Purpose:TP53 mutations in early-stage non–small cell lung cancer (NSCLC) may be associated with worse survival but their prognostic role in advanced NSCLC is controversial. In addition, it remains unclear whether mutated patients represent a clinically homogeneous group.

Experimental Design: We retrospectively examined TP53 mutations and outcome in a training cohort of 318 patients with stage IIIB–IV NSCLC: 125 epidermal growth factor receptor (EGFR) wild-type (wt) and 193 EGFR mutated (mut). An independent validation cohort of 64 EGFR-mut patients was subsequently analyzed. Mutations were classified as “disruptive” and “nondisruptive” according to their predicted degree of disturbance of the p53 protein structure and function.

Results: In the training cohort, TP53 mutations were found in 43 of the 125 EGFR-wt patients (34.4%). Of these, 28 had nondisruptive TP53 mutations and a median overall survival (OS) of 8.5 months, compared with 15.6 months for the remaining 97 patients (P = 0.003). In the EGFR-mut group, TP53 mutations were found in 50 of the 193 patients (25.9%). The OS for the 26 patients with TP53 nondisruptive mutations was 17.8 months versus 28.4 months for the remaining 167 patients (P = 0.04). In the validation cohort, the 11 patients with nondisruptive TP53 mutations had a median OS of 18.1 months compared with 37.8 months for the 53 remaining patients (P = 0.006). In multivariate analyses, nondisruptive TP53 mutations had an independent, significant association with a shorter OS.

Conclusions: Nondisruptive mutations in the TP53 gene are an independent prognostic factor of shorter survival in advanced NSCLC. Clin Cancer Res; 20(17); 4647–59. ©2014 AACR.

See related commentary by Govindan and Weber, p. 4419

Translational Relevance

The majority of patients diagnosed with non–small cell lung cancer (NSCLC) present with advanced disease stage and have extremely poor prognosis. The tumor suppressor gene TP53 is the most frequently mutated in NSCLC, but its prognostic role in advanced NSCLC is controversial, and it remains unclear whether mutated patients represent a clinically homogeneous group. This study shows that a specific group of mutations in the TP53 gene, namely nondisruptive mutations, are an independent prognostic factor of shorter survival in advanced NSCLC. Our results pave the way for more widespread testing of TP53 mutations in unresectable NSCLC and indicate that the disruptive/nondisruptive categorization is clinically relevant and should be applied in all studies of TP53 in lung cancer. Clinical trials are warranted to determine whether patients with advanced NSCLC with nondisruptive mutations can benefit from drugs that reactivate p53.

The majority of patients diagnosed with non–small cell lung cancer (NSCLC) present with advanced disease stage and have extremely poor prognosis (1). The tumor suppressor gene TP53 is the most frequently mutated in NSCLC (2). It encodes a 393-aa protein with three distinct domains (Supplementary Fig. S1). The transactivation domain is the target of posttranscriptional modification by proteins such as the serine-protein kinase ATM or the E3 ubiquitin-protein ligase Mdm2 (3). The DNA-binding domain (DBD), encoded by exons 5 to 8 of the TP53 gene, comprises residues 102–292 and recognizes a consensus sequence in the promoter of several genes involved in DNA repair, cell-cycle arrest, or apoptosis (CDKN1A, SFN, FAS, BAX, DDB2, and others; refs. 4–6). The sequence-specific transcriptional activity mediated by the DBD is the primary mechanism accounting for the tumor suppressor activities of p53. Within the DBD, loops L2 (residues 163–195) and L3 (residues 236–251) bind to a Zn atom and play a key role in the interaction with DNA (7). Finally, the C-terminal domains are responsible for oligomerization and negative regulation of the protein. Under normal conditions, the p53 protein binds to the ubiquitine ligase Mdm2 and is quickly degraded. Some cellular stresses (DNA damage, cell-cycle abnormalities, and hypoxia) block Mdm2 binding to p53 which then oligomerizes to an active tetramer (2).

Mutations in the TP53 gene are present in 32.5% of NSCLC tumors (8). Many of these are mutations in the DBD that lead to a stable protein, with significant loss of activity. The mutated protein accumulates in the nucleus of cells and exerts a dominant-negative effect by heterodimerizing with the (wild-type) wt p53 expressed from the remaining allele (9). Recent evidence also indicates that, in addition to abrogating tumor-suppressor properties, some mutations confer new “gain of function” (GOF) activities to the mutated p53 protein that contribute to tumor progression (10, 11). Mechanisms of mutated p53 GOF include interference with p53-related proteins (p63, p73, etc; ref. 12) and aberrant upregulation of genes that promote cancer progression or drug resistance (13).

A variety of criteria have been used to categorize TP53 mutations. Poeta and colleagues (14) proposed a classification in “disruptive” and “nondisruptive” according to their functional effects on the p53 protein. Disruptive mutations include (i) any mutation that originates a stop codon (nonsense, frameshift, and intronic), (ii) missense mutations located inside the L2 or L3 loops replacing one residue by another of a different polarity or charge and, (iii), in-frame deletions within the L2 or L3 loops. Nondisruptive mutations are all those not classified as disruptive and include (i) missense mutations and in-frame deletions located outside the L2-L3 loops and (ii) missense mutations within the L2-L3 loops but replacing one residue with another of the same polarity or charge. Disruptive mutations likely lead to a complete, or almost complete, loss of activity of the p53 protein. In contrast, nondisruptive mutations can retain some of the functional properties of wt-p53, and more importantly, experimental evidence shows that they often associate with GOF activities (15, 16).

Determination of TP53 mutations by standard techniques requires an amount of material that is often only available in resected tumors. For this reason, most studies of TP53 status in NSCLC have been performed in stages I to III. The small numbers of patients included in some of these studies, the differences in follow-up, and the various criteria used to classify TP53 mutations have led to contradictory results. This last aspect is particularly important as the heterogeneity of TP53 mutations has been demonstrated to correlate with similarly heterogeneous clinical outcomes in other types of tumor (14, 17, 18).

We have analyzed TP53 mutations in a training cohort of 318 patients with advanced NSCLC [125 EGFR-wt and 193 EGFR-mutated (mut)] and correlated TP53 mutations with clinical parameters. We validated our findings in an independent cohort of 64 EGFR-mut patients.

Patients

In the training cohort, we retrospectively analyzed a total of 318 patients with advanced NSCLC, divided into EGFR-wt and EGFR-mut groups. All patients were unresectable, stage IIIB, or IV NSCLC (sixth TNM staging system) with histologic tissue and clinical data available. Samples were obtained at the time of diagnosis of metastatic disease in all cases. Written informed consent was provided by all patients. Approval was obtained from each hospital's institutional review board and ethics committee. The EGFR-wt group consisted of 125 tumor samples from patients diagnosed between 2006 and 2012 in six different European hospitals. We selected all those patients with enough tumor tissue available and complete follow-up who also fulfilled the following criteria (i) stage IIIB–IV NSCLC, (ii) no surgery, (iii) first-line platinum-based chemotherapy, (iv) EGFR-wt. The EGFR-mut group was composed of 193 patients from the Spanish Lung Adenocarcinoma Database (19) and the EURTAC clinical trial (20). Since several molecular biology studies had already been performed on the biopsies from these patients (21), many of them had insufficient tumor sample remaining. We therefore selected all the patients from these trials who had sufficient tumor tissue.

The validation cohort consisted of 64 tumor samples from EGFR-mut patients diagnosed between 2006 and 2012 in two different hospitals in Spain and Colombia. All were stage IIIB–IV, nonsurgical patients with complete follow-up.

Processing and dissection of tumor samples

All samples were formalin-fixed, paraffin-embedded pretreatment tumor tissues, stained with haematoxilin/eosin, and evaluated by an expert pathologist. The small percentage of specimens with more than 60% tumor infiltration in an area greater than 2.2 mm2 were macrodissected; the rest of the samples were microdissected as previously described (21). In both cases, cells were dissected directly into 30 μL of PCR buffer (Ecogen) plus proteinase K (40 μg/mL) and incubated overnight at 60°C. Proteinase was inactivated by incubation at 100°C for 10 minutes and the resulting cell extract used for mutational analyses. Previous validation experiments had demonstrated that our genetic techniques can detect a mutation in TP53, KRAS, BRAF, and PIK3CA in samples containing as little as 10% tumor cells.

