Lorlatinib in patients with ALK-positive non-small-cell lung cancer: results from a global phase 2 study

Articles

Lorlatinib in patients with ALK-positive non-small-cell lung cancer: results from a global phase 2 study Benjamin J Solomon, Benjamin Besse, Todd M Bauer, Enriqueta Felip, Ross A Soo, D Ross Camidge, Rita Chiari, Alessandra Bearz, Chia-Chi Lin, Shirish M Gadgeel, Gregory J Riely, Eng Huat Tan, Takashi Seto, Leonard P James, Jill S Clancy, Antonello Abbattista, Jean-François Martini, Joseph Chen, Gerson Peltz, Holger Thurm, Sai-Hong Ignatius Ou, Alice T Shaw

Summary

Background Lorlatinib is a potent, brain-penetrant, third-generation inhibitor of ALK and ROS1 tyrosine kinases with broad coverage of ALK mutations. In a phase 1 study, activity was seen in patients with ALK-positive non-small-cell lung cancer, most of whom had CNS metastases and progression after ALK-directed therapy. We aimed to analyse the overall and intracranial antitumour activity of lorlatinib in patients with ALK-positive, advanced non-small-cell lung cancer. Methods In this phase 2 study, patients with histologically or cytologically ALK-positive or ROS1-positive, advanced, non-small-cell lung cancer, with or without CNS metastases, with an Eastern Cooperative Oncology Group performance status of 0, 1, or 2, and adequate end-organ function were eligible. Patients were enrolled into six different expansion cohorts (EXP1–6) on the basis of ALK and ROS1 status and previous therapy, and were given lorlatinib 100 mg orally once daily continuously in 21-day cycles. The primary endpoint was overall and intracranial tumour response by independent central review, assessed in pooled subgroups of ALK-positive patients. Analyses of activity and safety were based on the safety analysis set (ie, all patients who received at least one dose of lorlatinib) as assessed by independent central review. Patients with measurable CNS metastases at baseline by independent central review were included in the intracranial activity analyses. In this report, we present lorlatinib activity data for the ALK-positive patients (EXP1–5 only), and safety data for all treated patients (EXP1–6). This study is ongoing and is registered with ClinicalTrials.gov, number NCT01970865. Findings Between Sept 15, 2015, and Oct 3, 2016, 276 patients were enrolled: 30 who were ALK positive and treatment naive (EXP1); 59 who were ALK positive and received previous crizotinib without (n=27; EXP2) or with (n=32; EXP3A) previous chemotherapy; 28 who were ALK positive and received one previous non-crizotinib ALK tyrosine kinase inhibitor, with or without chemotherapy (EXP3B); 112 who were ALK positive with two (n=66; EXP4) or three (n=46; EXP5) previous ALK tyrosine kinase inhibitors with or without chemotherapy; and 47 who were ROS1 positive with any previous treatment (EXP6). One patient in EXP4 died before receiving lorlatinib and was excluded from the safety analysis set. In treatment-naive patients (EXP1), an objective response was achieved in 27 (90∙0%; 95% CI 73∙5–97∙9) of 30 patients. Three patients in EXP1 had measurable baseline CNS lesions per independent central review, and objective intracranial responses were observed in two (66∙7%; 95% CI 9∙4–99∙2). In ALK-positive patients with at least one previous ALK tyrosine kinase inhibitor (EXP2–5), objective responses were achieved in 93 (47∙0%; 39∙9–54∙2) of 198 patients and objective intracranial response in those with measurable baseline CNS lesions in 51 (63∙0%; 51∙5–73∙4) of 81 patients. Objective response was achieved in 41 (69∙5%; 95% CI 56∙1–80∙8) of 59 patients who had only received previous crizotinib (EXP2–3A), nine (32∙1%; 15∙9–52∙4) of 28 patients with one previous non-crizotinib ALK tyrosine kinase inhibitor (EXP3B), and 43 (38∙7%; 29∙6–48∙5) of 111 patients with two or more previous ALK tyrosine kinase inhibitors (EXP4–5). Objective intracranial response was achieved in 20 (87∙0%; 95% CI 66∙4–97∙2) of 23 patients with measurable baseline CNS lesions in EXP2–3A, five (55∙6%; 21∙2–86∙3) of nine patients in EXP3B, and 26 (53∙1%; 38∙3–67∙5) of 49 patients in EXP4–5. The most common treatment-related adverse events across all patients were hypercholesterolaemia (224 [81%] of 275 patients overall and 43 [16%] grade 3–4) and hypertriglyceridaemia (166 [60%] overall and 43 [16%] grade 3–4). Serious treatment-related adverse events occurred in 19 (7%) of 275 patients and seven patients (3%) permanently discontinued treatment because of treatment-related adverse events. No treatment-related deaths were reported. Interpretation Consistent with its broad ALK mutational coverage and CNS penetration, lorlatinib showed substantial overall and intracranial activity both in treatment-naive patients with ALK-positive non-small-cell lung cancer, and in those who had progressed on crizotinib, second-generation ALK tyrosine kinase inhibitors, or after up to three previous ALK tyrosine kinase inhibitors. Thus, lorlatinib could represent an effective treatment option for patients with ALK-positive non-small-cell lung cancer in first-line or subsequent therapy. Funding Pfizer. Copyright © 2018 Elsevier Ltd. All rights reserved. www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

