Abstract
Background
Prognostic variables are independently associated with survival and are fundamental to clinical trial design. We evaluated the impact of stage at initial diagnosis on overall survival in two independent, oncogene-defined cohorts.
Methods
All patients with EGFR-mutant and KRAS-mutant metastatic lung adenocarcinomas were identified through routine molecular testing from January 2005 to January 2011. Clinical characteristics were obtained. Overall survival from date of diagnosis of recurrent or de novo metastatic disease was estimated using the Kaplan-Meier method.
Results
635 patients with KRAS mutant and 496 patients with EGFR mutant metastatic lung adenocarcinomas were identified. Among patients with KRAS-mutant lung cancers, those with de novo metastatic disease have a shorter median overall survival compared to those with recurrent metastatic disease (13 vs 18 months, p=0.003). In a multivariable analysis of patients with KRAS mutant lung cancers, de novo metastatic disease at diagnosis (stage IV vs I–III, HR 1.5 [1.2–1.8], P<0.001) was independently associated with shorter overall survival. In patients with EGFR-mutant lung cancers, after controlling for age and performance status, de novo metastatic disease (IV vs I-III, HR 1.3 [1.0–1.7], P=0.03) was independently associated with shorter overall survival.
Conclusions
In KRAS-mutant lung adenocarcinomas, stage at diagnosis was associated with overall survival from diagnosis of recurrent/metastatic disease. In multivariable analyses, in both EGFR and KRAS mutant lung cancers, advanced stage at diagnosis was independently associated with shorter survival. Stage at diagnosis is a prognostic variable that should be accounted for in prospective studies in patients with metastatic lung cancers.
Background
Prognostic factors in lung cancers direct the clinical care of patients and set the framework for analyzing clinical research studies. Multiple prognostic variables have been identified in patients with advanced lung cancers, including performance status, age, sex, weight loss, LDH and hemoglobin [1–5]. We use these prognostic factors to make treatment recommendations and prospective clinical trials routinely record performance status, age and sex in order to interpret survival and treatment outcomes on study.
More recently, specific molecular alterations have been established to have both predictive and prognostic implications. The presence of an EGFR mutation predicts response to EGFR tyrosine kinase inhibitors (TKIs) [6] and patients with EGFR mutant lung cancers have a longer median overall survival from diagnosis of metastatic disease compared to patients with KRAS mutant lung cancers [9]. Differences in overall survival when comparing genotype-defined subgroups may reflect distinct tumor biology or may represent the differential benefit of targeted therapy in specific molecular subtypes.
Stage using the TNM staging system defines the appropriate therapeutic strategy and is the strongest predictor of survival in patients with lung cancers. Five year overall survival from initial diagnosis is closely associated with TNM stage [10]. Clinical trials of systemic therapies typically do not distinguish between those with metastatic disease at diagnosis (de novo metastatic disease) versus those with recurrent metastatic disease (i.e. those with a recurrent early stage lung cancer). It is unknown whether stage at initial diagnosis is a prognostic marker that impacts survival from time of diagnosis of metastatic disease. To examine whether stage at initial diagnosis is a prognostic variable, we reviewed a large population of patients identified at our institution with metastatic lung adenocarcinomas. In order to adjust for the potential prognostic impact of individual molecular alterations, we specifically reviewed patients with EGFR and KRAS mutant metastatic lung adenocarcinomas.
Methods
We identified all patients with EGFR-mutant and KRAS-mutant lung adenocarcinomas identified through routine molecular diagnostic testing performed between January 2005 and January 2011. Routine molecular testing including EGFR and KRAS began in 2004 at our institution; the patients included in our series had prospective molecular analysis performed. Data collection was approved by the institutional review board and privacy board. The list was manually reviewed to exclude patients that did not have metastatic or recurrent disease or a diagnosis of primary lung cancer. KRAS and EGFR mutation testing was done prior to this retrospective review by standard techniques [11–13].
