The treatment of advanced non small-cell lung cancer

Brief Report PTPRF Expression as a Potential Prognostic/Predictive Marker for Treatment with Erlotinib in Non Small-Cell Lung Cancer Denis Soulières, MD,* Fred R. Hirsch, MD, PhD, Frances A. Shepherd, MD, Walter Bordogna, PhD, Paul Delmar, PhD, David S. Shames, PhD,# and Barbara Klughammer, PhD# Introduction: EGFR mutations and anaplastic lymphoma kinase rearrangements are, to date, the only approved biomarkers to select treatment for non small-cell lung cancer (NSCLC). However, there is considerable interest in identifying other predictive markers. The PTPRF gene has been suggested as a marker of interest in NSCLC and other tumor types. Methods: This hypothesis-generating retrospective analysis examined data from two studies of erlotinib in NSCLC, Marker Identification Trial (MERIT; n = 102) and Sequential Tarceva in Unresectable NSCLC (SATURN; n = 262), to determine whether PTPRF expression was prognostic and/or predictive of patient outcomes. Exploratory analyses were conducted using quantitative reverse transcription polymerase chain reaction on existing formalin-fixed paraffin-embedded samples, to assess gene expression levels, including PTPRF. High versus low levels of expression were dichotomized using the median with B2M as a control comparator. Progression-free survival and overall survival were then compared for patients with high versus low levels of PTPRF in the two studies. Results: PTPRF expression was found to be prognostic for shorter overall survival but was also significantly predictive of improved survival with erlotinib versus placebo in SATURN (hazard ratio, 0.45 [95% confidence interval, CI, 0.30 0.69] in PTPRF high versus 0.96 [95% CI, 0.62 1.48] in PTPRF low; interaction p = 0.02), even in the EGFR wild-type subpopulation (adjusted hazard ratio, 0.44 [95% CI, 0.29 0.68] versus 0.96 [95% CI, 0.62 1.48]; interaction p = 0.01). *Department of Hematology, Centre Hospitalier de l Université de Montréal, Montréal, Canada; Department of Medicine, University of Colorado Cancer Center, Aurora, Colorado; Department of Medicine, University Health Network, Princess Margaret Cancer Centre, Toronto, Canada; Departments of Clinical Science, and Statistics, F. Hoffmann-La Roche Ltd, Basel, Switzerland; and #Department of Oncology Biomarkers, Genentech, Inc., SanFrancisco, California Disclosures: Support for third-party writing support was provided by F. Hoffmann-La Roche Ltd. Drs. Klughammer, Bordogna, and Delmar are employees of F. Hoffmann-La Roche Ltd; Dr. Shames is an employee of Genentech; Dr. Hirsch has received grants and consultancy fees from Genentech; Dr. Shepherd has received consultancy fees and honoraria from F. Hoffmann-La Roche Ltd; Dr. Soulières has received payment for lectures and fees for participation of review activities from F. Hoffmann-La Roche Ltd. Address for correspondence: Barbara Klughammer, F. Hoffmann-La Roche Ltd, Basel, Switzerland. E-mail: barbara.klughammer@roche.com DOI: 10.1097/JTO.0000000000000624 ISSN: 1556-0864/15/1009-1364 Conclusions: PTPRF may have value as a predictive marker to identify which patients can obtain the greatest benefit from erlotinib in the post first-line setting. Further research is warranted to determine the potential value of this marker in clinical decision-making. Key Words: Erlotinib, Non small-cell lung cancer, Biomarkers, Protein tyrosine phosphatase, PTPRF. (J Thorac Oncol. 2015;10: 1364 1369) The treatment of advanced non small-cell lung cancer (NSCLC) has been transformed by the advent of targeted therapies with defined biomarkers. Epidermal growth factor receptor tyrosine kinase inhibitors (EGFR TKIs) provide significant progression-free survival (PFS) benefits as first-line treatment for EGFR mutation-positive NSCLC, 1 3 but other exploratory subgroup analyses have suggested that some patients with wild-type EGFR also derive clinical benefit from erlotinib, as demonstrated in studies of unselected patients. 4,5 Approximately 70% to 85% of patients with stage IIIB/IV NSCLC have double-negative tumors, without either an ALK rearrangement or an EGFR mutation. These patients receive generic treatment not tailored to their individual disease, except for the use of histology to exclude specific chemotherapy agents. With the goal being to provide the most effective therapy in the era of molecularly defined lung cancer, the oncology community continues to look for new predictive biomarkers to identify cost-effective treatments for doublenegative patients, across all lines of therapy. One potential marker is the protein tyrosine phosphatase (PTP) receptor type F (PTPRF, sometimes called LAR), which plays a key role in regulating EGFR signaling, by dephosphorylating key tyrosine residues on the receptor. 