Browse the corpus

INTRODUCTION Targeted therapy has reshaped the care of many patients with lung cancers. Similar to sensitizing EGFR, BRAF, and MET mutations, recurrent gene rearrangements have emerged as actionable drivers in patients with ALK- and ROS1-rearranged lung cancers. In these individuals, dramatic improvements in response and progression-free survival compared to chemotherapy have been achieved with tyrosine kinase inhibition.1, 2 RET rearrangements are drivers of lung cancer oncogenesis.3 As with other recurrent gene rearrangements, the downstream RET gene maintains an intact tyrosine kinase domain, and is fused to a variety of upstream partners.4 While KIF5B-RET is the most common, multiple other fusion genes such as CCDC6-RET, NCOA4-RET, and TRIM33-RET have been reported.5 RET fusions are activating in vitro and in vivo.6 Upstream gene partners provide dimerization domains that result in ligand-independent signaling. Increased growth pathway activity downstream of the chimeric oncoprotein drives tumor cell proliferation and survival. The use of RET inhibitors results in the inhibition of downstream signaling and tumor growth.6–8

vivo.6 Upstream gene partners provide dimerization domains that result in ligand-independent signaling. Increased growth pathway activity downstream of the chimeric oncoprotein drives tumor cell proliferation and survival. The use of RET inhibitors results in the inhibition of downstream signaling and tumor growth.6–8 RET fusions are genomic alterations that can be routinely identified in the clinic.9 These are found in 1–2% of unselected lung cancers and tend to be mutually exclusive with other lung cancer drivers.3 Patients with RET-rearranged lung cancers are commonly never smokers or have a minimal history of prior tobacco exposure.10 From a pathologic perspective, RET fusions are identified largely in lung adenocarcinomas of the solid subtype or with signet ring cells.11 RET rearrangements can be identified by a number of tests including reverse transcriptase polymerase chain reaction (RT-PCR), fluorescence in situ hybridization (FISH), anchored multiplex polymerase chain reaction-based RNA sequencing, and broad, hybrid capture-based next-generation sequencing of DNA.12

t ring cells.11 RET rearrangements can be identified by a number of tests including reverse transcriptase polymerase chain reaction (RT-PCR), fluorescence in situ hybridization (FISH), anchored multiplex polymerase chain reaction-based RNA sequencing, and broad, hybrid capture-based next-generation sequencing of DNA.12 Cabozantinib is a multikinase inhibitor with low nanomolar (IC50 5·2 nM) activity against RET, in addition to its activity against ROS1, MET, VEGFR2, AXL, TIE2, and KIT.13 The use of cabozantinib results in the inhibition of lung cancer models harboring RET rearrangements. Shortly after publication of the first reports of the identification of RET rearrangements in tumors from patients with lung cancers in late 201114 and early 2012,3, 15, 16 we launched this phase 2 trial of cabozantinib for patients with RET-rearranged lung cancers. To our knowledge, this was the first prospective trial to test a RET inhibitor in a molecularly-enriched cohort of patients whose tumors harbored RET fusions.5

patients with lung cancers in late 201114 and early 2012,3, 15, 16 we launched this phase 2 trial of cabozantinib for patients with RET-rearranged lung cancers. To our knowledge, this was the first prospective trial to test a RET inhibitor in a molecularly-enriched cohort of patients whose tumors harbored RET fusions.5 METHODS Study design and patients This was an open-label, Simon two-stage17 phase 2 trial conducted at a single center in the USA. We included patients if they were 18 years of age or greater with metastatic or unresectable pathologically-confirmed lung cancers that harbored a RET rearrangement. Central pathologic confirmation was performed. Other eligibility criteria included a Karnofsky Performance Status of greater than 70%, adequate hematologic, renal, and hepatic function, and measurable disease by the Response Criteria Evaluation in Solid Tumors (RECIST) version 1·1.18 We included patients with treated or asymptomatic brain metastases. There were no restrictions on the number or type of prior systemic therapies except for cabozantinib.