Lancet Oncol 2018 Published Online November 6, 2018 http://dx.doi.org/10.1016/ S1470-2045(18)30649-1 See Online/Comment http://dx.doi.org/10.1016/ S1470-2045(18)30789-7 Peter MacCallum Cancer Centre, Melbourne, VIC, Australia (Prof B J Solomon MBBS); Gustave Roussy Cancer Campus, Villejuif, France (Prof B Besse MD); Department of Cancer Medicine, Paris-Sud University, Orsay, France (Prof B Besse); Sarah Cannon Cancer Research Institute/ Tennessee Oncology, PLLC, Nashville, TN, USA (T M Bauer MD); Vall d’Hebron Institute of Oncology, Barcelona, Spain (E Felip MD); National University Hospital Singapore, Singapore (R A Soo MD); Medical Oncology Department, University of Colorado, Aurora, CO, USA (Prof D R Camidge MD); Santa Maria della Misericordia Hospital, Azienda Ospedaliera di Perugia, Perugia, Italy (R Chiari MD); National Institute for Cancer Research, Aviano, Italy (A Bearz MD); National Taiwan University Hospital, Taipei, Taiwan (C-C Lin MD); Barbara Ann Karmanos Cancer Institute, University of Michigan, Ann Arbor, MI, USA (Prof S M Gadgeel MD); Memorial Sloan Kettering Cancer Center, New York, NY, USA (G J Riely MD); National Cancer Center, Singapore (E H Tan MBBS); National Kyushu Cancer Center, Fukuoka, Japan (Prof T Seto MD); Pfizer Oncology, New York, NY, USA (L P James MD, J Chen PharmD); Pfizer Oncology, Cambridge, MA, USA (J S Clancy BS); Pfizer Oncology, Milan, Italy (A Abbattista BSc); Pfizer Oncology, La Jolla, CA, USA (J-F Martini PhD, H Thurm MD); Pfizer Oncology, Groton, CT,

1

Articles

USA (G Peltz MD); University of California Irvine, Irvine, CA, USA (Prof S-H Ignatius Ou MD); and Massachusetts General Hospital, Boston, MA, USA (Prof A T Shaw MD) Correspondence to: Prof Benjamin J Solomon, Peter MacCallum Cancer Centre, Melbourne, VIC 3000, Australia [email protected]

Research in context Evidence before this study Although most patients with anaplastic lymphoma kinase (ALK)-positive non-small-cell lung cancer derive substantial clinical benefit from the first-in-class ALK tyrosine kinase inhibitor crizotinib, acquired resistance invariably develops. Second-generation ALK tyrosine kinase inhibitors such as ceritinib, alectinib, and brigatinib have been developed that have both systemic and intracranial antitumour activity with acceptable safety profiles in patients previously treated with crizotinib. Treatment options after failure of a second-generation ALK inhibitor have not been defined, and treatment with standard therapies, such as single-agent chemotherapy, has been minimally effective in this setting, while the efficacy of chemotherapy combinations such as platinum and pemetrexed has yet to be established. Therefore, an urgent need remains for additional active ALK tyrosine kinase inhibitors with broader coverage of ALK kinase domain mutations and improved CNS penetration in patients whose disease has progressed on a second-generation ALK tyrosine kinase inhibitor. Lorlatinib is a novel, highly potent, third-generation, macrocyclic inhibitor of ALK and c-ros oncogene 1 (ROS1) tyrosine kinases, which has activity in the CNS as well as broad-spectrum potency against most known ALK kinase domain resistance mutations that can develop during treatment with existing first-generation and second-generation ALK tyrosine kinase inhibitors. In the