Analysis of TP53 mutations

PCR followed by sequencing to determine mutations when they are distributed throughout a gene is an expensive, time- and sample-consuming procedure. Therefore, we set up and validated a high-resolution melting (HRM) technique to quickly and efficiently screen for mutations in exons 5 to 8 of TP53.

Cell extracts (1–2 μL) were mixed with 10 μL of MeltDoctor HRM Master Mix (Applied Biosystems), primers (0.3 μmol/L final, Supplementary Table S1) and, only in the case of exon 5b, MgCl2 (1 mmol/L additional). DNAs from cell lines with known TP53 status were used as controls (Supplementary Table S2). All cell lines were purchased from the American Type Culture Collection (ATCC) and passaged for fewer than 6 months after resuscitation, except for the PC9 cell line, which was kindly provided by F. Hoffman-La Roche Ltd. ATCC performs cell line authentication by short tandem repeat profiling. In addition, all the cell lines were validated in our laboratory by repeatedly genotyping them for EGFR, KRAS, PIK3CA, BRAF, TP53, and CTNNB1 genes. The genotyping was also performed in the same purified DNA samples that were used in our experiments. In all cases, the genotypes of the cells exactly matched those described in the COSMIC database.

HRM amplification reactions were carried out for 50 cycles (Supplementary Table S3). The melt analysis was performed with the 7900HT Fast Real-Time PCR System (Applied Biosystems). Conditions were 95°C for 1 second, 72°C for 90 seconds, and an HRM step rising from 72°C to 95°C at 0.1°C per second. The dissociation curves were analyzed using Applied Biosystems HRM Software v2.0 and the HRM products of all samples that seemed mutated or unclear were sequenced. Finally, all mutated samples were reconfirmed by standard PCR followed by Sanger sequencing (Supplementary Tables S1 and S3).

To address the issue of heterogeneity, we analyzed at least two separate tumor areas in a total of 74 of our samples (42 in the training and 32 in the validation cohort). The genotype of the different regions of the same sample was identical in all cases, indicating that the distribution of TP53 mutations within the tumor was homogeneous.

Description of the mutations was based on the Universal Mutation Database for TP53 (http://p53.fr/; ref. 13) and obtained using the MUT-TP53 2.0 Excel spreadsheet tool (22), which was also used to perform comparison of our dataset with previous publications using the mean and 95% confidence interval (CI) of p53 activity as measured by transactivation with the WAF1 promoter. In the validation cohort, the mean of our study was −1.27 for EGFR-wt and −1.34 for EGFR-mut, compared with −1.26 and −1.29 in the two comprehensive NSCLC studies reported in the tool.

Classification of TP53 mutations

Mutations were classified as “disruptive” and “nondisruptive” according to Poeta and colleagues (14), with the only modification being that, for our analyses, glycine was considered to be a nonpolar side chain residue.

Detection of hotspot KRAS, BRAF, and PIK3CA mutations

Controls used for detection of KRAS, BRAF, and PIK3CA hotspot mutations are presented in Supplementary Table S2, primers and probes in Supplementary Table S4, and amplification conditions in Supplementary Table S5. Mutations in codons 12 and 13 of KRAS were examined by HRM analysis and reconfirmed by standard sequencing, as described for TP53 mutations. Mutations V600E in BRAF and E542K, E545K and H1047R in PIK3CA were screened by an allelic discrimination assay. Amplification was carried out in 12.5-μL volumes using from 1 μL of DNA sample, 6.25 μL of Taqman Genotyping Master Mix (Applied Biosystems), 0.6 pmol of each primer, and 0.2 pmol of probes. All mutated samples were confirmed by standard PCR plus sequencing.

Statistical analyses

For patients from the EURTAC clinical trial, progression-free survival (PFS) and overall survival (OS) were calculated from time of randomization; maximum time between randomization and start of treatment was 15 days. For the remaining patients, PFS and OS were calculated from the start of treatment. Median PFS and OS were estimated using the Kaplan–Meier method and distribution curves were compared with a two-sided log-rank test. Cox regression univariate analysis was used to generate survival HRs. The effect of TP53 nondisruptive mutations on OS was assessed by correlating all the studied covariates with interaction terms. χ2 test or Fisher exact test were used when response was compared with prognostic factors. In the multivariate analysis of the training cohort, we included all standard covariates, although KRAS mutations and squamous histology were almost exclusively present in the EGFR-wt group. All statistical analyses were performed using the IBM Statistical Package for Social Science (SPSS) for Windows version 19. Level of significance was set bilaterally at 0.05.

Patient characteristics and TP53 status (training cohort)

Table 1 summarizes the characteristics of the 318 patients analyzed in the training cohort. Median follow-up was 14.36 months; 20.00 months for those still alive at the conclusion of the study. One hundred and twenty-five patients were EGFR-wt, and 193 were EGFR-mut. Mutations in the EGFR gene were significantly associated with female gender, never smoking status, older age, histology, and Eastern Cooperative Oncology Group Performance Status (ECOG PS) ≥1 (Table 1). More than 90% of EGFR-mut tumors were adenocarcinomas, whereas 27% of the EGFR-wt patients had squamous cell carcinomas. The percentage of samples carrying TP53 mutations was higher in the EGFR-wt than in the EGFR-mut group (34.4% vs. 25.9%) but the difference was not significant (P = 0.13). All EGFR-wt patients received platinum-based chemotherapy, whereas 146 EGFR-mut patients received erlotinib and 47 chemotherapy (Supplementary Table S6). Response to first-line therapy was significantly better in the EGFR-mut group than in the EGFR-wt group (P = 0.02). Finally, 72 EGFR-wt and at least 90 EGFR-mut patients received two or more lines of therapy (Supplementary Table S6).

Table 1.