ated along with exact 95% confidence intervals. All statistical analyses were conducted in R 3.2.2. While this trial has completed the accrual of patients with RET-rearranged lung cancers, it is ongoing as several patients remain on active treatment. This trial is registered with ClinicalTrials.gov, number NCT01639508. Role of the funding source Exelixis provided funding for study drug and research-related tests and assessments on this protocol. The authors were supported in part by funding from the National Institutes of Health/National Cancer Institute Cancer Center Support Grant P30 CA008748 that did not directly fund study costs. The authors wrote this article without any external funding or editorial support. AD designed the trial in cooperation with representatives of Exelixis. Exelixis was not involved in the collection, analysis, or interpretation of the data.

adequate hematologic, renal, and hepatic function, and measurable disease by the Response Criteria Evaluation in Solid Tumors (RECIST) version 1·1.18 We included patients with treated or asymptomatic brain metastases. There were no restrictions on the number or type of prior systemic therapies except for cabozantinib. Due to the potential antiangiogenic effects of cabozantinib mediated by its concomitant anti-VEGFR2 activity, patients were excluded if they had a history of significant bleeding, cavitating pulmonary lesions, tumors invading the tracheobronchial tree or major blood vessels, or a gastrointestinal disorder associated with a high risk of perforation or fistula formation. We excluded individuals receiving low molecular weight heparin, clopidogrel, or warfarin at therapeutic doses (appendix, p 1–2). This study was conducted in accordance with the provisions of the Declaration of Helsinki and Good Clinical Practice guidelines. The protocol was approved by an institutional review board and all patients provided written informed consent prior to participation. Tumor samples underwent either fluorescence in situ hybridization (FISH) or broad, hybrid capture-based next-generation sequencing in a Clinical Improvements Amendments (CLIA) laboratory to detect RET rearrangement. A dual-color break-apart FISH test was performed using institutional probes. Next-generation sequencing of tumor DNA was performed using one of two assays: MSK-IMPACT (Integrated Mutational Profiling of Actionable Cancer Targets ) or FoundationOne (appendix, p 2).

endments (CLIA) laboratory to detect RET rearrangement. A dual-color break-apart FISH test was performed using institutional probes. Next-generation sequencing of tumor DNA was performed using one of two assays: MSK-IMPACT (Integrated Mutational Profiling of Actionable Cancer Targets ) or FoundationOne (appendix, p 2). Procedures Cabozantinib was administered in tablet form at a starting dose of 60 mg orally once daily, the U.S. Food and Drug Administration (FDA)-approved dose for the treatment of patients with advanced renal cell carcinoma. Of note, plasma exposures (area under the plasma concentration-time curve) of the tablet formulation used in this study are comparable to the capsule formulation of cabozantinib that is administered at the FDA-approved dose of 140 mg daily for the treatment of patients with metastatic medullary thyroid carcinoma. Cabozantinib was administered in 28-day cycles. Treatment was continued until there was evidence of progression of disease or unacceptable toxicity. For patients who developed progression of disease according to RECIST v1·1,18 continued treatment with cabozantinib was permitted if the investigator felt that clinical benefit was maintained. Dose interruption and reduction followed a prescribed algorithm. A maximum of two dose reductions were allowed to 40 mg daily and 20 mg daily, respectively (appendix, p 3).

ease according to RECIST v1·1,18 continued treatment with cabozantinib was permitted if the investigator felt that clinical benefit was maintained. Dose interruption and reduction followed a prescribed algorithm. A maximum of two dose reductions were allowed to 40 mg daily and 20 mg daily, respectively (appendix, p 3). Computed tomography of the chest, abdomen, and pelvis was performed at baseline, 4 weeks after cabozantinib initiation, and every 8 weeks after the first response assessment scan (i.e. scans were performed at weeks 4, 12, 20, and so forth, appendix, p 2). Imaging of the brain was not required by the protocol. Treatment-related adverse events were graded using the Common Terminology Criteria for Adverse Events (CTCAE) version 4·0. Any patient who received one or more doses cabozantinib was included in the toxicity evaluations.