Introduction Rearrangements of the anaplastic lymphoma kinase (ALK) gene are found in 3–5% of non-small-cell lung cancers, and represent a clinically and molecularly distinct subtype that shows sensitivity to therapy with ALK tyrosine kinase inhibitors.1–3 Standard initial treatment for ALK-positive non-small-cell lung cancer includes first-line crizotinib4 and, more recently, first-line alectinib or ceritinib.5,6 However, most patients treated with crizotinib relapse over time because of acquired resistance, which occurs through several molecular mechanisms including secondary mutations of the ALK kinase domain and ALK-independent mechanisms.7 More potent, second-generation ALK tyrosine kinase inhibitors were developed to overcome crizotinib resistance and have demonstrated clinical benefit in patients with treatment-naive and crizotinib-refractory ALK-positive non-small-cell lung cancer.5,6,8,9 However, most patients will develop resistance to secondgeneration tyrosine kinase inhibitors or develop disease progression in the CNS.10,11 Lorlatinib (PF-06463922; Pfizer Oncology, Groton, CT, USA) is a novel, highly potent, selective third-generation tyrosine kinase inhibitor directed at the ALK and c-ros oncogene 1 (ROS1) kinases. Using structure-based drug design, this macrocyclic tyrosine kinase inhibitor was developed to penetrate the blood–brain barrier and to retain potency against most known ALK resistance 2

phase 1 portion of this phase 1–2 study, lorlatinib was generally well tolerated and showed both systemic and intracranial antitumour activity among patients with ALK-positive advanced non-small-cell lung cancer, most of whom had CNS metastases and had failed to respond to at least one previous ALK tyrosine kinase inhibitor. Added value of this study In phase 2 of this phase 1–2 study, treatment with lorlatinib led to frequent and durable responses in patients with ALK-positive non-small-cell lung cancer who were treatment naive or who had previously received at least one ALK tyrosine kinase inhibitor (including the second-generation drugs ceritinib, alectinib, and brigatinib). Substantial intracranial activity was observed in this cohort of patients, in whom brain metastases were frequent. Implications of all the available evidence Lorlatinib is a treatment option for patients with ALK-positive advanced non-small-cell lung cancer who are treatment naive or have been previously treated with at least one ALK tyrosine kinase inhibitor, including second-generation drugs, and who might have brain metastases. Based on these results, lorlatinib could represent a useful therapeutic option for these patients in first-line or subsequent therapy.

mutations that can develop during treatment with crizotinib and second-generation tyrosine kinase inhibitors, including the ALK Gly1202Arg solvent front mutation, which is located at the solvent-front region (ie, the solvent-facing surface) of ALK and can impair drug binding through steric hindrance.11,12 In the phase 1 portion of this phase 1–2 study, lorlatinib treatment led to a high proportion of patients achieving an objective response (19 [46%] of 41) and durable responses (median duration of response 12·4 months [95% CI 6·5–not reached]) in patients with ALK-positive non-small-cell lung cancer, many of whom were heavily pretreated and had CNS metastases. Responses were recorded in patients who had previously received a second-generation tyrosine kinase inhibitor as well as in those who had previously only received crizotinib. Additionally, a preliminary analysis of paired cere­ brospinal fluid and plasma samples demonstrated a high degree of penetration across the blood–brain barrier. Adverse events reported with lorlatinib were pre­ dominantly grade 1 or 2 in severity, with hyperchol­ esterolaemia and hypertri­glyceridaemia among the most frequently reported.13 On the basis of the phase 1 and preliminary phase 2 data, lorlatinib received accelerated approval from the US Food and Drug Administration (FDA) on Nov 2, 2018, for the treatment of patients with ALK-positive metastatic non-small-cell lung cancer whose disease has progressed

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

Articles

on crizotinib and at least one other ALK inhibitor for metastatic disease, or alectinib as the first ALK inhibitor therapy for metastatic disease, or ceritinib as the first ALK inhibitor therapy for metastatic disease. In this report, we present findings from the phase 2 portion of the phase 1–2 trial that investigated the activity of singleagent lorlatinib in patients with ALK-positive, advanced, non-small-cell lung cancer.