Characteristics of the patients, according to EGFR status

Training cohortValidation cohort
CharacteristicEGFR wt (n = 125) (%)EGFR mut (n = 193) (%)P valueaEGFR mut (n = 64) (%)
Median age (range) 62 (29–82) 66 (29–86) 0.003 61 (21–87) 
Gender   <0.001  
 Male 93 (74.4) 47 (24.4)  22 (34.4) 
 Female 32 (25.6) 146 (75.6)  42 (65.6) 
Smoking status   <0.001  
 Never smokers 17 (14.3) 141 (73.1)  29 (72.5) 
 Current and former smokers 102 (85.7) 52 (26.9)  11 (27.5) 
 Unknown  24 
Histology     
 Squamous 34 (27.2) 1 (0.5) <0.001b 2 (3.1) 
 Nonsquamous 91 (72.8) 192 (99.5)  62 (96.9) 
  Large cell 16 (12.8) 3 (1.6)  0 (0) 
  Adenocarcinoma 68 (54.4) 177 (91.7)  59 (92.2) 
  Undifferentiated 6 (4.8) 6 (3.1)  1 (1.6) 
  Others 1 (0.8) 6 (3.1)  2 (3.1) 
Stage   0.67  
 IIIB 11 (8.8) 14 (7.3)  5 (7.8) 
 IV 114 (91.2) 179 (92.7)  59 (92.2) 
ECOG PS   0.004  
 0 18 (15.3) 57 (29.8)  9 (15.8) 
 ≥1 100 (84.7) 134 (70.2)  48 (84.2) 
 Unknown  
Brain metastases   0.47  
 Yes 16 (13.8) 21 (10.9)  21 (35.0) 
 No 100 (86.2) 172 (89.1)  39 (65.0) 
 Unknown  
Bone metastases   0.09  
 Yes 40 (34.5) 48 (24.9)  26 (41.3) 
 No 76 (65.5) 145 (75.1)  37 (58.7) 
 Unknown  
Response   0.17  
 CR 6 (5.4) 10 (6.0)  8 (13.1) 
 PR 45 (40.2) 90 (53.9)  36 (59.0) 
 SD 44 (39.3) 50 (29.9)  16 (26.2) 
 PD 17 (15.2) 17 (10.2)  1 (1.6) 
 NE 13 26  
 Responders (CR+PR) 51 (45.5) 100 (59.9) 0.02 44 (72.1) 
 Non responders (SD+PD) 61 (54.5) 67 (40.1)  17 (27.9) 
TP53 mutated vs. wt   0.13  
 P53 mutated (D+NoD) 43 (34.4) 50 (25.9)  17 (26.6) 
 wt 82 (56.6) 143 (74.1)  47 (73.4) 
KRAS status (wt vs. mutated)   <0.001  
 wt 94 (75.2) 144 (99.3)  NA 
 Mutated 31 (24.8) 1 (0.7)   
 Not determined 48   
KRAS mutation   <0.001  
 wt 94 (75.2) 144 (99.3)  NA 
 G12C 15 (12.0) 1 (0.7)   
 Other mutations 16 (12.8) 0 (0.0)   
Training cohortValidation cohort
CharacteristicEGFR wt (n = 125) (%)EGFR mut (n = 193) (%)P valueaEGFR mut (n = 64) (%)
Median age (range) 62 (29–82) 66 (29–86) 0.003 61 (21–87) 
Gender   <0.001  
 Male 93 (74.4) 47 (24.4)  22 (34.4) 
 Female 32 (25.6) 146 (75.6)  42 (65.6) 
Smoking status   <0.001  
 Never smokers 17 (14.3) 141 (73.1)  29 (72.5) 
 Current and former smokers 102 (85.7) 52 (26.9)  11 (27.5) 
 Unknown  24 
Histology     
 Squamous 34 (27.2) 1 (0.5) <0.001b 2 (3.1) 
 Nonsquamous 91 (72.8) 192 (99.5)  62 (96.9) 
  Large cell 16 (12.8) 3 (1.6)  0 (0) 
  Adenocarcinoma 68 (54.4) 177 (91.7)  59 (92.2) 
  Undifferentiated 6 (4.8) 6 (3.1)  1 (1.6) 
  Others 1 (0.8) 6 (3.1)  2 (3.1) 
Stage   0.67  
 IIIB 11 (8.8) 14 (7.3)  5 (7.8) 
 IV 114 (91.2) 179 (92.7)  59 (92.2) 
ECOG PS   0.004  
 0 18 (15.3) 57 (29.8)  9 (15.8) 
 ≥1 100 (84.7) 134 (70.2)  48 (84.2) 
 Unknown  
Brain metastases   0.47  
 Yes 16 (13.8) 21 (10.9)  21 (35.0) 
 No 100 (86.2) 172 (89.1)  39 (65.0) 
 Unknown  
Bone metastases   0.09  
 Yes 40 (34.5) 48 (24.9)  26 (41.3) 
 No 76 (65.5) 145 (75.1)  37 (58.7) 
 Unknown  
Response   0.17  
 CR 6 (5.4) 10 (6.0)  8 (13.1) 
 PR 45 (40.2) 90 (53.9)  36 (59.0) 
 SD 44 (39.3) 50 (29.9)  16 (26.2) 
 PD 17 (15.2) 17 (10.2)  1 (1.6) 
 NE 13 26  
 Responders (CR+PR) 51 (45.5) 100 (59.9) 0.02 44 (72.1) 
 Non responders (SD+PD) 61 (54.5) 67 (40.1)  17 (27.9) 
TP53 mutated vs. wt   0.13  
 P53 mutated (D+NoD) 43 (34.4) 50 (25.9)  17 (26.6) 
 wt 82 (56.6) 143 (74.1)  47 (73.4) 
KRAS status (wt vs. mutated)   <0.001  
 wt 94 (75.2) 144 (99.3)  NA 
 Mutated 31 (24.8) 1 (0.7)   
 Not determined 48   
KRAS mutation   <0.001  
 wt 94 (75.2) 144 (99.3)  NA 
 G12C 15 (12.0) 1 (0.7)   
 Other mutations 16 (12.8) 0 (0.0)   

NOTE: KRAS mutations were not determined in the validation cohort, composed entirely of EGFR-mut patients.

Abbreviations: CR, complete response; D, disruptive; ECOG PS, Eastern Cooperative Oncology Group Performance Status; NA, not analyzed; NE, not evaluable; NoD, nondisruptive; PD, progressive disease; PR, partial response; SD, stable disease.

aP values calculated excluding the “unknown” or “not determined” samples.

bP value calculated for squamous versus nonsquamous.

Mutations in the TP53 gene were not significantly associated with gender, histology, stage (IIIB or IV), PS or presence of brain, and bone metastases in either the EGFR-wt or the EGFR-mut group (Supplementary Table S7). About smoking history, the frequency of TP53 mutations was significantly higher in former or current smokers only in the EGFR-wt group. The different types of TP53 mutations detected in our study and their distribution are shown in Table 2, Supplementary Table S8, and Supplementary Figs. S1 and S2. Remarkably, there were differences in the types of TP53 mutations between the two groups of patients. Frameshift and in-frame deletions were exclusively found in EGFR-mut patients, where they accounted for 18% of the total number of mutations. Transversions (replacing a purine by a pyrimidine or vice versa) were predominant in EGFR-wt patients, whereas transitions (that change a purine by a purine or a pyrimidine by a pyrimidine) made up the majority of point mutations in EGFR-mut patients. Tobacco smoking is known to induce transversions (23), and the percentage of former and current smokers was significantly higher in the EGFR-wt group than in the EGFR-mut group (85.7% vs. 26.9%, P < 0.001).

Table 2.

Summary of TP53 mutations identified in our study, classified according to their functional effects

Training cohortValidation cohort
Type of mutationEGFR wt (n = 43)EGFR mutated (n = 50)EGFR mutated (n = 17)
Disruptive 15 (35%) 24 (48%) 6 (35%) 
 Frameshift deletions 5 (10%) 2 (12%) 
 In-frame deletions within L2-L3 3 (6%) 
 Nonsense 8 (19%) 1 (2%) 3 (17%) 
 Missense within L2-L3a 7 (16%) 13 (26%) 1 (6%) 
 Intronic 2 (4%) 
Nondisruptive 28 (65%) 26 (52%) 11 (65%) 
 In-frame deletions outside L2-L3 1 (2%) 
 Missenseb 28 (65%) 25 (50%) 11 (65%) 
Training cohortValidation cohort
Type of mutationEGFR wt (n = 43)EGFR mutated (n = 50)EGFR mutated (n = 17)
Disruptive 15 (35%) 24 (48%) 6 (35%) 
 Frameshift deletions 5 (10%) 2 (12%) 
 In-frame deletions within L2-L3 3 (6%) 
 Nonsense 8 (19%) 1 (2%) 3 (17%) 
 Missense within L2-L3a 7 (16%) 13 (26%) 1 (6%) 
 Intronic 2 (4%) 
Nondisruptive 28 (65%) 26 (52%) 11 (65%) 
 In-frame deletions outside L2-L3 1 (2%) 
 Missenseb 28 (65%) 25 (50%) 11 (65%) 

aReplacing a residue with another of a different polarity or charge.

bOutside the L2-L3 loops or, if within L2-L3, replacing a residue with another of the same polarity or charge.

Two recent reports have shown that KRAS mutations are associated with shorter survival in chemotherapy-treated advanced NSCLC (24, 25). Therefore, samples were analyzed for KRAS mutations (exons 12–13). Where material was still available, BRAF and PIK3CA hotspots were also tested (Supplementary Fig. S3). As expected, only one of the 32 KRAS-mutated patients had a concomitant EGFR-mut (Supplementary Table S7). Finally, mutations in the PIK3CA gene were infrequent and only one patient had a BRAF mutation (Supplementary Table S6).

Patient characteristics and TP53 status (validation cohort)

The characteristics of the 64 patients analyzed in the validation cohort are presented in Table 1. Median follow-up was 27.32 months for all patients and 29.32 months for those still alive at the conclusion of the study. All patients were EGFR-mut, and 92.2% had adenocarcinomas. The percentage of samples carrying TP53 mutations was similar to that observed among the EGFR-mut patients of the training cohort (26.6% vs. 25.9%). The profile of TP53 mutations was also comparable; 12% were deletions (Table 2) and 60% of point mutations were transitions (Supplementary Fig. S2). First-line treatment was erlotinib or gefitinib in 89% of the patients and 59.5% received at least two lines of therapy (Supplementary Table S6).