eks 4, 12, 20, and so forth, appendix, p 2). Imaging of the brain was not required by the protocol. Treatment-related adverse events were graded using the Common Terminology Criteria for Adverse Events (CTCAE) version 4·0. Any patient who received one or more doses cabozantinib was included in the toxicity evaluations. Outcomes The primary objective was to determine the overall response rate, defined as the proportion of patients with a confirmed complete response or partial response according to RECIST version 1·1. Secondary outcomes were progression-free survival, overall survival, and safety. Progression-free survival was measured from the date of initiation of cabozantinib until radiologic progression by RECIST version 1·1 or death. Patients alive and progression-free at the time of the last data cutoff were censored at the time of the last follow-up. Overall survival was measured from the date of initiation of cabozantinib treatment until death. Patients alive at the time of the last data cutoff were censored at the time of the last follow-up. Patients were deemed evaluable for an analysis of the activity of cabozantinib if they received at least one dose of study drug, and computed tomography imaging was performed at baseline and at least one protocol-specified follow up time point. One patient who did not meet these criteria was replaced, as is described later.

nts were deemed evaluable for an analysis of the activity of cabozantinib if they received at least one dose of study drug, and computed tomography imaging was performed at baseline and at least one protocol-specified follow up time point. One patient who did not meet these criteria was replaced, as is described later. Statistical analysis We used a Simon two-stage17 minimax design to test the null hypothesis of a 10% overall response rate, the historical response rate to chemotherapy in an unselected population who have previously received a platinum doublet, against the desired alternative of 30% overall response rate. This had a one-sided type I error of 10% and a power of 90%. In the first stage of this design, 16 patients were accrued. If no responses or one response was observed, the study was to be terminated and declared negative. If at least two responses were observed, an additional nine patients were accrued to the second stage. The study was deemed to have met its primary endpoint if confirmed responses were observed in five or more patients out of a total of 25 response-evaluable patients. Patients were deemed not assessable if they did not receive any cabozantinib or did not undergo any post-baseline protocol-defined computed tomography scan. Safety analyses were based on the intention to treat population that received at least one dose of cabozantinib. The objective response rate was calculated along with an exact 95% confidence interval. The progression-free survival, overall survival, duration of treatment, and duration of response were evaluated by the Kaplan-Meier method. The progression-free survival and overall survival were compared between subgroups by log-rank tests. The toxicity rates were calculated along with exact 95% confidence intervals. All statistical analyses were conducted in R 3.2.2. While this trial has completed the accrual of patients with RET-rearranged lung cancers, it is ongoing as several patients remain on active treatment. This trial is registered with ClinicalTrials.gov, number NCT01639508.

Cancer Center Support Grant P30 CA008748 that did not directly fund study costs. The authors wrote this article without any external funding or editorial support. AD designed the trial in cooperation with representatives of Exelixis. Exelixis was not involved in the collection, analysis, or interpretation of the data. RESULTS Between July 13, 2012 and April 30, 2016, 26 patients were enrolled. Data are presented up to June 7, 2016. Patient baseline characteristics are shown in table 1. Comprehensive molecular profiling, including next-generation sequencing in 19 cases, did not reveal concurrent activating alterations involving known targets of cabozantinib including ROS1, MET, and AXL. The most common RET fusion identified was KIF5B-RET in 16 patients (62%). Thirteen patients (50%) received one prior line of chemotherapy. No patients received a RET tyrosine kinase inhibitor prior to therapy with cabozantinib. The median follow-up time was 8·9 months (interquartile range 4.1–18.8). Twenty five of the 26 enrolled patients were evaluable for an analysis of the activity of cabozantinib. The remaining patient did not undergo repeat protocol imaging and was replaced as prespecified. This patient remains included in the toxicity analysis.

an follow-up time was 8·9 months (interquartile range 4.1–18.8). Twenty five of the 26 enrolled patients were evaluable for an analysis of the activity of cabozantinib. The remaining patient did not undergo repeat protocol imaging and was replaced as prespecified. This patient remains included in the toxicity analysis. Of the 25 evaluable patients, while no complete responses were observed, this trial met its primary endpoint with confirmed partial responses were observed in seven patients (figure 1A). No patient had disease progression as their best overall response. The overall response rate was 28% (n=7, 95% CI 12–49%). Response to therapy was observed early. Of the seven patients with a confirmed partial response, disease shrinkage of 30% or greater was noted at the first response assessment in five patients (71%). The median duration of response was 7·0 months (95% CI 3·7–38·9). Response by fusion type were as follows: confirmed responses were observed in three of 15 patients (20%) with KIF5B-RET, and two of six patients (33%) with unknown upstream partners (FISH positive). Responses were observed in patients whose tumors harbored TRIM33-RET or CLIP1-RET, and no responses were observed in patients whose tumors harbored CCDC6-RET or ERC1-RET (figure 1B).

