Identification of a KRAS mutation in a patient with non-small cell lung cancer treated with chemoradiotherapy and panitumumab

Abstract

RTOG 0839 is a Phase II study of pre-operative chemoradiotherapy with or without panitumumab in potentially operable locally advanced non-small cell lung cancer (NSCLC). The investigational agent, panitumumab, is an anti-epithelial growth factor receptor (EGFR) antibody that improves progression-free survival in chemorefractory metastatic colorectal cancer (mCRC). Recently, both KRAS mutational status (i.e., mutated or not) and subtype (i.e., activating or inactivating) have been shown to be predictive of response to anti-EGFR therapy in mCRC. However, in NSCLC, it is unknown if KRAS mutational status or subtype predict benefit to anti-EGFR therapies because of unique genetic and epigenetic factors unique to each cancer. We present a patient with stage III NSCLC containing a KRAS G12D activating mutation who had a partial pathologic response, with disappearance of a minor KRAS mutant clone. This case suggests possible eradication of the G12D KRAS lung cancer clones by concurrent chemoradiation with panitumumab.

Introduction

Lung cancer is the leading cause of cancer-related mortality in the United States, with an estimated annual morality of 160 000. Non-small cell lung cancer (NSCLC) accounts for 85% of these cases, and about a third of NSCLC patients present with stage III disease where the primary tumor has directly invaded local structures outside the lung (T3–T4) and/or the cancer has spread to the mediastinal lymph nodes (N2–3). These tumors are generally not amenable to surgical resection. Patients treated with standard therapies, including concurrent chemotherapy with definitive radiation therapy (RT) followed by surgical resection in select patients, have median survival times of 18 to 24 mo. Most patients are not cured and have therapy-related complications. Investigations are focused on molecularly targeted therapies, which can greatly benefit selected advanced NSCLC patients.¹

Panitumumab, an anti-epithelial growth factor receptor (EGFR) antibody, is a targeted molecular agent under investigation in the Phase II study RTOG 0839.² RTOG 0839 compares pre-operative chemo-RT with or without panitumumab in potentially resectable locally advanced NSCLC. When the RTOG 0839 protocol was initially written (November 2010), the hypothesis was that panitumumab, in combination with chemoradiotherapy, would improve mediastinal pathologic complete response (pCR), a surrogate marker for improved overall survival (OS), compared with chemo-RT alone.³ This hypothesis was based on studies demonstrating improved outcomes in patients with metastatic colorectal cancer (mCRC) treated with the anti-EGFR agent cetuximab (FDA approved in 2009).⁴

However, there has been controversy regarding the efficacy of anti-EGFR agents in mCRC: in 2006, it was hypothesized that any KRAS mutation precluded benefit from anti-EGFR therapy;⁵ in 2012, this was narrowed to only activating KRAS mutations.⁶ Currently, the utility of determining KRAS mutational status or subtype to predict benefit to anti-EGFR therapies in NSCLC remains unclear, and there is no a priori reason the think that the same mutation should result in the same phenotype when expressed in cells because of unique genetic and epigenetic factors unique to NSCLC and mCRC.

Case Report

A 57-y-old woman with a heavy smoking history presented to her primary care physician with a nonproductive cough. Imaging studies showed a 4 cm right upper lobe mass. A PET-CT showed avidity in the mass, and pretracheal and right hilar adenopathy. DNA Sanger sequencing of bronchoscopic biopsy of a right upper lobe mass identified wild-type (i.e., no mutation) EGFR and a minor clone containing an activating, guanine to adenine mutation on chromosome 12 associated with a KRAS missense mutation (glycine to aspartate [G12D]; [Fig. 1A and B]). The patient was diagnosed with a stage III (T2aN2M0) NSCLC and enrolled on the RTOG 0839 clinical trial.

Figure 1. A PET-CT showed avidity in the lung mass (A). Bronchoscopic biopsy with subsequent DNA Sanger sequencing (B) identified a minor clone containing an activating, KRAS G12D mutation. A PET/CT after completion of radiation, chemotherapy, and panitumumab showed a partial response with reduced FDG uptake and residual disease in the mediastinum (C). Ion torrent sequencing revealed that the residual tumor cells harbor the same cancer gene mutations as the pretreatment tumor tissues, with the exception of G12D KRAS mutation being absent (D). In mCRC, which are predominantly adenocarcinomas, panitumumab halts the phosphorylation of KRAS-GDP, preventing activation of the KRAS/RAF/MEK/ERK pathway ([E], blocked pathways grayed out); however, activating mutations cause constitutive activation of the downstream pathway ([F], pathway no longer gray). In NSCLC, even when an activating KRAS mutation is present, all pathways contribute to cell cycle progression, proliferation, angiogenesis, metastasis, and production of EGFR ligands that cause radioresistance ([G], without panitumumab all pathways are active); with panitumumab, although the KRAS pathways may be constitutively activated, the overall number of effectors contributing to cancer cell proliferation is decreased ([H], panitumumab causes inactivation of pathways, now grayed out).