Survival and clinical characteristics, EGFR, and KRAS mutational status

For all 318 patients in the training cohort, survival was significantly associated with conventional prognostic factors (Table 3). OS was 26.5 months in the EGFR-mut group versus 12.8 months in the EGFR-wt group (P < 0.001, Supplementary Fig. S4). Within the EGFR-wt group, OS was significantly associated with gender, smoking history, histology, and presence of bone metastases. The G12C mutation correlated with a significantly worse outcome in this group, whereas the rest of the mutations in the KRAS gene did not (Supplementary Table S9). OS for the 15 patients with a G12C mutation was 7.6 months, compared with 14.4 months for patients with other genotypes (P = 0.03; Supplementary Fig. S5). The number of patients harboring PIK3CA or BRAF mutations was too low for reliable statistical analyses. Finally, in the EGFR-mut group, only an ECOG PS of 0 was significantly associated with a better OS (Supplementary Table S9).

Table 3.

Results of the univariate and multivariate analyses of selected factors for OS in the training cohort

Univariate analysisMultivariate analysis
VariableNHR for death (95% CI)P valueHR for death (95% CI)P value
EGFR status 
 Mutated 193   
 wt 125 1.96 (1.46–2.62) <0.001 1.22 (0.77–1.93) 0.41 
Age 315 0.99 (0.98–1.01) 0.69 1.01 (0.98–1.02) 0.84 
Gender 
 Female 178   
 Male 140 1.77 (1.32–2.36) <0.001 1.29 (0.85–1.94) 0.24 
Smoking status 
 Never smokers 17   
 Current + former smokers 102 1.78 (1.33–2.40) <0.001 1.18 (0.74–1.88) 0.48 
Histology 
 Nonsquamous 283   
 Squamous 35 2.49 (1.67–3.72) <0.001 2.32 (1.32–4.08) 0.003 
Stage 
 IIIB 25   
 IV 293 1.70 (0.95–3.05) 0.08 2.08 (0.75–5.78) 0.16 
ECOG PS 
 0 75   
 ≥1 234 1.83 (1.26–2.65) 0.001 2.03 (1.24–3.34) 0.005 
Brain metastases 
 No 272   
 Yes 37 1.46 (0.95–2.24) 0.09 1.56 (0.94–2.58) 0.09 
Bone metastases 
 No 221   
 Yes 88 1.80 (1.32–2.46) <0.001 2.25 (1.53–3.29) <0.001 
TP53 
 wt+ D 264   
 NoD 54 2.08 (1.46–2.97) <0.001 1.89 (1.09–3.26) 0.02 
KRAS status 
 wt+ others 254   
 G12C 16 2.19 (1.42–3.36) <0.001 2.10 (1.19–3.71) 0.01 
Univariate analysisMultivariate analysis
VariableNHR for death (95% CI)P valueHR for death (95% CI)P value
EGFR status 
 Mutated 193   
 wt 125 1.96 (1.46–2.62) <0.001 1.22 (0.77–1.93) 0.41 
Age 315 0.99 (0.98–1.01) 0.69 1.01 (0.98–1.02) 0.84 
Gender 
 Female 178   
 Male 140 1.77 (1.32–2.36) <0.001 1.29 (0.85–1.94) 0.24 
Smoking status 
 Never smokers 17   
 Current + former smokers 102 1.78 (1.33–2.40) <0.001 1.18 (0.74–1.88) 0.48 
Histology 
 Nonsquamous 283   
 Squamous 35 2.49 (1.67–3.72) <0.001 2.32 (1.32–4.08) 0.003 
Stage 
 IIIB 25   
 IV 293 1.70 (0.95–3.05) 0.08 2.08 (0.75–5.78) 0.16 
ECOG PS 
 0 75   
 ≥1 234 1.83 (1.26–2.65) 0.001 2.03 (1.24–3.34) 0.005 
Brain metastases 
 No 272   
 Yes 37 1.46 (0.95–2.24) 0.09 1.56 (0.94–2.58) 0.09 
Bone metastases 
 No 221   
 Yes 88 1.80 (1.32–2.46) <0.001 2.25 (1.53–3.29) <0.001 
TP53 
 wt+ D 264   
 NoD 54 2.08 (1.46–2.97) <0.001 1.89 (1.09–3.26) 0.02 
KRAS status 
 wt+ others 254   
 G12C 16 2.19 (1.42–3.36) <0.001 2.10 (1.19–3.71) 0.01 

Abbreviations: D, disruptive; ECOG PS, Eastern Cooperative Oncology Group Performance Status; NoD, nondisruptive.

Survival and TP53 mutations

In the training cohort, the 93 patients with TP53 mutations had a median OS of 17.5 months versus 22.9 months for the 225 TP53-wt patients (Supplementary Fig. S6), but the difference was not statistically significant (HR, 1.45; CI 95%, 0.95–2.22; P = 0.09). However, because different types of mutations in the TP53 gene are known to have different effects on the functionality of the protein, we classified TP53 mutations into disruptive and nondisruptive and found that patients with advanced NSCLC with these two types of TP53 mutation constitute distinct prognostic groups. In the training cohort, the median OS in patients with nondisruptive mutations was 13.3 months compared with 24.6 months in patients TP53-wt or carrying disruptive mutations (HR, 2.08; 95% CI, 1.46–2.97; P < 0.001; Fig. 1, Table 3). The association of nondisruptive mutations with shorter survival was maintained when the patients were divided according to EGFR status. In the EGFR-wt group, median OS was 8.5 months for patients with nondisruptive TP53 mutations versus 15.6 months for other patients (HR, 2.04; 95% CI, 1.27–3.26; P = 0.003). In the EGFR-mut group, OS was 17.8 months for patients with TP53 nondisruptive mutations compared with 28.4 months for the remaining patients (HR, 1.79; 95% CI, 1.02–3.13; P = 0.04; Fig. 2, Supplementary Table S9).

Figure 1.

Kaplan–Meier plots of OS among patients in the training and validation cohorts, according to TP53 mutation status. A, median survival for patients in the training cohort with nondisruptive mutations (green; N = 54, of whom 40 died) was 13.3 months, compared with 26.2 months for patients with disruptive mutations (blue; N = 39, of whom 21 died) and 22.9 months for patients with wt TP53 (brown; N = 225, of whom 126 died; P < 0.001). B, median survival for patients in the training cohort with nondisruptive mutations was 13.3 months (green), whereas for the remaining patients it was 24.6 months (red; P < 0.001). C, median survival for patients in the validation cohort (all of them EGFR-mut) with nondisruptive mutations (n = 11, of whom 6 died) was 18.1 months, compared with 37.8 months for TP53-wt patients (n = 47, of whom 18 died) and not reached among patients with disruptive mutations (n = 6, of whom 2 died; P = 0.01). D, median survival for EGFR-mut patients in the validation cohort with nondisruptive mutations was 18.1 months, whereas it was 37.8 months for the remaining patients (P = 0.006).

Figure 1.

Kaplan–Meier plots of OS among patients in the training and validation cohorts, according to TP53 mutation status. A, median survival for patients in the training cohort with nondisruptive mutations (green; N = 54, of whom 40 died) was 13.3 months, compared with 26.2 months for patients with disruptive mutations (blue; N = 39, of whom 21 died) and 22.9 months for patients with wt TP53 (brown; N = 225, of whom 126 died; P < 0.001). B, median survival for patients in the training cohort with nondisruptive mutations was 13.3 months (green), whereas for the remaining patients it was 24.6 months (red; P < 0.001). C, median survival for patients in the validation cohort (all of them EGFR-mut) with nondisruptive mutations (n = 11, of whom 6 died) was 18.1 months, compared with 37.8 months for TP53-wt patients (n = 47, of whom 18 died) and not reached among patients with disruptive mutations (n = 6, of whom 2 died; P = 0.01). D, median survival for EGFR-mut patients in the validation cohort with nondisruptive mutations was 18.1 months, whereas it was 37.8 months for the remaining patients (P = 0.006).

Close modal
Figure 2.

Kaplan–Meier plots of OS for the EGFR-wt and EGFR-mut populations in the training cohort, according to TP53 mutation status. A, median survival for EGFR-wt patients in the training cohort with nondisruptive mutations (green; N = 28, of whom 25 died) was 8.5 months, compared with 22.6 months for patients with disruptive mutations (blue; N = 15, of whom 12 died) and 14.0 months for the patients with wt TP53 (brown; N = 82, of whom 56 died; P = 0.01). B, median survival for EGFR-wt patients in the training cohort with nondisruptive mutations was 8.5 months (green), compared with 15.6 months (red) for the rest of the patients (P = 0.003). C, median survival for EGFR-mut patients in the training cohort with nondisruptive mutations (n = 26, of whom 15 died) was 17.8 months, compared with 27.0 months for TP53-wt patients (n = 143, of whom 70 died) and 30.0 months among patients with disruptive mutations (n = 24, of whom 9 died; P = 0.061). D, median survival for EGFR-mut patients in the training cohort with nondisruptive mutations was 17.8 months, whereas it was 28.4 months for the rest of the patients (P = 0.04).