Methods

Study design and participants This ongoing, multicentre, open-label, single-arm phase 2 trial of lorlatinib was done at 47 centres across 13 countries (appendix p 3). Eligible patients (aged ≥18 years) had a histologically or cytologically confirmed diagnosis of metastatic non-small-cell lung cancer that carried either an ALK or ROS1 gene rearrangement. Positivity for ALK gene rearrangement was determined locally based on the FDA-approved fluorescence in-situ hybridisation (FISH) assay (Abbott Molecular, Abbott Park, IL, USA) or by immuno­histochemistry (Ventana Medical Systems, Tucson, AZ, USA). ROS1 positivity was established by FISH or reverse transcriptase PCR or nextgeneration sequencing via a local diagnostic test. Patients were required to have at least one measurable target extracranial lesion according to Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. All patients were to have archival or de-novo tissue sample available and collected before enrolment. Asymptomatic treated or untreated CNS metastases were permitted. Additionally, all patients were required to have a baseline Eastern Cooperative Oncology Group (ECOG) performance status of 0, 1, or 2, and have adequate bone marrow (absolute neutrophil count ≥1·5 × 10⁹/L, platelets ≥100 × 10⁹/L, and haemoglobin ≥9 g/dL), pancreatic (serum amylase [pancreatic isoenzyme] ≤1·5  ×  upper limit of normal [ULN] and serum lipase ≤1·5  ×  ULN), renal (serum creatinine ≤1·5 × ULN or estimated creatinine clearance ≥60 mL/min as calculated using the method standard for the institution), and liver (total serum bilirubin ≤1·5 × ULN, and aspartate aminotransferase and alanine aminotransferase ≤2·5 × ULN [≤5·0 × ULN in the event of liver metastases]) function. Patients were not eligible for inclusion if they had spinal cord compression; active and clinically significant bacterial, fungal, or viral infection; clinically significant cardiovascular disease; predisposing characteristics for acute pancreatitis such as uncontrolled hyperglycaemia or current gallstone disease; history of extensive, dis­ seminated, bilateral, or grade 3–4 (National Cancer Institute Common Terminology Criteria for Adverse Events) interstitial fibrosis or interstitial lung disease; severe, acute, or chronic psychiatric conditions; acute malignant disease (other than non-small-cell lung cancer, non-melanoma skin cancer, in-situ cervical cancer, papillary thyroid cancer, ductal carcinoma in situ of the breast, or localised and presumed cured prostate cancer)

within the past 3 years; active inflammatory gastro­ intestinal disease, chronic diarrhoea, symptomatic diverticular disease, previous gastric resection, or gastric band; or abnormal left ventricular ejection fraction. The full list of inclusion and exclusion criteria is in the appendix (pp 4–5). Patients were enrolled into expansion cohorts (EXP) on the basis of their ALK or ROS1 status and previous treatment history. Specific enrolment criteria for each expansion cohort were as follows: ALK-positive, treatmentnaive patients (EXP1); ALK-positive patients with disease progression following previous crizotinib only (EXP2); ALK-positive patients with disease progression following previous crizotinib and one or two regimens of chemo­ therapy given before or after crizotinib (EXP3A); ALKpositive patients with disease progression following one previous non-crizotinib ALK tyrosine kinase inhibitor with any number of chemo­therapy regimens (EXP3B); ALKpositive patients with disease progression following two (EXP4) or three (EXP5) previous ALK tyrosine kinase inhibitors with any number of chemotherapy regimens. The institutional review board or independent ethics committee at each participating centre approved the protocol, which complied with the International Ethical Guidelines for Biomedical Research Involving Human Subjects, Good Clinical Practice guidelines, the Declaration of Helsinki, and local laws. The protocol is included in the appendix (pp 15–214). All patients provided written, informed consent before participation.