Per the RTOG 0839 protocol,² she received external beam radiation therapy (EBRT) at 2 Gy/day, 5 times per week, to 60 Gy with concomitant panitumumab once per week for 6 weeks and paclitaxel and carboplatin once per week for 6 weeks. A PET/CT after completion of radiation, chemotherapy, and panitumumab showed a partial response with reduced FDG uptake and residual disease in the mediastinum. She had a right thoracotomy and upper lobectomy. Pathology with double immunohistochemical stain, TTF1 (nuclear), and Napsin A (cytoplasmic), confirmed residual moderately-differentiated adenocarcinoma. Ion Torrent sequencing revealed that the residual tumor cells harbor the same cancer gene mutations as the pretreatment tumor tissues, with the exception of G12D KRAS mutation being absent (Fig. 1C and D). Other mutations (e.g., in EGFR) were not identified on either specimen.

Discussion

The investigational agent of RTOG 0839, panitumumab, is a human monoclonal IgG2 antibody specific to the extracellular domain of the EGFR. Panitumumab improves progression-free survival in chemorefractory mCRC.⁴ Further, panitumumab may have less toxicity than cetuximab, which has been shown to improve pCR and OS in certain advanced NSCLC patients receiving platinum-based chemotherapy.⁷

In mCRC, which are predominantly adenocarcinomas, panitumumab halts the phosphorylation of KRAS-GDP, preventing activation of the KRAS/RAF/MEK/ERK pathway, which is responsible for cell cycle progression, survival, proliferation, angiogenesis, and metastasis (Fig. 1E, blocked pathways grayed out). However, many mCRC tumors that have wild-type KRAS and are sensitive to EGFR blockage develop resistance within several months of therapy.⁸

Resistance to anti-EGFR therapy occurs due to a subpopulation of KRAS-mutated cells (found in 30–50% of CRC tumors) that are preferentially selected once wild type cells are damaged with panitumumab.⁹ In mCRC the EGFR–KRAS signaling cascade is considered to function as a unidirectional, linear, outside-in signaling cascade.¹⁰ In 2005 it was hypothesized that tumors harboring mutant KRAS are independent of EGFR activation because KRAS is constitutively activated.¹¹ In 2012, the KRAS mutation subtype was shown to be predictive of outcome more so than mutational status alone: activating mutations cause constitutive activation of the downstream pathway (Fig. 1F, pathway no longer gray), while inactivating mutations do not. Once activating KRAS mutations are detected in mCRC, anti-EGFR therapy is not indicated.⁵

About 87% of all lung cancers are NSCLC, and 50% of these are adenocarcinomas. EGFR is overexpressed in 40–80% of NSCLC adenocarcinomas.¹² The hypothesis in RTOG 0839 is that chemoradiotherapy and anti-EGFR therapy would have pCR and OS rates similar to RTOG 0324 (which investigated cetuximab in NSCLC), but with less toxicity.⁷ KRAS mutations are found in 15–30% of all patients with NSCLC (30% of adenocarcinomas, 5% of squamous cell carcinomas)¹¹ and predict poor outcome to conventional carbo/taxol treatment regimens.¹³ Unlike in mCRC, in NSCLC, KRAS mutation (1) status has not been shown to predict benefit from anti-EGFR monoclonal antibody cetuximab, emphasizing a biological difference between mCRC and NSCLC;¹⁴ and (2) subtypes have not yet been investigated as a predictor of response to anti-EGFR therapy in NSCLC.

Neither KRAS mutation status nor subtype was considered in the design of RTOG 0839 for a few reasons. In 2006 it was shown that EGFR antibodies had marked radiosensitizing properties that resulted in improved outcomes in a locally advanced head and neck cancer.¹⁵ Although head and neck cancers are mostly squamous cell carcinomas, and KRAS mutations occur in <7% of patients,¹⁶ the investigators of RTOG 0839 wanted to include as many patients as possible, regardless of KRAS status.