Stage at diagnosis, age, gender, smoking history, and Karnofsky performance status (KPS) were obtained from the medical record. For each mutation specific group (EGFR and KRAS), overall survival (OS) from date of diagnosis of recurrent/metastatic disease was estimated using Kaplan-Meier method and compared across groups using log-rank test. Multivariable Cox proportional hazards models were used to test the associations between stage at initial cancer diagnosis and overall survival, controlling for factors associated with overall survival on the univariate analysis at a p value level of significance of 0.1. Patients alive at the end of the study period were censored at the time of the last available follow up.
Results
Between January 2005 and January 2011, 635 patients were identified with metastatic KRAS-mutant lung cancers and 496 patients were identified with metastatic EGFR-mutant lung cancers. Sixty one percent of patients with KRAS-mutant lung cancers and 74% of patients with EGFR-mutant lung cancers were diagnosed with de novo metastatic disease at diagnosis. The clinical characteristics of the patients are detailed in Table 1 and 2. Among the patients with KRAS-mutant lung adenocarcinoma, there were no differences among the patients with recurrent metastatic disease and de novo metastatic disease with respect to age, sex, smoking history, and performance status. Among patients with EGFR-mutant lung adenocarcinomas, there were similarly no differences in the clinical features between those with recurrent vs de novo metastatic cancer with the exception of smoking status. There were more never-smokers with EGFR-mutant lung cancer that were diagnosed with de novo metastatic cancer than with recurrent disease (p=0.01). Twenty-six percent (34/130) of the patients with recurrent EGFR-mutant lung cancer received adjuvant erlotinib or gefitinib for their early-stage disease.
Table 1.
Clinical characteristics of patients with KRAS–mutant lung adenocarcinomas: recurrent vs de novo metastatic disease
| Patient Characteristics |
Total N=635 |
Recurrent N=250 |
De Novo N=385 |
|
|---|---|---|---|---|
| Age at Diagnosis | ||||
| Median | 66 | 66 | 65 | |
| Range | 31–89 | 33–89 | 31–86 | |
| Sex (%) | ||||
| Men | 247 (39) | 86 (34) | 161 (42) | p=0.092 |
| Women | 388 (61) | 164 (66) | 224 (58) | |
| Smoking (%) | ||||
| Never-smoker | 44 (7) | 15 (6) | 29 (8) | p=0.41 |
| Smoker | 587 (92) | 233 (93) | 354 (92) | |
| KPS (%) | ||||
| KPS≥80% | 400 (63) | 161 (64) | 239 (62) | p=0.83 |
| KPS≤70% | 118 (19) | 49 (20) | 69 (118) | |
| Not recorded | 117 (18) | 40 (16) | 77 (20) |
Table 2.
Clinical characteristics of patients with EGFR–mutant lung adenocarcinomas: recurrent vs de novo metastatic disease
| Patient Characteristics |
Total N=496 |
Recurrent N=130 |
De Novo N=366 |
|
|---|---|---|---|---|
| Age at Diagnosis | ||||
| Mean | 63 | 68 | 61 | |
| Range | 26–89 | 35–89 | 26–88 | |
| Sex (%) | ||||
| Men | 170 (34) | 36 (28) | 134 (33) | p=0.068 |
| Women | 326 (66) | 94 (72) | 232 (63) | |
| Smoking (%) | ||||
| Never-smoker | 292 (59) | 63 (48) | 229 (68) | p=0.007 |
| Smoker | 204 (41) | 67 (52) | 137 (32) | |
| KPS (%) | ||||
| Median | 80 | 80 | 80 | |
| KPS≥80 | 401 (81) | 98 (75) | 303 (83) | p=0.071 |
| KPS≤70 | 95 (19) | 32 (25) | 63 (17) |
Patients with de novo metastatic KRAS-mutant lung cancers had shorter median overall survival from date of metastatic disease, compared to patients with recurrent metastatic disease (12 vs 18 months, p=0.002)(Figure 1). Among the patients with EGFR-mutant cancer, unadjusted overall survival was similar between patients with de novo vs recurrent metastatic disease (30 vs 32 months, p=0.14) (Figure 2).
Figure 1.
Overall survival from metastatic disease in patients with KRAS mutant lung cancers
Figure 2.