6,7 Aberrations in the PTP gene family are reported as carcinogenic in numerous solid tumors, including head and neck, prostate and breast cancers. Müller et al. 8 concluded that loss of PTPRF expression could lead to malignancy. A recent study found that high expression of PTPRF was associated with decreased levels of phosphorylated EGFR in breast cancer, which resulted in reduced cell invasion, migration, and metastasis. 9 In contrast, expression of PTPRF is upregulated in some breast cancer tissues compared with controls and has been associated with increased metastatic potential in animal 1364 Journal of Thoracic Oncology Volume 10, Number 9, September 2015

Journal of Thoracic Oncology Volume 10, Number 9, September 2015 PTPRF Expression as a Marker for Erlotinib in NSCLC models. 10 The role of PTPRF in tumorigenesis and metastatic behavior remains unclear and may depend on its effect on the expression of other proteins in different tumor types. In lung cancer, specifically, the PTP gene family is of great interest in NSCLC 11,12 and other tumor types, with 107 PTPs identified in the human genome. 13 Motiwala et al. 11 found that overexpression of PTPRO appeared to lead to suppression of growth and increased susceptibility to apoptosis in the A549 lung cancer cell line, whereas Scrima et al. 12 found that mrna expression of PTPRF, along with that of PTPN13, PTPRE, and PTPRD, was significantly downregulated in NSCLC cell lines (p = 0.024). Both groups hypothesized that these genes may act as tumor suppressors. 11,12 Considering the role of PTPRF in the modulation of EGFR signaling and its carcinogenic potential, plus the use of EGFR-targeted therapy in metastatic NSCLC, a retrospective analysis was carried out to evaluate its potential as a predictive marker for treatment with erlotinib and to determine the prognostic role of PTPRF in NSCLC. This study is supported by the fact that PTPRF was found to be significantly differentially expressed by quantitative reverse transcription polymerase chain reaction on fresh-frozen and fixed material in Marker Identification Trial (MERIT, later described as the discovery cohort), a study designed to evaluate markers associated with erlotinib efficacy in metastatic NSCLC patients who had failed first-line chemotherapy. In that study, PTPRF was specifically interesting as it was one of the five markers most associated with clinical efficacy. It was decided to further investigate this marker before the others because of its higher association with clinical efficacy with a differential expression factor of more than five, and because it was potentially linked to the EGFR signaling pathway without specifically addressing the target gene (EGFR). MATERIALS AND METHODS Clinical Study Designs MERIT study (discovery cohort) The MERIT was a phase II, nonrandomized, gene expression profiling study conducted at 26 centers across 12 countries. A total of 264 eligible patients were enrolled and received oral erlotinib (150 mg/d) until progression, unacceptable toxicity, or death. Full methodology has been reported previously in the primary publication. 14 The main end point of the study was to correlate response to markers that were differentially expressed by a factor of eight. No marker meeting this criterion was identified. Other markers were sought in the database through additional exploratory analyses, to identify potential genes of interest for further research. EGFR, PSPH, RAPGEF5, and PTPRF were among the markers most strongly associated with clinical benefit. 14 SATURN study (validation cohort) The Sequential Tarceva in Unresectable NSCLC (SATURN) study was an international, phase III, randomized, placebo-controlled study to assess the efficacy and safety of maintenance erlotinib after first-line platinum-doublet chemotherapy. For this study, 1949 eligible patients received four cycles of first-line chemotherapy during the run-in phase, following which those without disease progression who had adequate performance status and good hematologic/renal/ hepatic function (n = 889) were randomized to receive maintenance erlotinib (orally, 150 mg/d) or placebo until progression, unacceptable toxicity or death. The full methodology has been reported previously. 4 Provision of a tumor sample was mandatory for SATURN, and extensive biomarker analyses have been conducted and reported. 