were observed in three of 15 patients (20%) with KIF5B-RET, and two of six patients (33%) with unknown upstream partners (FISH positive). Responses were observed in patients whose tumors harbored TRIM33-RET or CLIP1-RET, and no responses were observed in patients whose tumors harbored CCDC6-RET or ERC1-RET (figure 1B). The duration of cabozantinib therapy is shown in figure 2. The median duration of treatment was 4·7 months (interquartile range 3·1–8·4). Twelve patients (48%) were treated with cabozantinib beyond six months, including four patients (16%) who were treated with cabozantinib beyond one year. At the time of analysis, four patients (16%) remained on cabozantinib, including one patient on treatment more than three years after drug was first administered. Three patients were treated with cabozantinib beyond radiologic disease progression due to ongoing clinical benefit. Of these three patients, one had asymptomatic progression in the central nervous system with the development of new brain metastases that were radiated, and two had asymptomatic extracranial radiologic progression.

nts were treated with cabozantinib beyond radiologic disease progression due to ongoing clinical benefit. Of these three patients, one had asymptomatic progression in the central nervous system with the development of new brain metastases that were radiated, and two had asymptomatic extracranial radiologic progression. At the time of the data cutoff 19 patients (76%) either had disease progression or died (figure 3a). The median progression-free survival was 5·5 months (95% CI 3·8 to 8·4). There were 16 death events. The median overall survival was 9·9 months (95% CI 8·1-not reached, figure 3b). Post-hoc exploratory analyses of the activity of cabozantinib by prior lines of therapy, previous bevacizumab exposure, and fusion type, and the activity of cabozantinib in patients with brain metastases are included in the appendix (appendix 3–7, 9). These analyses were not prespecified by the protocol.

d, figure 3b). Post-hoc exploratory analyses of the activity of cabozantinib by prior lines of therapy, previous bevacizumab exposure, and fusion type, and the activity of cabozantinib in patients with brain metastases are included in the appendix (appendix 3–7, 9). These analyses were not prespecified by the protocol. Twenty six patients treated were evaluable for toxicity. Treatment-related adverse events were predominantly grade 1 or grade 2 (table 2) and one or more drug-related toxicities of any grade were observed in 25 patients (overall toxicity rate of 96·2%, 95% CI 80·4–99·9%). The most common treatment-related adverse events of any grade were increased alanine aminotransferase, increased aspartate aminotransferase, hypothyroidism, diarrhea, palmar plantar erythrodysesthesia, and skin hypopigmentation. The most common grade 3 treatment-related adverse events were lipase elevation in four patients (15%), increased alanine aminotransferase in two patients (8%), increased alanine aminotransferase in two patients (8%), decreased platelet count in two patients (8%), and hypophosphatemia in two patients (8%). Patients in whom these toxicities were observed were asymptomatic and these adverse events resolved to grade 1 or better with dose modification. No grade 4 or grade 5 treatment-related events were observed. While no drug-related deaths were observed, 16 patients died during the course of follow up on this study. The reasons for these deaths included disease progression in 14 patients, and acute respiratory failure in two patients, one immediately following a thoracentesis, and one from suspected disease-related pulmonary embolism.

drug-related deaths were observed, 16 patients died during the course of follow up on this study. The reasons for these deaths included disease progression in 14 patients, and acute respiratory failure in two patients, one immediately following a thoracentesis, and one from suspected disease-related pulmonary embolism. Nineteen patients (73%) required a cabozantinib dose reduction due to intolerable grade 2 or grade 3 drug-related toxicities. The most common reasons for dose reduction included palmar plantar erythrodysesthesia in seven patients (37%), fatigue in three patients (16%), and diarrhea in two patients (11%). Other reasons for dose reduction included transaminitis, thrombocytopenia, proteinuria, nausea, oral mucositis, and hypertension. One dose reduction to 40 mg daily was required in 19 patients (58%), and two dose reductions to 20 mg daily were required in four patients (15%). For the majority of patients who required a dose reduction, their dose was first reduced within the first two cycles as is shown in the appendix (appendix, p 8). Two patients (8%) discontinued cabozantinib due to drug-related toxicity, specifically retroperitoneal hemorrhage in one patient (4%), and thrombocytopenia in one patient (4%).