Overall Survival from metastatic disease in patients with EGFR mutant lung cancers
In patients with KRAS-mutant lung cancers, sex, performance status, and stage at initial diagnosis were associated with survival in univariate analysis (Table 3). In a multivariate analysis, only performance status and stage at initial diagnosis remained highly significant (Table 4). Patients with de novo metastatic KRAS-mutant cancers have shorter overall survival from diagnosis of metastatic disease (HR 1.41 95% CI 1.15–1.74, p=0.001). In univariate analysis of patients with EGFR-mutant lung cancers, age and performance status were associated with survival (Table 4). In a multivariate analysis controlling for age and performance status, de novo metastatic disease status was associated with shorter survival (HR 1.36 95% CI 1.05–1.76, p=0.019) (Table 5).
Table 3.
Univariate analysis of prognostic factors in metastatic KRAS mutant lung cancers
| Variable | Survival (yrs) |
HR (95% CI) | P value |
|---|---|---|---|
| Sex (male vs female) | 1.0 vs 1.3 | 1.25 (1.05–1.49) | 0.012 |
| Age | 1.04 (1.0–1.09) | 0.064 | |
| Smoking (former/current vs never) | 1.1 vs 1.6 | 1.29 (0.92–1.81) | 0.15 |
| KPS (≤70 vs ≥ 80) | 0.5 vs 1.3 | 2.33 (1.85-2.93) | <0.001 |
Table 4.
Multivariate analysis of prognostic factors in metastatic KRAS mutant lung cancers
| Variable | Hazard Ratio (95% CI) | P value |
|---|---|---|
| Recurrent vs De novo | 1.41 (95% CI 1.15–1.74) | 0.001 |
| Sex (male vs female) | 1.22 (95% CI 1.00–1.49) | 0.056 |
| Age | 1.00 (95%CI 0.95 – 1.05 | 0.99 |
| KPS (≤ 70 vs ≥ 80) | 2.28 (95% CI 1.79–2.89) | <0.001 |
Table 5.
Univariate analysis of prognostic factors in metastatic EGFR mutant lung cancers
| Variable | Survival (yrs) |
HR (95% CI) | P value |
|---|---|---|---|
| Sex (male vs female) | 2.6 vs 2.8 | 1.05 (0.84–1.32) | 0.68 |
| Age | 1.05 (1.00–1.10) | 0.037 | |
| Smoking (former/current vs never) | 2.8 vs 2.7 | 0.84 (0.67–1.04) | 0.11 |
| KPS (≤70 vs ≥ 80) | 1.9 vs 2.3 | 1.53 (1.18–1.99) | 0.001 |
Discussion
In our review of 1131 patients with EGFR and KRAS mutant lung adenocarcinomas, stage at initial diagnosis was a prognostic variable for overall survival from date of diagnosis of metastatic disease. Patients with de novo metastatic KRAS-mutant lung adenocarcinomas had a shorter overall survival from diagnosis of metastatic disease compared to patients with recurrent metastatic disease. This difference remained significant upon multivariate analysis. In patients with EGFR-mutant lung adenocarcinoma, when controlling for age and sex, de novo metastatic cancer was associated with shorter overall survival from diagnosis of metastatic disease compared to recurrent metastatic disease. As expected, poor performance status, older age and male sex were also predictors of shorter survival in both mutation-specific cohorts.
The underlying mechanism of the survival difference between those with recurrent vs de novo metastatic disease is unknown. Previously identified early stage disease could represent distinct, more indolent tumor biology. It is also possible that there are differences in the treatment patterns of patients with recurrent vs de novo metastatic lung cancers. In more than a third of cases, disease recurrence is localized to the chest and local therapy including radiation and surgery may be viable treatment options [14]. Patients with intrathoracic only recurrence have longer post-recurrence survival compared to patients with extrathoracic recurrences (1 year post recurrence survival 50% vs 26%, p<0.01) [14]. The inclusion of patients with limited site recurrences may contribute to the improved post-recurrence survival. A limitation of our findings is that we did not note the sites of recurrent or metastatic disease for the patients in this series. It is possible that a proportion of the included patients with recurrent disease had localized involvement that may have been amenable definitive therapy contributing to the improved outcome of the recurrent metastatic patients. An additional hypothesis is that regular imaging surveillance performed after definitive treatment for locally advanced lung cancers may contribute to lead time bias as disease recurrence may be found earlier when patients are asymptomatic. Conversely, de novo metastatic lung cancers typically present when symptoms are present, at a later time point in the disease course.