Figure 2.

Kaplan–Meier plots of OS for the EGFR-wt and EGFR-mut populations in the training cohort, according to TP53 mutation status. A, median survival for EGFR-wt patients in the training cohort with nondisruptive mutations (green; N = 28, of whom 25 died) was 8.5 months, compared with 22.6 months for patients with disruptive mutations (blue; N = 15, of whom 12 died) and 14.0 months for the patients with wt TP53 (brown; N = 82, of whom 56 died; P = 0.01). B, median survival for EGFR-wt patients in the training cohort with nondisruptive mutations was 8.5 months (green), compared with 15.6 months (red) for the rest of the patients (P = 0.003). C, median survival for EGFR-mut patients in the training cohort with nondisruptive mutations (n = 26, of whom 15 died) was 17.8 months, compared with 27.0 months for TP53-wt patients (n = 143, of whom 70 died) and 30.0 months among patients with disruptive mutations (n = 24, of whom 9 died; P = 0.061). D, median survival for EGFR-mut patients in the training cohort with nondisruptive mutations was 17.8 months, whereas it was 28.4 months for the rest of the patients (P = 0.04).

Close modal

Because we had found that the KRAS G12C mutation was associated with worse outcome in EGFR-wt patients, we excluded the 15 patients carrying this mutation from the analysis of the EGFR-wt population. For the remaining patients, OS was 9.0 months for those with nondisruptive mutations, compared with 22.6 months for those with disruptive mutations and 16.3 months for those TP53-wt (P = 0.005; Supplementary Fig. S7).

In the multivariate Cox proportional-hazard model (Table 3), the presence of a nondisruptive TP53 mutation was significantly associated with decreased OS (HR, 1.89; 95% CI, 1.09–3.26; P = 0.02) after adjustment for all covariates. The interaction terms of nondisruptive TP53 mutations with the rest of the covariates were not statistically significant. The presence of a nondisruptive TP53 mutation remained an independent prognostic factor for shorter OS in the EGFR-wt group (HR, 1.78; 95% CI, 1.03–3.07; P = 0.04). Squamous histology also emerged as an independent prognostic factor in this group. In the EGFR-mut patients, only ECOG PS ≥ 1 was associated with survival, but the presence of a nondisruptive TP53 mutation almost reached statistical significance (HR, 1.71; 95% CI, 0.97–3.02; P = 0.06; Supplementary Table S10).

The association of nondisruptive mutations with shorter survival was maintained in the validation cohort of stage IIIB–IV EGFR-mut patients, that had a median follow-up considerably longer than the training cohort (27.32 vs. 14.36 months). The OS for the 11 patients with TP53 nondisruptive mutations was 18.1 months versus 37.8 months for the 53 patients TP53-wt or carrying disruptive mutations (HR, 3.84; 95% CI, 1.46–10.11; P = 0.006; Fig. 1, Table 4). In the multivariate analysis, the presence of a nondisruptive TP53 mutation emerged as the only independent prognostic factor for OS (HR, 6.11; 95% CI, 1.43–26.09; P = 0.01) when either a backward of a forward stepwise selection variable procedure was carried out (Table 4).

Table 4.

Results of the univariate and multivariate analyses of selected factors for OS in the in the validation cohort

Univariate analysisMultivariate analysis
VariableNHR for death (95% CI)P valueHR for death (95% CI)P value
Age 63 0.98 (0.95–1.01) 0.24 0.99 (0.95–1.03) 0.60 
Gender 
 Female 42   
 Male 22 1.27 (0.56–2.89) 0.57 1.60 (0.57–4.46) 0.37 
Stage 
 IIIB   
 IV 59 2.14 (0.45–10.14) 0.34 0.66 (0.04–10.38) 0.76 
ECOG 
 0   
 ≥1 48 1.52 (0.44–5.30) 0.51 3.15 (0.60–16.73) 0.18 
Brain metastases 
 No 39   
 Yes 21 0.86 (0.35–2.11) 0.75 0.99 (0.35–2.81) 0.99 
Bone metastases 
 No 37   
 Yes 26 1.70 (0.74–3.88) 0.21 1.44 (0.46–4.53) 0.53 
TP53 
 wt+ D 53   
 NoD 11 3.84 (1.46–10.11) 0.006 6.11 (1.43–26.09) 0.01 
Univariate analysisMultivariate analysis
VariableNHR for death (95% CI)P valueHR for death (95% CI)P value
Age 63 0.98 (0.95–1.01) 0.24 0.99 (0.95–1.03) 0.60 
Gender 
 Female 42   
 Male 22 1.27 (0.56–2.89) 0.57 1.60 (0.57–4.46) 0.37 
Stage 
 IIIB   
 IV 59 2.14 (0.45–10.14) 0.34 0.66 (0.04–10.38) 0.76 
ECOG 
 0   
 ≥1 48 1.52 (0.44–5.30) 0.51 3.15 (0.60–16.73) 0.18 
Brain metastases 
 No 39   
 Yes 21 0.86 (0.35–2.11) 0.75 0.99 (0.35–2.81) 0.99 
Bone metastases 
 No 37   
 Yes 26 1.70 (0.74–3.88) 0.21 1.44 (0.46–4.53) 0.53 
TP53 
 wt+ D 53   
 NoD 11 3.84 (1.46–10.11) 0.006 6.11 (1.43–26.09) 0.01 

NOTE: All patients in the validation cohort were EGFR-mut. EGFR and KRAS status were not included in the analyses. Smoking history was also excluded due to the high number of patients with no data available.

Abbreviations: D, disruptive; NoD, nondisruptive.

Response, PFS, and TP53 mutations

Among all 318 patients in the training cohort, median PFS was 7.0 months in the TP53 nondisruptive group versus 8.3 months for the remaining patients (P = 0.05; Supplementary Fig. S8). However, no difference of PFS was observed when the patients were divided in the EGFR-wt and EGFR-mut groups (Supplementary Fig. S9). When patients were stratified by treatment type rather than EGFR mutation status, among the 172 patients that received erlotinib, PFS was 11.0 months for those harboring TP53 nondisruptive mutations versus 15.0 months for the remaining patients (P = 0.14). No difference in PFS was observed according to the type of TP53 mutation among the 146 patients treated with chemotherapy (Supplementary Fig. S10).

Finally, no association between nondisruptive TP53 mutations and response to therapy was observed. In the training cohort, 23 of the 54 patients (50%) with nondisruptive mutations had complete or partial responses to therapy, compared with 128 among the remaining 264 patients (55%; P = 0.63).

In the present study, we have analyzed TP53 mutations in a large cohort of patients with advanced, nonresectable NSCLC and we provide evidence of an association of TP53 status and OS. Our results demonstrate that nondisruptive TP53 mutations define a distinct prognostic group of patients with significantly shorter survival. In contrast, those patients with disruptive mutations showed a nonsignificant trend toward better OS, compared with TP53-wt patients. In the multivariate analysis, the HRs for nondisruptive TP53 mutations were similar to those obtained for other widely used markers of poor prognosis, such as the PS ≥ 1. These findings, which have been validated in an independent cohort, indicate that TP53 nondisruptive mutations could be a clinically useful prognostic marker in advanced NSCLC.

It is now universally accepted that EGFR mutations define two types of NSCLC with different biology, therapeutic options, and outcome (19). In the present study, we have found that patients with the G12C KRAS mutation also have significantly worse outcome. Although the issue of KRAS mutations in lung cancer is controversial, a recent study of 484 patients found that they are significantly associated with shorter survival in advanced NSCLC (OS of 7.7 for patients with the G12C mutation vs. 15.0 months for those KRAS wt; ref. 24). In the present study, the prognostic value of TP53 nondisruptive mutations was not dependent on EGFR or KRAS status, and was observed both in chemotherapy-treated, EGFR-wt patients as well as in EGFR-mut patients treated in first line with erlotinib or chemotherapy.