15 Further exploratory analyses were performed based on previous observations from MERIT. Biomarker Assessment Methodology Exploratory analyses were conducted on material extracted from existing formalin-fixed, paraffin-embedded samples. For MERIT, RNA was extracted using the RNeasy Mini Kit (QIAGEN GmbH, Hilden, Germany). For SATURN, protocols and processing instructions developed by Roche Diagnostics (Penzberg, Germany) were followed. Quantitative reverse transcription polymerase chain reaction was then used to assess the prognostic and predictive value of marker gene expression levels, including PTPRF. High versus low levels of expression were dichotomized using the median, ie, high: greater than or equal to median and low: less than median with the housekeeping gene B2M being used as a control comparator. The methodology used was as previously described for similar gene expression analyses in the relevant studies. 14,15 Statistical Methodology For MERIT, hazard ratios (HRs) were estimated using Kaplan Meier methodology, with statistical significance expressed as p values using Wald s test. For SATURN, HRs were estimated using Kaplan Meier methodology, with interaction p values based on a Cox proportional hazard model. Treatment (erlotinib versus placebo), PTPRF expression level (high versus low), and treatment PTPRF level interaction were used as covariates for this analysis. Subgroup analysis in the population of patients previously identified as EGFR wild type was also performed, as this is the population for whom new predictive markers are required. As retrospective analyses of small patient subgroups, these analyses are not powered to demonstrate statistical significance but may identify trends. RESULTS MERIT Study (Discovery Cohort) In total, 102 representative samples were available for analysis of PTPRF gene expression; baseline characteristics were well balanced between the PTPRF subgroups (Table 1). In an exploratory ranking of genes based on their association with clinical benefit, PTPRF was in the top five, although formal statistical significance was not reached after correction for multiple testing. Examination of PFS and overall survival (OS) relative to PTPRF expression level suggested a differential effect (Fig. 1), but it was impossible to distinguish between the 1365

Soulières et al. Journal of Thoracic Oncology Volume 10, Number 9, September 2015 TABLE 1. Key Baseline Characteristics of the SATURN Validation Cohort by PTPRF Expression Characteristic SATURN Validation Cohort PTPRF High (n = 131) PTPRF Low (n = 131) Median age, yr (range) 60 (44 83) 59 (35 82) P value 0.370 Gender, n (%) Male 103 (79) 86 (66) Female 28 (21) 45 (34) P value 0.019 ECOG PS, n (%) 0 40 (31) 45 (34) 1 91 (69) 86 (66) P value 0.509 Histology, n (%) Adenocarcinoma 45 (34) 80 (61) Squamous 78 (60) 36 (27) Other 8 (6) 15 (11) P value <0.001 Smoking status, n (%) Current smoker 86 (66) 65 (50) Former smoker 27 (21) 37 (28) Never smoker 18 (14) 29 (22) P value 0.029 EGFR mutation status, n (%) Activating mutation 13 (10) 15 (11) Wild type 104 (79) 93 (71) Resistance mutation 0 (0) 1 (1) Indeterminate/missing/other 14 (11) 22 (17) P value 0.528 SATURN, Sequential Tarceva in Unresectable NSCLC; EGFR, epidermal growth factor receptor; PTPRF, protein tyrosine phosphatase receptor type F; ECOG PS, Eastern Cooperative Oncology Group performance status. prognostic and predictive values of the marker in this singlearm study. Further validation was performed in SATURN. SATURN Study (Validation Cohort) Samples were available from 262 patients for PTPRF analysis. The population for whom PTPRF expression level was analyzed was representative of the overall study population (Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/jto/a854) and the subgroups were well balanced (Table 1); the results are, therefore, unlikely to be affected by sampling bias. PFS was improved with erlotinib versus placebo in the subgroup with high PTPRF expression (PTPRF high; HR, 0.47, 95% confidence interval [CI], 0.33 0.69) and a difference was also noted in the subgroup with low PTPRF expression (PTPRF low; HR, 0.77, 95% CI, 0.54 1.11; Fig. 2), although the CI for the latter subgroup did exceed 1.0. The treatment by PTPRF level interaction p value for PTPRF high versus PTPRF low was 0.06, showing a trend toward statistical significance. OS also was improved with erlotinib versus placebo in the PTPRF high subgroup (HR, 0.45, 95% CI, 0.30 0.69), but there was a negligible difference in the PTPRF low subgroup (HR, 0.96, 95% CI, 0.62 1.48; Fig. 3). The treatment by PTPRF level interaction p value was significant (p = 0.02). Examining the PFS and OS outcomes for the placebo group demonstrates that high PTPRF