Information regarding prognostic implications of recurrent versus de novo metastatic cancer is not widely reported. The only published reports are in patients with breast cancer where the findings are opposite. In metastatic breast cancer, patients with de novo metastatic disease have longer overall survival from diagnosis of metastatic disease compared to earlier stage patients that relapse with metastatic disease (median 39 vs 29 mo, p<0.001)[15]. These findings were corroborated in a cohort of patients with HER2 positive breast cancer. Overall survival from diagnosis of metastatic disease was 42 months for de novo HER2 positive metastatic patients compared to 33 months for patients with recurrent HER2 positive breast cancers [16]. This discrepant finding may be related to selection of chemotherapy resistant clones in the recurrent breast cancer setting. Eighty percent of breast cancer patients with recurrent disease received adjuvant chemotherapy [15, 16] while less than 20 percent of recurrent lung cancer patients received perioperative chemotherapy for their initial lung cancer diagnosis [14]. In addition, breast cancer is typically diagnosed in the locally advanced setting and only a small minority of locally advanced breast cancers recur. This pattern is in striking contrast to lung cancers, where a minority (40%) of patients are diagnosed when locally advanced, and a large majority of those with locally advanced cancers recur. Consequently, recurrent metastatic disease in breast cancer represents the most aggressive disease within the disease spectrum, however the opposite may be true in lung cancer. This difference may explain the contradictory findings in our report compared to the breast cancer literature.
In summary, we have found that stage at initial diagnosis is an important prognostic factor in metastatic lung adenocarcinomas. The association between improved survival in those with recurrent compared to de novo metastatic disease was found in both EGFR mutant and KRAS mutant lung adenocarcinomas. These findings should be validated in prospective datasets and confirmed in other mutation subtypes. In future studies of patients with recurrent or metastatic lung cancers, stage at diagnosis should be recorded and stratification based on this clinical characteristic considered in phase 3 studies. Stage at diagnosis should be noted as a prognostic factor when interpreting treatment and survival outcomes on clinical trials.
Table 6.
Multivariate analysis of prognostic factors in metastatic EGFR mutant lung cancers
| Variable | Hazard Ratio (95% CI) | P value |
|---|---|---|
| Recurrent vs De novo | 1.36 (95% CI 1.05–1.76) | 0.019 |
| Age | 1.05 (95%CI 1.00–1.10) | 0.058 |
| KPS (≤ 70 vs ≥ 80) | 1.54 (95% CI 1.18-2.02) | 0.002 |
Acknowledgments
Dr. Yu has consulted for Clovis Oncology, and has received research support from Clovis Oncology, Astra Zenica, Pfizer, Astellas and Incyte. Dr. Riely has consulted for Ariad, Celgene, Mersana and Novartis and has received research support from Novartis, Roche/Genentech, Millennium, GlaxoSmithKline, Pfizer, and Infinity. Dr. Kris has consulted for Astra Zenica, Clovis Oncology, Pfizer, Boehringer Ingelheim, Roche/Genentech and has received research support from Pfizer and Puma Biotechnology.
Footnotes
Disclosures: Dr. Hellmann, Dr. Sima, Dr. Naidoo, Ms. Busby, and Ms. Rodriguez have no disclosures to report.