We found TP53 mutations in 34% (43/125) of the EGFR-wt patients, in accordance with the frequency described in the COSMIC database for NSCLC (8). The frequency of TP53 mutations dropped to 26% in the EGFR-mut patients. Patients with tobacco-associated lung cancer have a higher frequency of TP53 mutations than patients who never smoked (26). Tobacco-associated lung cancer is also characterized by a higher number of transversions in the TP53 gene, whereas a high percentage of transition mutations are found in never smokers. In our study, 63% of the TP53 point mutations in the EGFR-wt group were transversions, compared with 53% (training cohort) or 60% (validation cohort) of transitions in the EGFR-mut group. In addition, frameshift or in-frame deletions were found exclusively in EGFR-mut patients.

Although most studies on the prognostic role of p53 in human cancers have only dealt with wt versus mutated patients, recent reports have demonstrated the usefulness of categorizing TP53 mutations since different mutant proteins can have widely disparate biologic effects (14, 17, 18). Our study strongly supports this new approach: we found that only nondisruptive mutations were significantly associated with worse OS, whereas TP53 mutations as a whole did not significantly correlate with outcome.

A weak prognostic role for TP53 mutations in NSCLC has been suggested by two meta-analyses (27, 28), but the issue is still controversial. Reports are contradictory probably due to the fact that some investigators consider TP53 mutations as a whole, whereas others do categorize them, but use a wide range of criteria. Most authors have analyzed early-stage, surgically resected tumors where sufficient tissue is available for mutational testing by standard techniques. When TP53 mutations were uncategorized, some studies found no association with outcome (29) or only identified a trend, which was lost in the multivariate analysis (30). Other studies observed an association between TP53 mutations and response to adjuvant therapy (31) or with shorter OS, either alone (32; only in stage I disease) or in combination with other molecular markers (33). In studies of early-stage NSCLC, a prognostic value for particular types of TP53 mutations is usually found, although results are contradictory. Truncated, but not missense, mutations have been associated with shorter PFS or OS in some studies (34, 35), whereas others have reached the opposite conclusion (36). Finally, some authors have reported an association between shorter OS and the presence of particular types of TP53 mutations, such as “severe flexible and contact” mutants (37), “truncated, structural, and DNA contact mutations” (38), or mutations in particular exons and codons (39, 40). In the case of advanced NSCLC, a significant percentage of patients cannot be biopsied and, in those where a biopsy is feasible, the amount of tumor tissue obtained is often scarce and can be easily consumed by routine testing. To the best of our knowledge, only two reports in stage IIIB–IV NSCLC have analyzed the clinical relevance of TP53 mutations, but both studies included a limited number of patients and did not classify the mutations. One of these studies found an association of TP53 mutations with shorter median OS in 70 cases of advanced NSCLC (4 vs. 9 months; ref. 41), whereas the other reported no association in a population of 88 patients (42). Taken together, all these discordant reports highlight the importance of establishing a widely accepted, clinically relevant classification of TP53 mutations that can allow the comparison of different studies. Our results indicate that the disruptive/nondisruptive categorization can fulfill this criterion.

It remains to be explained why nondisruptive mutations, that likely result in a p53 protein which can retain some functionality, are associated with shorter OS in advanced NSCLC, whereas disruptive mutations are not. Nondisruptive mutations could be more frequent in smokers whose tumors develop a high genetic instability (29) and other, as yet unknown, genetic alterations may be responsible for the worse outcomes. This would make our findings spurious, positioning nondisruptive TP53 mutations as a bystander of those alterations. However, several lines of evidence suggest that this is not the case and that nondisruptive TP53 mutations are responsible for the poor outcome of the patients.

First, nondisruptive mutations were predictive of shorter survival in the EGFR-mut patients, both in the training and in the validation cohorts. More than 70% of these patients were never smokers, whose tumors have a lower genetic instability and fewer mutations than tobacco-associated lung cancer. In addition, in contrast with the EGFR-wt group, the frequency of TP53 mutations among the EGFR-mut patients did not depend on the smoking status (Supplementary Table S7). Furthermore, nondisruptive mutations represented 65% of the TP53 mutations in the validation cohort, composed entirely of EGFR-mut patients, a majority of whom are never smokers.

Second, p53 is a “master protein” that regulates multiple cell processes. It plays a key role in tumorigenesis and TP53 mutations are well-known drivers in many types of human cancer. Experimental evidence shows that many nondisruptive mutations, rather than causing simple loss of function of wt p53, induce GOF activities that can be exerted through direct transcriptional regulation or through inactivation of p63/p73 (12). These GOF activities are dominant over the TP53-wt allele and lead to increased tumorigenicity, growth rate, motility, metastasis and invasiveness, and decreased chemosensitivity in cell models (43). However, these phenotypes do not always appear together and different p53 mutants have been demonstrated to have different, complex patterns of GOF. At least 11 of the nondisruptive mutations found in our study (Supplementary Table S8) have been shown to induce GOF activities in cell models (Supplementary Table S11), including hotspot mutations such as R175H or R273H. Examples of these GOF activities are the downregulation of apoptotic (FAS) and cell-arrest (CDKN1A) genes and the upregulation of immortalizing (TERT), mitogenic (EGR1, MYC), stress-protective (HSPA1A), angiogenic (ANGPT1), or drug resistance (ABCB1, AXL) genes. Overexpression of the Axl tyrosine-kinase receptor (43), associated with the epithelial-to-mesenchimal transition, leads to resistance to tyrosine-kinase inhibitors.

The tumorigenic GOF activities can explain the worse survival observed in patients carrying nondisruptive TP53 mutations. In contrast, disruptive mutations are less likely to acquire GOF activity, which can explain why these mutations do not affect survival in advanced NSCLC. In addition, many of them are missense or frameshift mutations that lead to a truncated p53 protein that might be unable to form tetramers and thus to exert a dominant-negative effect (44). A wide-range analysis of mRNA and miRNA can determine what particular sets of genes show an altered expression in patients with different types of TP53 mutations and can further refine the prognostic significance of such mutations.

The number of published studies that have used the disruptive/nondisruptive categorization of TP53 mutations is very limited and all of them have been performed in surgically resected head and neck squamous cell carcinoma (HNSCC). In the first report proposing this classification, Poeta and colleagues (14) found that nondisruptive mutations were associated with shorter OS (3.9 vs. 5.4 years for the TP53-wt patients), although the difference did not reach statistical significance. In contrast, patients with disruptive mutations showed a significantly shorter OS (2.0 years). These results were confirmed in an independent cohort of patients (45). TP53 disruptive mutations also led to treatment failure through locoregional recurrence in HNSCC (46). In contrast with these studies, we found no association of TP53 disruptive mutations with a shorter OS in advanced NSCLC. Several reasons might help to explain this discrepancy. First, TP53 mutations can function differently in different cell contexts. For instance, the R175H nondisruptive mutant upregulated the transcription of the human telomerase reverse transcriptase gene in osteosarcoma cells, whereas it had no effect on a lung adenocarcinoma cell line (13). Second, we analyzed stage IIIB–IV, nonresectable lung cancer, while the patients in the HNSCC studies were all surgically resected. Several biomarkers are known to have opposing prognostic values in early and advanced tumors, such as the DNA repair gene ERCC1. Overexpression of ERCC1 was associated with adverse prognosis in advanced-stage NSCLC (47) but correlated with longer OS in early-stage, surgically resected patients (48). High levels of ERCC1 expression indicate an efficient DNA repair system that can prevent the appearance of new genetic alterations that promote invasiveness and metastases, thus explaining its association with better outcome in early-stage, resected tumors. In contrast, in advanced, nonresectable NSCLC, where the standard treatment is platinum-based chemotherapy, a proficient DNA repair capacity within the tumor can eliminate the adducts generated by platinum and therefore correlate with a poor prognosis. Similar mechanisms may be operating in disruptive mutations in TP53, which can drastically reduce the DNA repair capacity of tumor cells (49).