f patients who required a dose reduction, their dose was first reduced within the first two cycles as is shown in the appendix (appendix, p 8). Two patients (8%) discontinued cabozantinib due to drug-related toxicity, specifically retroperitoneal hemorrhage in one patient (4%), and thrombocytopenia in one patient (4%). DISCUSSION We demonstrate that the multikinase RET inhibitor cabozantinib is active in patients with advanced RET-rearranged lung cancers. The overall response rate of 28% with cabozantinib is comparable to the activity of single-agent BRAF tyrosine kinase inhibitor therapy (response rate of 33% with dabrafenib) in patients with advanced BRAF V600E-mutant lung cancers19 and exceeds that of single-agent ERBB2 (HER2) tyrosine kinase inhibitor therapy in ERBB2 exon 20-mutant lung cancers (response rate of 12% with dacomitinib).20 Furthermore, it exceeds the activity of single-agent immune checkpoint inhibition (response rate of 20% with nivolumab)21 and single-agent chemotherapy (response rate of 9% with pemetrexed and 8% with docetaxel) after progression on initial platinum doublet therapy in unselected patients with advanced non-small cell lung cancers.22 Responses on this trial were brisk and durable, with two patients remaining on therapy past two and a half years.

ingle-agent chemotherapy (response rate of 9% with pemetrexed and 8% with docetaxel) after progression on initial platinum doublet therapy in unselected patients with advanced non-small cell lung cancers.22 Responses on this trial were brisk and durable, with two patients remaining on therapy past two and a half years. While clinically meaningful benefit was observed with cabozantinib, its activity was lower than that observed with ALK- and ROS1-directed tyrosine kinase inhibitor therapy (response rates of 57% and 72%) in patients with ALK- and ROS1-rearranged lung cancers, respectively.2,23 It was also lower than the response rates achieved with EGFR tyrosine kinase inhibitor therapy in treatment-naïve patients with EGFR-mutant lung cancers.24 The median PFS and the median OS of cabozantinib in RET-rearranged lung cancers were likewise lower than that observed for single-agent tyrosine kinase inhibitor therapy in EGFR-mutant, and ALK- and ROS1-rearranged lung cancers.2, 23, 24 A number of factors might account for these discrepancies.

-mutant lung cancers.24 The median PFS and the median OS of cabozantinib in RET-rearranged lung cancers were likewise lower than that observed for single-agent tyrosine kinase inhibitor therapy in EGFR-mutant, and ALK- and ROS1-rearranged lung cancers.2, 23, 24 A number of factors might account for these discrepancies. First, dose reductions were required in the majority of patients due to drug-related toxicities, as has been observed in studies of cabozantinib in other solid tumors. Cabozantinib is a multikinase inhibitor that is much more effective at inhibiting VEGFR2 (IC50 0·04 nM) than RET (IC50 5·20 nM) and its other targets including ROS1 and MET.13 Dose-limiting palmar-plantar erythrodysesthesia, gastrointestinal toxicities, and other events were likely mediated by inhibition of VEGFR2 and other kinases. While the average peak concentrations of cabozantinib at the reduced doses of 40 mg and 20 mg daily exceed the cellular IC50 required to inhibit RET,25 this nevertheless raises the possibility of decreased on-target inhibition of RET at the deliverable doses. Moving forward, it would not be unreasonable to explore alternative dosing regimens that both minimize drug-related toxicities and potentially maximize target inhibition of cabozantinib and other multikinase inhibitors with activity against RET.