See Online for appendix

Procedures Lorlatinib was administered orally in a tablet form at a starting dose of 100 mg once daily continuously in 21-day cycles. Treatment continued until investigator-assessed disease progression, unacceptable toxicity, withdrawal of consent, or death. Patients were allowed to continue treatment with lorlatinib after objective progression as long as there was evidence of clinical benefit in the investigator’s opinion. Dose delays and reductions were permitted to manage toxicities per investigator discretion. Patients who required more than three dose reductions were discontinued from treatment. Visits were scheduled to occur on days 1, 8, and 15 of the first 21-day cycle and then every 3 weeks (reduced to every 6 weeks after 38 cycles [more than 30 months]), at treatment discontinuation, and at 28–35 days posttreatment. Follow-up for overall survival and subsequent therapy continued every 2 months after treatment discontinuation. All patients underwent baseline tumour imaging, including CT scans of the chest, abdomen, and pelvis, and brain imaging by MRI. CT and MRI scans were to be done every 6 weeks for the first 30 months and then every 12 weeks thereafter until disease progression or the start of a new anticancer therapy. CNS lesions could be selected as targeted lesions provided the individual

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

3

Articles

lesions were a minimum of 5 mm using gadolinium contrast enhancement and in a site of no previous radiotherapy or progression following radiotherapy.14 Responses were confirmed 4 or more weeks after initial response (assessed according to RECIST version 1.1). Bone scans or bone MRIs were required at baseline and were repeated every 12 weeks during treatment only if bone disease was identified. Peripheral blood samples were collected at screening, at the beginning of cycle 3, and at the end of treatment visit for circulating cell-free DNA analysis of ALK gene rearrangements and kinase domain mutations using next-generation sequencing (Guardant360, Guardant Health, Redwood City, CA, USA). Additionally, a baseline de-novo biopsy was to be done at screening if deemed safe and feasible for assessment of biomarkers potentially associated with sensitivity or resistance to lorlatinib. Safety assessments were done in all patients at baseline and at every subsequent visit. Adverse events were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03 and were assessed from the start of treatment until at least 28 days after the final lorlatinib administration. Additional assessments of cognitive function using the Cogstate assessment tool (Cogstate Inc, New Haven, CT, USA), mood using the Beck Depression Inventory-II scale, and suicidal ideation and behaviour using the ColumbiaSuicide Severity Rating Scale were collected at the beginning of each cycle up to cycle 6 and then every other cycle, and were analysed by Cogstate. Further details of the study procedures are included in the appendix and the protocol.

Outcomes The primary endpoint was objective tumour response (defined as a confirmed complete response or partial response) and intracranial tumour response according to modified RECIST version 1.1, which allowed for up to five CNS target lesions, as assessed by independent central radiology review (ICR) and assessed in pooled subgroups of ALK-positive patients (ie, EXP1, EXP2–3A, EXP3B, EXP4–5 and EXP2–5). Key secondary endpoints included: duration of response, defined as the time from the first documen­ tation of objective tumour response to the first docu­ mentation of disease progression or to death from any cause; intracranial duration of response; time to first tumour response, defined as the time from first dose to first documentation of objective tumour response; progression-free survival, defined as the time from first dose to first documentation of objective disease progression or to death on study due to any cause; safety and tolerability including adverse events, vital signs, and laboratory abnormalities; and further assessment of mood, suicidal ideation and behaviour, and cognitive function. Selected molecular profiling of circulating free DNA and tumour tissue, which was also a prespecified 4

secondary endpoint, will be reported at a later date. Details of further exploratory endpoints are listed in the appendix. The secondary endpoint patient-reported outcomes (PROs) was also assessed using the European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire–Core 30 (QLQ-C30) and its corresponding Lung Cancer Module (QLQ-LC13). Other secondary endpoints (not reported here) were CSF concentration of lorlatinib, QTc interval, left ventricular ejection fraction, disease control rate, overall survival, probability of CNS progression, time to progression and response to previous therapies, and pharmacokinetics.