Although our study shows a prognostic impact of nondisruptive TP53 mutations in advanced NSCLC, it also indicates that they have no predictive value. We found no significant association of any type of TP53 mutations with duration or type of response, although there was a trend toward shorter PFS in erlotinib-treated patients carrying nondisruptive mutations. We lack an adequate explanation for this finding. However, we can hypothesize that most GOF activities of the nondisruptive TP53 mutations in our study could promote an aggressive behavior of the tumor after progression rather than induce a quicker resistance to drugs. In transformed cell models, resistance to DNA-damaging chemotherapeutic agents is not necessarily associated with increased proliferating or metastatic capacity, at least in some GOF p53 mutants (50).

In conclusion, we have demonstrated that nondisruptive mutations in TP53 are an independent prognostic factor of shorter survival in advanced NSCLC. Clinical trials are warranted to determine whether patients with this type of mutation would benefit from drugs that reactivate mutant p53.

B. Massuti reports receiving speakers bureau honoraria from and is a consultant/advisory board member for Roche. No potential conflicts of interest were disclosed by the other authors.

Conception and design: M.A. Molina-Vila, J. Bertran-Alamillo, A. Gascó, B. Massuti, R. Rosell

Development of methodology: M.A. Molina-Vila, J. Bertran-Alamillo, C. Mayo-de-las-Casas, R. Rosell

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M.A. Molina-Vila, A. Gascó, L. Pujantell-Pastor, L. Bonanno, A.G. Favaretto, M. Majem, B. Massuti, T. Morán, E. Carcereny, S. Viteri, R. Rosell

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M.A. Molina-Vila, J. Bertran-Alamillo, M. Sánchez-Ronco, B. Massuti, R. Rosell

Writing, review, and/or revision of the manuscript: M.A. Molina-Vila, J. Bertran-Alamillo, A. Gascó, C. Mayo-de-las-Casas, M. Sánchez-Ronco, L. Bonanno, A.G. Favaretto, A. Vergnenègre, M. Majem, B. Massuti, T. Morán, S. Viteri, R. Rosell

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A.G. Favaretto, R. Rosell

Study supervision: M.A. Molina-Vila, R. Rosell

The authors thank Niki Karachaliou, Ana Drozdowskyj, and José Javier Sánchez for critical review of the article and Kate Williams and Renée O'Brate for revision of English.