ibility of decreased on-target inhibition of RET at the deliverable doses. Moving forward, it would not be unreasonable to explore alternative dosing regimens that both minimize drug-related toxicities and potentially maximize target inhibition of cabozantinib and other multikinase inhibitors with activity against RET. Second, multikinase inhibition may not be the most effective strategy for inhibiting RET fusions. RET inhibitors that are currently in clinical development for RET-rearranged lung cancers, including vandetanib, lenvatinib, sunitinib, and ponatinib, are multikinase inhibitors that, similar to cabozantinib, may be limited in their ability to inhibit RET relative to their other kinase targets.26 Alectinib, a multikinase inhibitor with activity against RET,7 may represent an agent that could be dosed to more effectively target RET due to its favorable safety profile in comparison to other RET inhibitors, however, highly RET-specific tyrosine kinase inhibitors are already in preclinical development. These RET-specific inhibitors are likely to achieve much more effective inhibition of the RET kinase in comparison to currently available RET-directed therapies. In addition, these newer drugs are likely to have a wider therapeutic window in patients, and thus may be more tolerable and amenable to chronic administration at full doses. On the other hand, the possibility that the concurrent inhibition of angiogenesis by multikinase RET inhibitors is responsible, in part, for the activity of these drugs cannot be fully discounted.

wider therapeutic window in patients, and thus may be more tolerable and amenable to chronic administration at full doses. On the other hand, the possibility that the concurrent inhibition of angiogenesis by multikinase RET inhibitors is responsible, in part, for the activity of these drugs cannot be fully discounted. Third, the biology of RET-rearranged lung cancers may dictate the need for combination therapy. As mentioned previously, similar to cabozantinib, an overall response rate of 33% can be achieved with the use of single-agent BRAF inhibition (dabrafenib) in BRAF V600E-mutant lung cancers. This response rate almost doubles to 63% with the use of combined BRAF and MEK inhibition (dabrafenib and trametinib) in comparable patients.27 Similarly, RET-rearranged lung cancers may rely on bypass pathways that are not addressed by the variety of RET inhibitors that are currently available. Lung cancers with RET fusions may also harbor additional genomic alterations that blunt the response to targeted therapy. While at least one trial is currently exploring combinatorial therapy for patients with RET-rearranged lung cancers,28 the potential for increased toxicity with dual tyrosine kinase inhibitor therapy must be kept in mind. In this respect, RET-specific tyrosine kinase inhibitors or monoclonal antibody therapy with a potentially more favorable toxicity profile in comparison to tyrosine kinase inhibitor therapy may ultimately serve as better candidates for combination treatments.

dual tyrosine kinase inhibitor therapy must be kept in mind. In this respect, RET-specific tyrosine kinase inhibitors or monoclonal antibody therapy with a potentially more favorable toxicity profile in comparison to tyrosine kinase inhibitor therapy may ultimately serve as better candidates for combination treatments. This trial has limitations. Due to tissue constraints, not all tumors underwent broad hybrid capture-based next-generation sequencing on this study. Given the hypothesis that upstream gene partners may affect response to RET inhibition, future trials would benefit from comprehensive molecular profiling that elucidates both the upstream gene partner and identifies concurrent genomic alterations that may affect response. In addition, this single-center study may not represent the breadth of patients with RET-rearranged lung cancers. Fortunately, early data from other series29–31 have confirmed that multikinase RET inhibitors are active in patients with RET-rearranged lung cancers. Lastly, confirmation of the results of our phase 2 trial in a larger group of patients will likely be required in order to obtain regulatory approval considering the response rate achieved in this series.32

ries29–31 have confirmed that multikinase RET inhibitors are active in patients with RET-rearranged lung cancers. Lastly, confirmation of the results of our phase 2 trial in a larger group of patients will likely be required in order to obtain regulatory approval considering the response rate achieved in this series.32 In conclusion, this phase 2 trial met its primary endpoint. The RET inhibitor cabozantinib can produce rapid and durable responses in patients with RET-rearranged lung cancers. Dose reductions, likely related to "off-target" toxicities due to concomitant VEGFR2 inhibition, are frequently required. We look forward to the final results from ongoing trials of other multi-kinase RET inhibitors. Furthermore, we anticipate the transition to RET-specific tyrosine kinase inhibitors in the clinic shortly. Supplementary Material 1 2 3 4 5 6 7 8 9 Funding. Exelixis, NIH/NCI Cancer Center Support Grant P30 CA008748 The authors would like to acknowledge the patients, their families, and their caregivers for their support. Declaration of interest AD has received honoraria from Exelixis. The other authors declared no conflicts of interests. Author contributions