expression is a potential prognostic marker for shorter OS (respectively, p = 0.06 and p = 0.01; Fig. 4). Analysis in the confirmed EGFR wild-type population found similar results for PFS and OS in the PTPRF high versus PTPRF low subgroups (Fig. 5), suggesting that PTPRF expression is an independent marker for erlotinib treatment outcomes, rather than a surrogate for mutation status. When the data were adjusted for clinical covariates (gender, histology, Eastern Cooperative Oncology Group performance status), HRs for PFS were 0.48 (95% CI, 0.33 0.69) in the PTPRF high group and 0.78 (95% CI, 0.54 1.14) in the PTPRF low group, with an interaction p value of 0.05. For OS, the HRs were 0.44 (95% CI, 0.29 0.68) and 0.96 (95% CI, 0.62 1.48), respectively, with a significant interaction p value of 0.01. Further exploratory analyses from additional studies are included as Supplementary material (Supplemental Digital Content 2, http://links.lww.com/jto/a855). DISCUSSION The MERIT study was initiated to generate hypotheses on factors associated with the efficacy of erlotinib. It was a single-arm study, and the results were of sufficiently significant interest for some genes found to be differentially expressed, including PTPRF, to test the prognostic, but more importantly predictive value, in a cohort randomized to erlotinib or placebo. However, it must be noted that these analyses were unplanned and should be considered as hypothesis generating until further validation is obtained. On the basis of the results from the translational analyses presented in this article of specimens obtained from clinical studies of erlotinib in patients with advanced NSCLC, PTPRF expression above the median appears to be a negative prognostic factor for OS in patients who have completed first-line therapy and for whom treatment with erlotinib is envisioned, although no prognostic influence on PFS could be discerned. This may be because of crossover, whereby the predictive effect of PTPRF expression is lost in the analysis of OS, whereas the prognostic effect, which is independent of treatment, remains identifiable. To confirm the prognostic value of PTPRF expression, however, we will require large cohorts of patients receiving multiple types and lines of therapies and evaluated in multivariate analysis considering other currently known prognostic factors. The presented analysis also suggests the hypothesis that PTPRF expression above the median may be a positive predictive factor for erlotinib treatment, regardless of EGFR mutation status. Further research could help to elucidate the mechanism by which PTPRF may influence erlotinib outcomes in lung cancer. If PTPRF expression is indeed independent of EGFR mutation status, as suggested by the wild-type results in this analysis, it may have value as a potential predictive marker for erlotinib 1366

Journal of Thoracic Oncology Volume 10, Number 9, September 2015 PTPRF Expression as a Marker for Erlotinib in NSCLC FIGURE 1. Progression-free survival (PFS, A) and overall survival (OS, B) relative to protein tyrosine phosphatase receptor type F (PTPRF) expression levels in the Marker Identification Trial (MERIT; quantitative reverse transcription polymerase chain reaction data). FIGURE 2. Progression-free survival (PFS) relative to protein tyrosine phosphatase receptor type F (PTPRF) high (A) and PTPRF low (B) status for erlotinib versus placebo in Sequential Tarceva in Unresectable NSCLC (SATURN). FIGURE 3. Overall survival (OS) relative to protein tyrosine phosphatase receptor type F (PTPRF) high (A) and PTPRF low (B) status for erlotinib versus placebo in Sequential Tarceva in Unresectable NSCLC (SATURN). 1367

Soulières et al. Journal of Thoracic Oncology Volume 10, Number 9, September 2015 FIGURE 4. Progression-free survival (PFS, A) and overall survival (OS, B) relative to protein tyrosine phosphatase receptor type F (PTPRF) status in the placebo group. HR, hazard ratio; CI, confidence interval. FIGURE 5. Progression-free survival (PFS, A) and overall survival (OS, B) relative to protein tyrosine phosphatase receptor type F (PTPRF) status in the epidermal growth factor receptor (EGFR) wild-type group. 1368