This work was partially supported by a grant from “Redes Tematicas de Investigacion en Cancer” (RD12/0036/0072), Spanish Ministry of Health.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
D'Addario
G
,
Fruh
M
,
Reck
M
,
Baumann
P
,
Klepetko
W
,
Felip
E
. 
Metastatic non-small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up
.
Ann Oncol
2010
;
21
Suppl 5
:
v116
9
.
2.
Goh
AM
,
Coffill
CR
,
Lane
DP
. 
The role of mutant p53 in human cancer
.
J Pathol
2011
;
223
:
116
26
.
3.
Xu
Y
. 
Regulation of p53 responses by post-translational modifications
.
Cell Death Differ
2003
;
10
:
400
3
.
4.
Levine
AJ
. 
p53, the cellular gatekeeper for growth and division
.
Cell
1997
;
88
:
323
31
.
5.
Jin
S
,
Levine
AJ
. 
The p53 functional circuit
.
J Cell Sci
2001
;
114
:
4139
40
.
6.
Riley
T
,
Sontag
E
,
Chen
P
,
Levine
A
. 
Transcriptional control of human p53-regulated genes
.
Nat Rev Mol Cell Biol
2008
;
9
:
402
12
.
7.
Cho
Y
,
Gorina
S
,
Jeffrey
PD
,
Pavletich
NP
. 
Crystal structure of a p53 tumor suppressor-DNA complex: understanding tumorigenic mutations
.
Science
1994
;
265
:
346
55
.
8.
Forbes
SA
,
Bindal
N
,
Bamford
S
,
Cole
C
,
Kok
CY
,
Beare
D
, et al
COSMIC: mining complete cancer genomes in the Catalogue of Somatic Mutations in Cancer
.
Nucleic Acids Res
2011
;
39
:
D945
50
.
9.
Lehmann
BD
,
Pietenpol
JA
. 
Targeting mutant p53 in human tumors
.
J Clin Oncol
2012
;
30
:
3648
50
.
10.
Xu
J
,
Reumers
J
,
Couceiro
JR
,
De Smet
F
,
Gallardo
R
,
Rudyak
S
, et al
Gain of function of mutant p53 by coaggregation with multiple tumor suppressors
.
Nat Chem Biol
2011
;
7
:
285
95
.
11.
Brosh
R
,
Rotter
V
. 
When mutants gain new powers: news from the mutant p53 field
.
Nat Rev Cancer
2009
;
9
:
701
13
.
12.
Oren
M
,
Rotter
V
. 
Mutant p53 gain-of-function in cancer
.
Cold Spring Harb Perspect Biol
2010
;
2
:
a001107
.
13.
Leroy
B
,
Fournier
JL
,
Ishioka
C
,
Monti
P
,
Inga
A
,
Fronza
G
, et al
The TP53 website: an integrative resource centre for the TP53 mutation database and TP53 mutant analysis
.
Nucleic Acids Res
2013
;
41
:
D962
9
.
14.
Poeta
ML
,
Manola
J
,
Goldwasser
MA
,
Forastiere
A
,
Benoit
N
,
Califano
JA
, et al
TP53 mutations and survival in squamous-cell carcinoma of the head and neck
.
N Engl J Med
2007
;
357
:
2552
61
.
15.
Scian
MJ
,
Stagliano
KE
,
Deb
D
,
Ellis
MA
,
Carchman
EH
,
Das
A
, et al
Tumor-derived p53 mutants induce oncogenesis by transactivating growth-promoting genes
.
Oncogene
2004
;
23
:
4430
43
.
16.
Vikhanskaya
F
,
Lee
MK
,
Mazzoletti
M
,
Broggini
M
,
Sabapathy
K
. 
Cancer-derived p53 mutants suppress p53-target gene expression–potential mechanism for gain of function of mutant p53
.
Nucleic Acids Res
2007
;
35
:
2093
104
.
17.
Trbusek
M
,
Smardova
J
,
Malcikova
J
,
Sebejova
L
,
Dobes
P
,
Svitakova
M
, et al
Missense mutations located in structural p53 DNA-binding motifs are associated with extremely poor survival in chronic lymphocytic leukemia
.
J Clin Oncol
2011
;
29
:
2703
8
.
18.
Vegran
F
,
Rebucci
M
,
Chevrier
S
,
Cadouot
M
,
Boidot
R
,
Lizard-Nacol
S
. 
Only missense mutations affecting the DNA binding domain of p53 influence outcomes in patients with breast carcinoma
.
PLoS ONE
2013
;
8
:
e55103
.
19.
Rosell
R
,
Moran
T
,
Queralt
C
,
Porta
R
,
Cardenal
F
,
Camps
C
, et al
Screening for epidermal growth factor receptor mutations in lung cancer
.
N Engl J Med
2009
;
361
:
958
67
.
20.
Rosell
R
,
Carcereny
E
,
Gervais
R
,
Vergnenegre
A
,
Massuti
B
,
Felip
E
, et al
Erlotinib versus standard chemotherapy as first-line treatment for European patients with advanced EGFR mutation-positive non-small-cell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial
.
Lancet Oncol
2012
;
13
:
239
46
.
21.
Rosell
R
,
Molina
MA
,
Costa
C
,
Simonetti
S
,
Gimenez-Capitan
A
,
Bertran-Alamillo
J
, et al
Pretreatment EGFR T790M mutation and BRCA1 mRNA expression in erlotinib-treated advanced non-small-cell lung cancer patients with EGFR mutations
.
Clin Cancer Res
2011
;
17
:
1160
8
.
22.
Soussi
T
,
Hamroun
D
,
Hjortsberg
L
,
Rubio-Nevado
JM
,
Fournier
JL
,
Beroud
C
. 
MUT-TP53 2.0: a novel versatile matrix for statistical analysis of TP53 mutations in human cancer
.
Hum Mutat
2010
;
31
:
1020
5
.
23.
Hainaut
P
,
Pfeifer
GP
. 
Patterns of p53 G–>T transversions in lung cancers reflect the primary mutagenic signature of DNA-damage by tobacco smoke
.
Carcinogenesis
2001
;
22
:
367
74
.
24.
Sun
JM
,
Hwang
DW
,
Ahn
JS
,
Ahn
MJ
,
Park
K
. 
Prognostic and predictive value of KRAS mutations in advanced non-small cell lung cancer
.
PLoS ONE
2013
;
8
:
e64816
.
25.
Johnson
ML
,
Sima
CS
,
Chaft
J
,
Paik
PK
,
Pao
W
,
Kris
MG
, et al
Association of KRAS and EGFR mutations with survival in patients with advanced lung adenocarcinomas
.
Cancer
2013
;
119
:
356
62
.
26.
Subramanian
J
,
Govindan
R
. 
Molecular genetics of lung cancer in people who have never smoked
.
Lancet Oncol
2008
;
9
:
676
82
.
27.
Szymanowska
A
,
Jassem
E
,
Dziadziuszko
R
,
Skrzypski
M
,
Kobierska-Gulida
G
,
Holm
K
, et al
Analysis of prognostic value of TP53 gene mutations in non-small cell lung cancer
.
Pneumonol Alergol Pol
2005
;
73
:
264
9
.
28.
Scoccianti
C
,
Vesin
A
,
Martel
G
,
Olivier
M
,
Brambilla
E
,
Timsit
JF
, et al
Prognostic value of TP53, KRAS and EGFR mutations in nonsmall cell lung cancer: the EUELC cohort
.
Eur Respir J
2012
;
40
:
177
84
.
29.
Ko
JL
,
Cheng
YW
,
Chang
SL
,
Su
JM
,
Chen
CY
,
Lee
H
. 
MDM2 mRNA expression is a favorable prognostic factor in non-small-cell lung cancer
.
Int J Cancer
2000
;
89
:
265
70
30.
Kosaka
T
,
Yatabe
Y
,
Onozato
R
,
Kuwano
H
,
Mitsudomi
T
. 
Prognostic implication of EGFR, KRAS, and TP53 gene mutations in a large cohort of Japanese patients with surgically treated lung adenocarcinoma
.
J Thorac Oncol
2009
;
4
:
22
9
.
31.
Tsao
MS
,
Aviel-Ronen
S
,
Ding
K
,
Lau
D
,
Liu
N
,
Sakurada
A
, et al
Prognostic and predictive importance of p53 and RAS for adjuvant chemotherapy in non small-cell lung cancer
.
J Clin Oncol
2007
;
25
:
5240
7
.
32.
Chien
WP
,
Wong
RH
,
Wu
TC
,
Cheng
YW
,
Chen
CY
,
Lee
H
. 
Potential increase in the prognostic value of p53 mutation by Pro72 allele in stage I non-small-cell lung cancer
.
Ann Surg Oncol
2009
;
16
:
1918
24
.
33.
Burke
L
,
Flieder
DB
,
Guinee
DG
,
Brambilla
E
,
Freedman
AN
,
Bennett
WP
, et al
Prognostic implications of molecular and immunohistochemical profiles of the Rb and p53 cell cycle regulatory pathways in primary non-small cell lung carcinoma
.
Clin Cancer Res
2005
;
11
:
232
41
.
34.
Hashimoto
T
,
Tokuchi
Y
,
Hayashi
M
,
Kobayashi
Y
,
Nishida
K
,
Hayashi
S
, et al
p53 null mutations undetected by immunohistochemical staining predict a poor outcome with early-stage non-small cell lung carcinomas
.
Cancer Res
1999
;
59
:
5572
7
.
35.
de Anta
JM
,
Jassem
E
,
Rosell
R
,
Martínez-Roca
M
,
Jassem
J
,
Martínez-López
E
, et al
TP53 mutational pattern in Spanish and Polish non-small cell lung cancer patients: null mutations are associated with poor prognosis
.
Oncogene
1997
;
15
:
2951
8
.
36.
Tomizawa
Y
,
Kohno
T
,
Fujita
T
,
Kiyama
M
,
Saito
R
,
Noguchi
M
, et al
Correlation between the status of the p53 gene and survival in patients with stage I non-small cell lung carcinoma
.
Oncogene
1999
;
18
:
1007
14
.
37.
Koga
T
,
Hashimoto
S
,
Sugio
K
,
Yoshino
I
,
Mojtahedzadeh
S
,
Matsuo
Y
, et al
Clinicopathological and molecular evidence indicating the independence of bronchioloalveolar components from other subtypes of human peripheral lung adenocarcinoma
.
Clin Cancer Res
2001
;
7
:
1730
8
.
38.
Ahrendt
SA
,
Hu
Y
,
Buta
M
,
McDermott
MP
,
Benoit
N
,
Yang
SC
, et al
p53 mutations and survival in stage I non-small-cell lung cancer: results of a prospective study
.
J Natl Cancer Inst
2003
;
95
:
961
70
.
39.
Skaug
V
,
Ryberg
D
,
Kure
EH
,
Arab
MO
,
Stangeland
L
,
Myking
AO
, et al
p53 mutations in defined structural and functional domains are related to poor clinical outcome in non-small cell lung cancer patients
.
Clin Cancer Res
2000
;
6
:
1031
7
.
40.
Huang
C
,
Taki
T
,
Adachi
M
,
Konishi
T
,
Higashiyama
M
,
Miyake
M
. 
Mutations in exon 7 and 8 of p53 as poor prognostic factors in patients with non-small cell lung cancer
.
Oncogene
1998
;
16
:
2469
77
.
41.
Murakami
I
,
Hiyama
K
,
Ishioka
S
,
Yamakido
M
,
Kasagi
F
,
Yokosaki
Y
. 
p53 gene mutations are associated with shortened survival in patients with advanced non-small cell lung cancer: an analysis of medically managed patients
.
Clin Cancer Res
2000
;
6
:
526
30
.
42.
Lim
EH
,
Zhang
SL
,
Li
JL
,
Yap
WS
,
Howe
TC
,
Tan
BP
, et al
Using whole genome amplification (WGA) of low-volume biopsies to assess the prognostic role of EGFR, KRAS, p53, and CMET mutations in advanced-stage non-small cell lung cancer (NSCLC)
.
J Thorac Oncol
2009
;
4
:
12
21
.
43.
Vaughan
CA
,
Singh
S
,
Windle
B
,
Yeudall
WA
,
Frum
R
,
Grossman
SR
, et al
Gain-of-function activity of mutant p53 in lung cancer through up-regulation of receptor protein tyrosine kinase Axl
.
Genes Cancer
2012
;
3
:
491
502
.
44.
Chan
WM
,
Siu
WY
,
Lau
A
,
Poon
RY
. 
How many mutant p53 molecules are needed to inactivate a tetramer?
Mol Cell Biol
2004
;
24
:
3536
51
.
45.
Lindenbergh-van der Plas
M
,
Brakenhoff
RH
,
Kuik
DJ
,
Buijze
M
,
Bloemena
E
,
Snijders
PJ
, et al
Prognostic significance of truncating TP53 mutations in head and neck squamous cell carcinoma
.
Clin Cancer Res
2011
;
17
:
3733
41
.
46.
Skinner
HD
,
Sandulache
VC
,
Ow
TJ
,
Meyn
RE
,
Yordy
JS
,
Beadle
BM
, et al
TP53 disruptive mutations lead to head and neck cancer treatment failure through inhibition of radiation-induced senescence
.
Clin Cancer Res
2012
;
18
:
290
300
.
47.
Roth
JA
,
Carlson
JJ
. 
Prognostic role of ERCC1 in advanced non-small-cell lung cancer: a systematic review and meta-analysis
.
Clin Lung Cancer
2011
;
12
:
393
401
.
48.
Pesta
M
,
Kulda
V
,
Fiala
O
,
Safranek
J
,
Topolcan
O
,
Krakorova
G
, et al
Prognostic significance of ERCC1, RRM1 and BRCA1 in surgically-treated patients with non-small cell lung cancer
.
Anticancer Res
2012
;
32
:
5003
10
.
49.
Hanel
W
,
Moll
UM
. 
Links between mutant p53 and genomic instability
.
J Cell Biochem
2012
;
113
:
433
9
.
50.
Bristow
RG
,
Peacock
J
,
Jang
A
,
Kim
J
,
Hill
RP
,
Benchimol
S
. 
Resistance to DNA-damaging agents is discordant from experimental metastatic capacity in MEF ras-transformants-expressing gain of function MTP53
.
Oncogene
2003
;
22
:
2960
6
.

Supplementary data