Statistical analysis The sample size of each cohort was based on an estimation design with no specific hypothesis testing. EXP1 had a target enrolment of 30 patients. Cohorts EXP2 and EXP3 were originally regarded as similar and thus had a combined target enrolment of 80 patients. Most patients to be enrolled in the study were expected to have been pretreated with crizotinib and a second generation ALK TKI (EXP4), so the target enrolment was set to 70 patients, while EXP5 and EXP6 were set targets of 40 patients each. We decided to split EXP3 into EXP3A and EXP3B and then subsequently pool EXP2 with EXP3A, but keep EXP3B separately because patients in EXP2 and EXP3A had all been treated with crizotinib (first generation ALK TKI) as their only ALK TKI, whereas patients in EXP3B had been pre-treated with a second generation ALK TKI only. Patients in EXP4 and EXP5 were exposed to two or three previous ALK TKIs, most of which were crizotinib and one or two second generation ALK TKIs. These cohorts represented the more advanced lines of treatment where no treatment options with other ALK TKIs would be available; thus a pooled analysis of efficacy was planned on these two last-line cohorts. Analyses of activity and safety in this report were based on the safety analysis set (ie, all patients who received at least one dose of lorlatinib, as assessed by ICR). Patients with measurable CNS metastases at baseline by ICR were included in the intracranial activity analyses. The PRO-evaluable analysis set was defined as all enrolled patients who received at least one dose of lorlatinib and completed a baseline and at least one post-baseline PRO assessment. The proportions of patients with objective response and objective intracranial response were defined as those who achieved a confirmed complete response or partial response according to RECIST version 1.1 in the safety analysis as their best overall or intracranial response, respectively. The corresponding 95% CIs were calculated using the exact method based on the binomial distribution. For time-to-event endpoints, such as duration of response and progression-free survival, we estimated median values and two-sided 95% CIs using Kaplan-Meier methods. Time to first

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

Articles

276 patients enrolled

1 excluded (died before receiving first dose)

275 patients enrolled and received at least one dose of lorlatinib

228 ALK-positive patients

30 treatment-naive patients (EXP1)

27 patients treated with previous crizotinib only (EXP2)

32 patients treated with previous crizotinib and chemotherapy (EXP3A) 28 patients treated with one second-generation ALK TKI with or without chemotherapy (EXP3B)

47 ROS1-positive patients

65 patients treated with two previous ALK TKIs* with or without chemotherapy (EXP4)

46 patients treated with three previous ALK TKIs* with or without chemotherapy (EXP5)

47 patients treated with any line of treatment (EXP6)

Figure 1: Summary of expansion cohorts ALK=anaplastic lymphoma kinase. EXP=expansion cohort. ROS=c-Ros oncogene. TKI=tyrosine kinase inhibitor. *Lines of therapy; if the same TKI was given twice, it was counted as two previous lines of treatment.

tumour response and PROs were summarised with descriptive statistics. The data cutoff for this analysis was March 15, 2017. All analyses were done using SAS version 9.4. This study is registered with ClinicalTrials. gov, number NCT01970865.

Role of the funding source This study was designed by the sponsor (Pfizer), study investigators, and members of the steering committee. Data were collected by investigators and analysed by the sponsor. All authors, including those employed by the sponsor of the study, contributed to the interpretation of the data and the development, writing, and approval of the manuscript. Medical writing support was funded by the sponsor. All authors had full access to the raw data in the study, and the corresponding author had final responsibility for the decision to submit for publication.

Results Between Sept 15, 2015, and Oct 3, 2016, 276 patients were enrolled across all cohorts (EXP1–6), of whom 275 received at least one dose of lorlatinib and were evaluable for safety (figure 1); one patient enrolled in EXP4 died before receiving lorlatinib and was excluded from all analyses. Here, we present lorlatinib activity data for the ALK-positive patients (EXP1–5 only), and safety data for all treated patients (EXP1–6). Antitumour activity results for the ROS1 cohort (EXP6) will be reported separately. Baseline characteristics of the enrolled patients are shown in table 1. At the time of enrolment, CNS meta­ stases were present in eight (27%) of 30 treatment-naive, ALK-positive patients (EXP1), in 133 (67%) of 198 patients