Journal of Thoracic Oncology Volume 10, Number 9, September 2015 PTPRF Expression as a Marker for Erlotinib in NSCLC treatment in advanced wild-type NSCLC. Indeed, this could be the first predictive marker to offer clinical efficacy in restricting erlotinib treatment to those patients most likely to have high clinical benefit and could also identify a population that does not seem to derive benefit from erlotinib therapy. However, further studies would be needed to validate PTPRF as a biomarker in the clinical setting, with the determination of a set level of expression as a cutoff rather than the median, to allow for a test that could be administered consistently between different laboratories and across different sample types. With the knowledge currently available, targeted therapy for NSCLC based on gene alterations linked with oncogenic addiction represents less than 20% of patients, although the advent of next-generation sequencing may identify as yet unknown alterations in a higher percentage of patients. 16 In the meantime, additional biomarkers are still required to guide treatment choice for all other patients for whom treatment with an EGFR inhibitor is considered. The results reported here indicate that PTPRF expression in two cohorts of patients presents a marker of significant interest and should be the object of further studies to provide more guidance on the use of erlotinib therapy in metastatic NSCLC. REFERENCES 1. Fukuoka M, Wu YL, Thongprasert S, et al. Biomarker analyses and final overall survival results from a phase III, randomized, open-label, firstline study of gefitinib versus carboplatin/paclitaxel in clinically selected patients with advanced non-small-cell lung cancer in Asia (IPASS). J Clin Oncol 2011;29:2866 2874. 2. Rosell R, Carcereny E, Gervais R, et al.; Spanish Lung Cancer Group in collaboration with Groupe Français de Pneumo-Cancérologie and Associazione Italiana Oncologia Toracica. 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;13:239 246. 3. Sequist LV, Yang JC, Yamamoto N, et al. Phase III study of afatinib or cisplatin plus pemetrexed in patients with metastatic lung adenocarcinoma with EGFR mutations. J Clin Oncol 2013;31:3327 3334. 4. Cappuzzo F, Ciuleanu T, Stelmakh L, et al.; SATURN investigators. Erlotinib as maintenance treatment in advanced non-small-cell lung cancer: a multicentre, randomised, placebo-controlled phase 3 study. Lancet Oncol 2010;11:521 529. 5. Zhu CQ, da Cunha Santos G, Ding K, et al.; National Cancer Institute of Canada Clinical Trials Group Study BR.21. Role of KRAS and EGFR as biomarkers of response to erlotinib in National Cancer Institute of Canada Clinical Trials Group Study BR.21. J Clin Oncol 2008;26:4268 4275. 6. Hashimoto N, Zhang WR, Goldstein BJ. Insulin receptor and epidermal growth factor receptor dephosphorylation by three major rat liver protein-tyrosine phosphatases expressed in a recombinant bacterial system. Biochem J 1992;284 (Pt 2):569 576. 7. Kulas DT, Goldstein BJ, Mooney RA. The transmembrane protein-tyrosine phosphatase LAR modulates signaling by multiple receptor tyrosine kinases. J Biol Chem 1996;271:748 754. 8. Müller T, Choidas A, Reichmann E, Ullrich A. Phosphorylation and free pool of beta-catenin are regulated by tyrosine kinases and tyrosine phosphatases during epithelial cell migration. J Biol Chem 1999;274:10173 10183. 9. Du WW, Fang L, Li M, et al. MicroRNA mir-24 enhances tumor invasion and metastasis by targeting PTPN9 and PTPRF to promote EGF signaling. J Cell Sci 2013;126(Pt 6):1440 1453. 10. Whitmore TE, Peterson A, Holzman T, et al. Integrative analysis of N-linked human glycoproteomic data sets reveals PTPRF ectodomain as a novel plasma biomarker candidate for prostate cancer. J Proteome Res 2012;11:2653 2665. 11. Motiwala T, Kutay H, Ghoshal K, et al. Protein tyrosine phosphatase receptor-type O (PTPRO) exhibits characteristics of a candidate tumor suppressor in human lung cancer. Proc Natl Acad Sci U S A 2004;101:13844 13849. 12. Scrima M, De Marco C, De Vita F, et al. The nonreceptor-type tyrosine phosphatase PTPN13 is a tumor suppressor gene in non-small cell lung cancer. Am J Pathol 2012;180:1202 1214. 13. Alonso A, Sasin J, Bottini N, et al. Protein tyrosine phosphatases in the human genome. Cell 2004;117:699 711. 14. Tan EH, Ramlau R, Pluzanska A, et al. A multicentre phase II gene expression profiling study of putative relationships between tumour biomarkers and clinical response with erlotinib in non-small-cell lung cancer. Ann Oncol 2010;21:217 222. 15. Brugger W, Triller N, Blasinska-Morawiec M, et al. Prospective molecular marker analyses of EGFR and KRAS from a randomized, placebo-controlled study of erlotinib maintenance therapy in advanced non-small-cell lung cancer. J Clin Oncol 2011;29:4113 4120. 16. Drilon A, Wang L, Arcila ME, et al. Broad, hybrid capture-based nextgeneration sequencing identifies actionable genomic alterations in lung adenocarcinomas otherwise negative for such alterations by other genomic testing approaches. Clin Cancer Res 2015 [Epub ahead of print]. 1369