Second, the “off target” effects of the targeted therapy (i.e., cetuximab) were not known in 2006, and KRAS mutation subtype was only shown to predict outcome to anti-EGFR therapy in mCRC in 2012.⁶ If the RTOG decided that specific KRAS mutations precluded activity of cetuximab (for which 0839 was initially designed) or panitumumab, or if it had the resources to require testing for the mutations before patients were entered on study, this patient would not have been entered and the case report would not exist. The understanding of anticipated and unanticipated “off target” effects of a targeted therapy are important in understanding of the underlying mechanisms of the agent.

Third, NSCLC has intracellular signaling pathways that are different from those in mCRC. Evidence from the last decade has shown that, through an autocrine feedback loop, EGFR can also be viewed as a downstream effector of the KRAS pathway in NSCLC, which is different from the unidirectional mechanism in mCRC. In NSCLC, both EGFR and KRAS use multiple effector pathways, including PLC-γ, PI3K, and JAK-STAT. Even when an activating KRAS mutation is present, all pathways contribute to cell cycle progression, proliferation, angiogenesis, metastasis, and production of EGFR ligands that cause radioresistance (Fig. 1G, without panitumumab all pathways are active).^17-19 The differences in these KRAS alternative signaling pathways may be due to the different (1) KRAS mutation subtypes, which have different affinities to downstream effectors; (2) KRAS mutation subtype frequencies; and (3) different risk factors (e.g., tobacco) associated with KRAS mutations in NSCLC.²⁰

KRAS-mutant, genetically engineered mouse models of NSCLC suggest that dual inhibition of PI3K and KRAS pathways are required to fully block oncogenic KRAS signaling.²¹ Theoretically, in KRAS G12D mutant NSCLC, panitumumab should still be able to prevent activation of the PLC-γ, PI3K, and JAK-STAT pathways. Thus, although the KRAS pathways may be constitutively activated, the overall number of effectors contributing to cancer cell proliferation is decreased (Fig. 1H, panitumumab causes inactivation of pathways, now grayed out).

Currently, the utility of determining KRAS mutational status or subtype to predict benefit to anti-EGFR therapies in NSCLC remains unclear, and KRAS mutations are not a validated biomarker in the negative selection of lung cancer patients who are otherwise candidates for anti-EGFR therapy. Moreover, a recent analysis of several trials of adjuvant chemotherapy in patients with resected early stage NSCLC has not shown a clear impact of KRAS mutation status or subtype on the effect of cisplatin-containing adjuvant therapy.²² Nonetheless, at least 8 mutant KRAS subtypes exist in NSCLC,²³ and the effect of these mutations should be further explored.

This case suggests possible eradication of the G12D KRAS lung cancer clones by concurrent chemoradiation with panitumumab. However, this report has limitations: first, it is a single case event, and we do not know if a similar scenario would occur in other KRAS G12D patients on RTOG 0839; second, sampling error may have occurred. Since patient received RT, chemotherapy and panitumumab, it is unclear whether the panitumumab played any role in the response to therapy, as chemo-RT by itself is capable of producing a complete response in 10–30% of patients when given prior to resection. Finally, since the initial G12D KRAS lung cancer cell clones represented a subpopulation of the initial tumor mass (Fig. 1B), similar heterogeneity in the tumor could account for the “disappearance” of this clone in the post-treatment biopsy without invoking any effect of panitumumab.²⁴

Conclusion

In conclusion, the role of KRAS mutational status and subtype in predicting a response to EGFR-directed therapies in NSCLC is currently unknown. This case suggests possible eradication of the G12D KRAS lung cancer clones by concurrent chemoradiation with panitumumab. To confirm our findings, we recommend (1) testing for KRAS mutations in all patients on study and distinguishing activating vs. inactivating mutations; (2) analyzing both the initial biopsy specimens and the post-treatment surgical specimens; and (3) correlating these to patient outcomes.

Glossary

Abbreviations:

EBRT

external beam radiation therapy

EGFR

epithelial growth factor receptor

mCRC

metastatic colorectal cancer

KRAS

Kirsten rat sarcoma

NSCLC

non-small cell lung cancer

pCR

pathologic complete response

OS

overall survival

RT

radiation therapy

RTOG

Radiation Therapy Oncology Group

Disclosure of Potential Conflicts of Interest

No potential conflict of interest was disclosed.