who received at least one previous ALK tyrosine kinase inhibitor (pooled cohorts EXP2–5), and 25 (53%) of 47 ROS1-positive patients (EXP6), most of whom underwent previous brain radiotherapy that could have included whole-brain radiotherapy, stereotactic radio­ surgery, or both (data not shown). Most patients were white or of Asian ethnicity and had an ECOG performance status of 0 or 1. Approximately two-thirds of ALK-positive patients had been treated with at least one previous ALK tyrosine kinase inhibitor and 36 (77%) ROS1-positive patients had also received previous chemotherapy (table 1). 30 patients had not previously received systemic treatment for their advanced or metastatic disease (EXP1). Of these 30 patients, 27 (90∙0%; 95% CI 73∙5–97∙9) had an objective response, with one patient achieving a complete response and 26 achieving a partial response. Of these 27 confirmed responses, 23 (85%) were ongoing and the median duration of response was not reached (95% CI 10∙0 months–not reached [NR]; table 2, figure 2A). Median time to first tumour response was 1·4 months (IQR 1·3–2·7). The estimated median duration of follow-up for response was 6∙9 months (IQR 5·6–12·5; table 2). Three patients in EXP1 had measurable baseline CNS lesions per ICR, and objective intracranial responses (both partial responses) were observed in two (66∙7%; 95% CI 9∙4–99∙2; table 2, figure 3A). Median intracranial duration of response was not reached (95% CI NR–NR) and median progressionfree survival was not reached (95% CI 11·4–NR) at the time of the analysis (figure 4). A total of 198 patients received at least one previous ALK tyrosine kinase inhibitor before enrolling in the

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

5

Articles

Treatment naive Previous (EXP1; n=30) crizotinib with or without chemotherapy (EXP2–3A; n=59)

Previous non-crizotinib ALK TKI with or without chemotherapy (EXP3B; n=28)

≥2 previous ALK TKIs* with or without chemotherapy (EXP4–5; n=111)

Pooled activity ROS1-positive Pooled safety group group patients (EXP2–5; n=198) (EXP6; n=47) (EXP1–6; n=275)

51∙0 (43·0–59·0)

53∙0 (44·0–62·0)

Age, years 59∙0 (48·0–68·0)

54∙0 (46·0–66·0)

54∙0 (46·5–64·0)

Mean (SD)

57∙4 (12∙1)

54∙9 (12∙5)

55∙0 (11∙6)

51∙9 (11∙5)

53∙2 (11∙9)

52·8 (12·9)

53∙6 (12∙1)

Range

27–75

30–85

33–77

29–83

29–85

19–77

19–85

Female

13 (43%)

39 (66%)

16 (57%)

62 (56%)

117 (59%)

27 (57%)

157 (57%)

Male

17 (57%)

20 (34%)

12 (43%)

49 (44%)

81 (41%)

20 (43%)

118 (43%)

31 (53%)

7 (25%)

59 (53%)

97 (49%)

25 (53%)

132 (48%)

1 (1%)

1 (2%)

3 (1%)

70 (35%)

16 (34%)

103 (37%)

Median (IQR)

54·0 (44·0–61·0)

54∙0 (45.0–62.0)

Sex

Race White

10 (33%)

Black

1 (3%)

Asian

17 (57%)

0 17 (29%)

1 (4%) 16 (57%)

0 37 (33%)

Other

1 (3%)

2 (3%)

1 (4%)

5 (5%)

8 (4%)

3 (6%)

12 (4%)

Unspecified†

1 (3%)

9 (15%)

3 (11%)

10 (9%)

22 (11%)

2 (4%)

25 (9%)

0

13 (43%)

28 (47%)

15 (54%)

46 (41%)

89 (45%)

17 (36%)

119 (43%)

1

16 (53%)

30 (51%)

13 (46%)

59 (53%)

102 (52%)

28 (60%)

146 (53%)

ECOG performance status

2 Brain metastases present at baseline‡

1 (3%)

1 (2%)

8 (27%)

37 (63%)

6 (5%)

7 (4%)

2 (4%)

10 (4%)

13 (46%)

0

83 (75%)

133 (67%)

25 (53%)

166 (60%)

Number of brain metastases at baseline‡ 1–3

4 (50%)

13 (35%)

4 (31%)

34 (41%)

51 (38%)

10 (40%)

65 (39%)

4–6

2 (25%)

12 (32%)

6 (46%)

25 (30%)

43 (32%)

11 (44%)

56 (34%)

7–9

2 (25%)

7 (19%)