Supplementary Material 1 2 3 4 5 6 7 8 9 Funding. Exelixis, NIH/NCI Cancer Center Support Grant P30 CA008748 The authors would like to acknowledge the patients, their families, and their caregivers for their support. Declaration of interest AD has received honoraria from Exelixis. The other authors declared no conflicts of interests. Author contributions AD designed the trial in cooperation with representatives of Exelixis. D, MA, MV, AP, MG, GR, CR, and MK were involved in the treatment of patients and data collection. NR, LW, MA, ML, JM, RS, and RSS were involved in molecular and pathologic analysis of study samples. AD and AN had access to the raw data. AN performed statistical analyses. AD wrote the first draft of the manuscript. All authors were involved in editing and writing the final draft of this manuscript. All authors agreed to submit this article for publication. AD had full access to all the data in the study and had final responsibility for the decision to submit for publication. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

f the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Figure 1 Tumour response The waterfall plot of maximal reduction in the size of indicator lesions in response to cabozantinib in 25 patients with evaluable disease is depicted by type of response (A) and by fusion type (B). Stars represent patients whose maximum reduction of disease burden was 0%. Figure 2 Duration of therapy Duration of cabozantinib therapy is shown for 25 evaluable patients. Each bar represents the period of time from the first dose to the last dose of cabozantinib. Arrows denote ongoing treatment at the time of data cutoff. White circles denote the development of radiographic progression by RECIST v1.1 in patients who were treated past progression for ongoing clinical benefit. Figure 3 Kaplan-Meier curves of progression-free survival and overall survival Both curves include 25 evaluable patients who were treated with cabozantinib. Dotted lines represent 95% CI. Table 1 Baseline Characteristics Patients with RET-rearranged lung cancers who received cabozantinib (n=26) Age median 59 (interquartile range 54–67) Sex Male 11 (42%) Female 15 (58%) Race Caucasian 19 (73%) Asian 6 (21%) African American 1 (6%) Karnofsky performance status 100% 0 90% 7 (27%) 80% 19 (73%) Cigarette smoking history Never smoker 17 (65%) >0–15 pack years 8 (31%) >15 pack years 1 (4%)

Patients with RET-rearranged lung cancers who received cabozantinib (n=26) Age median 59 (interquartile range 54–67) Sex Male 11 (42%) Female 15 (58%) Race Caucasian 19 (73%) Asian 6 (21%) African American 1 (6%) Karnofsky performance status 100% 0 90% 7 (27%) 80% 19 (73%) Cigarette smoking history Never smoker 17 (65%) >0–15 pack years 8 (31%) >15 pack years 1 (4%) Prior chemotherapy regimens 0 6 (23%) 1 13 (50%) ≥2 7 (27%) Adenocarcinoma 26 (100%) Fusion type KIF5B-RET 16 (62%) CCDC6-RET 1 (4%) TRIM33-RET 1 (4%) CLIP1-RET 1 (4%) ERC1-RET 1 (4%) Unknown (FISH-positive) 6 (22%) Brain metastases at baseline Not present 16 (62%) Present, treated 5 (19%) Present, untreated and asymptomatic 5 (19%) Table 2 Treatment-related toxicities of cabozantinib are summarized. Grade 1 and 2 adverse events observed in 10% or more of patients are listed while any grade 3, 4, or 5 event observed in any patient regardless of frequency is included.

