values were corrected via step- down multiple testing procedures (23,24). We compared the obtained results to those from TCGA (25,26). Known tumor suppressor genes and oncogenes were obtained from previously published results (27).

Chromosomal instability

The number of breakpoints (BPs) in the CN was used as measure of chromosomal instability. Its association with clinicopathological factors was first tested in univariate analyses (Kruskal-Wallis tests), then in a multivariate analysis using a log-linear quasi-Poisson model. Its association with outcomes and treatment effect was studied in Cox models comparing the 20% of patients with the highest number of BPs (≥314) to the 20% of patients with the lowest (≤109).

Software

The bioinformatics pre-processing and the statistical analyses were performed using R software v3.3, with the following packages: biospear, CGHbase, CGHcall, CGHregions, DNAcopy, glmnet, gplots, parallel, qvalue, scales, survival, TxDb.Hsapiens.UCSC.hg19.knownGene, XLConnect.

Ethical Statement: The study was approved by institutional review boards (No. UHN 04-0333-T).