Works Cited
- 1.Albain KS, et al. Survival determinants in extensive-stage non-small-cell lung cancer: the Southwest Oncology Group experience. J Clin Oncol. 1991;9(9):1618–1626. doi: 10.1200/JCO.1991.9.9.1618. [DOI] [PubMed] [Google Scholar]
- 2.Finkelstein DM, Ettinger DS, Ruckdeschel JC. Long-term survivors in metastatic non-small-cell lung cancer: an Eastern Cooperative Oncology Group Study. J Clin Oncol. 1986;4(5):702–709. doi: 10.1200/JCO.1986.4.5.702. [DOI] [PubMed] [Google Scholar]
- 3.Buccheri G, Ferrigno D, Tamburini M. Karnofsky and ECOG performance status scoring in lung cancer: a prospective, longitudinal study of 536 patients from a single institution. Eur J Cancer. 1996;32A(7):1135–1141. doi: 10.1016/0959-8049(95)00664-8. [DOI] [PubMed] [Google Scholar]
- 4.Radzikowska E, Glaz P, Roszkowski K. Lung cancer in women: age, smoking, histology, performance status, stage, initial treatment and survival. Population-based study of 20 561 cases. Ann Oncol. 2002;13(7):1087–1093. doi: 10.1093/annonc/mdf187. [DOI] [PubMed] [Google Scholar]
- 5.Stanley KE. Prognostic factors for survival in patients with inoperable lung cancer. J Natl Cancer Inst. 1980;65(1):25–32. [PubMed] [Google Scholar]
- 6.Pao W, et al. EGF receptor gene mutations are common in lung cancers from "never smokers" and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A. 2004;101(36):13306–13311. doi: 10.1073/pnas.0405220101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Rosell R, 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(3):239–246. doi: 10.1016/S1470-2045(11)70393-X. [DOI] [PubMed] [Google Scholar]
- 8.Sequist LV, et al. Phase III Study of Afatinib or Cisplatin Plus Pemetrexed in Patients With Metastatic Lung Adenocarcinoma With EGFR Mutations. J Clin Oncol. 2013;31(27):3327–3334. doi: 10.1200/JCO.2012.44.2806. [DOI] [PubMed] [Google Scholar]
- 9.Johnson ML, et al. Association of KRAS and EGFR mutations with survival in patients with advanced lung adenocarcinomas. Cancer. 2012;119(2):356–362. doi: 10.1002/cncr.27730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Groome PA, et al. The IASLC Lung Cancer Staging Project: validation of the proposals for revision of the T, N, and M descriptors and consequent stage groupings in the forthcoming (seventh) edition of the TNM classification of malignant tumours. J Thorac Oncol. 2007;2(8):694–705. doi: 10.1097/JTO.0b013e31812d05d5. [DOI] [PubMed] [Google Scholar]
- 11.Pan Q, Pao W, Ladanyi M. Rapid polymerase chain reaction-based detection of epidermal growth factor receptor gene mutations in lung adenocarcinomas. J Mol Diagn. 2005;7(3):396–403. doi: 10.1016/S1525-1578(10)60569-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Pao W, et al. KRAS mutations and primary resistance of lung adenocarcinomas to gefitinib or erlotinib. PLoS Med. 2005;2(1):e17. doi: 10.1371/journal.pmed.0020017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Brevet M, et al. Detection of EGFR mutations in plasma DNA from lung cancer patients by mass spectrometry genotyping is predictive of tumor EGFR status and response to EGFR inhibitors. Lung Cancer. 2011;73(1):96–102. doi: 10.1016/j.lungcan.2010.10.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sugimura H, et al. Survival after recurrent nonsmall-cell lung cancer after complete pulmonary resection. Ann Thorac Surg. 2007;83(2):409–417. doi: 10.1016/j.athoracsur.2006.08.046. discussioin 417–8. [DOI] [PubMed] [Google Scholar]
- 15.Dawood S, et al. Survival differences among women with de novo stage IV and relapsed breast cancer. Ann Oncol. 2010;21(11):2169–2174. doi: 10.1093/annonc/mdq220. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Yardley DA, et al. Treatment patterns and clinical outcomes for patients with de novo versus recurrent HER2-positive metastatic breast cancer. Breast Cancer Res Treat. 2014;145(3):725–734. doi: 10.1007/s10549-014-2916-8. [DOI] [PMC free article] [PubMed] [Google Scholar]