ent 16 (62%) Present, treated 5 (19%) Present, untreated and asymptomatic 5 (19%) Table 2 Treatment-related toxicities of cabozantinib are summarized. Grade 1 and 2 adverse events observed in 10% or more of patients are listed while any grade 3, 4, or 5 event observed in any patient regardless of frequency is included. Treatment-Related Adverse Event All Grades Grade 1 Grade 2 Grade 3 Grade 4 Grade 5 Alanine aminotransferase increased 25 (96%) 21 (81%) 2 (8%) 2 (8%) 0 0 Aspartate aminotransferase increased 19 (73%) 16 (62%) 1 (4%) 2 (8%) 0 0 Hypothyroidism 18 (69%) 4 (15%) 14 (54%) 0 0 0 Diarrhea 16 (62%) 12 (46%) 4 (15%) 0 0 0 Palmar plantar erythrodysesthesia 15 (58%) 9 (31%) 6 (23%) 1 (4%) 0 0 Skin hypopigmentation 13 (50%) 0 13 (50%) 0 0 0 Platelet count decreased 13 (50%) 8(31%) 3 (12%) 2 (8%) 0 0 Fatigue 12 (46%) 4 (15%) 7 (27%) 1 (4%) 0 0 Oral mucositis 12 (46%) 11 (42%) 0 1 (4%) 0 0 Lipase increased 9 (35%) 3 (12%) 2 (8%) 4 (15%) 0 0 Nausea 8 (31%) 5 (19%) 3 (12%) 0 0 0 Dysgeusia 9 (31%) 6 (23%) 2 (8%) 0 0 0 Serum amylase increased 7 (27%) 5 (19%) 2 (8%) 0 0 0 Hypomagnesemia 7 (27%) 6 (23%) 1 (4%) 0 0 0 Vomiting 6 (23%) 6 (23%) 0 0 0 0 Weight loss 6 (23%) 5 (19%) 1 (4%) 0 0 0 Hypophosphatemia 5 (19%) 0 3 (12%) 2 (8%) 0 0 Constipation 5 (19%) 5 (19%) 0 0 0 0 Anorexia 5 (19%) 1 (4%) 4 (15%) 0 0 0 Hypertension 5 (19%) 0 4 (15%) 1 (4%) 0 0 Blood bilirubin increased 5 (19%) 3 (12%) 2 (8%) 0 0 0 Dry skin 5 (19%) 5 (19%) 0 0 0 0 Hoarseness 4 (15%) 2 (8%) 2 (8%) 0 0 0 Anemia 3 (12%) 2 (8%) 1 (4%) 0 0 0 Alopecia 3 (12%) 3 (12%) 0 0 0 0 Alkaline phosphatase increased 3 (12%) 2 (8%) 1 (4%) 0 0 0 Retroperitoneal hemorrhage 1 (4%) 0 0 1 (4%) 0 0 Hypertension 1 (4%) 0 0 1 (4%) 0 0 PANEL: RESEARCH IN CONTEXT Evidence before this study The search terms “RET rearrangement” and “non-small cell lung cancer” were used to search PubMed for articles published between 2000 and 2016. RET is a proto-oncogene. Activation by mutation and rearrangement drive oncogenesis in thyroid and lung cancers. The first reports of RET rearrangements in lung cancers were published in late 2011 and early 2012. The most common gene rearrangement in RET-rearranged lung cancers is KIF5B-RET. RET rearrangements were found predominantly in lung adenocarcinomas from patients with minimal to no prior tobacco exposure. Multikinase inhibitors were tested in RET-rearranged models in vitro and in vivo and resulted in decreased cell viability and inhibition of tumor growth.

RET-rearranged lung cancers is KIF5B-RET. RET rearrangements were found predominantly in lung adenocarcinomas from patients with minimal to no prior tobacco exposure. Multikinase inhibitors were tested in RET-rearranged models in vitro and in vivo and resulted in decreased cell viability and inhibition of tumor growth. Prior to this publication, case reports of clinical responses to the multikinase inhibitors cabozantinib and vandetanib described durable benefit with the use of these drugs. Added value of this study This study demonstrated in a prospective fashion that cabozantinib is active in patients with advanced RET-rearranged lung cancers with an overall response rate of 28%, a median progression-free survival of 6 months, and a median overall survival of 10 months. Responses were brisk and durable. Dose reductions were frequent due to drug-related adverse events. Implications of all the available evidence Launched in 2012 shortly after the first reports of RET fusions in lung cancer, this protocol represented the first prospective clinical trial initiated to establish the activity of a RET inhibitor for patients with RET-rearranged lung cancers. The activity observed was comparable to the activity of single-agent tyrosine kinase inhibition in other molecular cohorts of patients with lung cancers (dabrafenib in BRAF V600E-mutant lung cancers). It also exceeded the response rate of cabozantinib in unselected patients, and single-agent chemotherapies used in the second line setting for the treatment of non-small cell lung cancers.