3 (23%)

14 (17%)

24 (18%)

2 (8%)

28 (17%)

≥10

0

5 (14%)

0

10 (12%)

15 (11%)

2 (8%)

17 (10%)

Median

3 (1–6)

5 (3–7)

6 (3–6)

4 (2–7)

5 (2–7)

4 (2–5)

5 (2–7)

Previous radiotherapy

6 (20%)

30 (51%)

12 (43%)

83 (75%)

125 (63%)

23 (49%)

154 (56%)

Previous brain-directed radiotherapy

2 (7%)

19 (32%)

8 (29%)

59 (53%)

86 (43%)

15 (32%)

103 (37%)

105 (38%)

Number of previous chemotherapy regimens 0

29 (97%)

24 (41%)

15 (54%)

26 (23%)

65 (33%)

11 (23%)

1

1 (3%)§

30 (51%)

10 (36%)

43 (39%)

83 (42%)

12 (26%)

96 (35%)§

2

0

2 (3%)

2 (7%)

26 (23%)

30 (15%)

13 (28%)

43 (16%)

3

0

3 (5%)

1 (4%)

8 (7%)

12 (6%)

10 (21%)

22 (8%)

≥4

0

0

0

8 (7%)

8 (4%)

1 (2%)

9 (3%)

Number of previous ALK or ROS1 TKI regimens 0 1

30 (100%) 0

0 59 (100%)

0 28 (100%)

0 0

0 87 (44%)

13 (28%)

43 (16%)

30 (64%)

117 (43%)

2

0

0

0

65 (59%)

65 (33%)

2 (4%)

67 (24%)

3

0

0

0

42 (38%)

42 (21%)

2 (4%)

44 (16%)

≥4

0

0

0

4 (4%)

4 (2%)

0

4 (1%)

Data are n (%) unless indicated otherwise· ALK=anaplastic lymphoma kinase. ECOG=Eastern Cooperative Oncology Group. EXP=expansion cohort. ROS1=c-ros oncogene 1. TKI=tyrosine kinase inhibitor. *Lines of therapy; if the same TKI was given twice, it was counted as two previous lines of treatment. †In France, information about race was not allowed to be collected per local regulations. ‡By independent central review; includes measurable and non-measurable CNS lesions at baseline. §One patient in EXP1 received previous adjuvant chemotherapy but no previous treatment for metastatic disease.

Table 1: Baseline characteristics and patient demographics (safety analysis set)

study (these patients comprised those enrolled in cohorts EXP2–5). Of these patients, 77 (39%) received crizotinib as their last previous tyrosine kinase inhibitor, 62 (31%) 6

received alectinib, 47 (24%) received ceritinib, eight (4%) received brigatinib, and four (2%) received another ALK tyrosine kinase inhibitor (ensartinib [Xcovery] or

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

Articles

Treatment naive (EXP1)

Previous crizotinib with or without chemotherapy (EXP2–3A)

Previous non-crizotinib ALK TKI with or without chemotherapy (EXP3B)

≥2 previous ALK TKIs* with or without chemotherapy (EXP4–5)

≥1 previous ALK TKI with or without chemotherapy (pooled EXP2–5)

30

59

28

111

198

Overall responses Number of patients Best overall response Complete response† Partial response†

1 (3%)

1 (2%)

1 (4%)

2 (2%)

4 (2%)

26 (87%)

40 (68%)

8 (29%)

41 (37%)

89 (45%) 58 (29%)

Stable disease

2 (7%)

10 (17%)

10 (36%)

38 (34%)

Objective progression

1 (3%)

6 (10%)

7 (25%)

20 (18%)

33 (17%)

Indeterminate

0

2 (3%)

2 (7%)

10 (9%)

14 (7%)

9 (32∙1%; 15∙9–52∙4)

43 (38∙7%; 29∙6–48∙5)

93 (47∙0%; 39∙9–54∙2)

Patients with confirmed objective response (%; 95% CI)‡ Median time to first tumour response, months (IQR) Median duration of response, months (95% CI)§ Median duration of follow-up for response, months (IQR)¶

27 (90∙0%; 73∙5–97∙9)

41 (69∙5%; 56∙1–80∙8)

1∙4 (1∙3–2∙7)

1∙4 (1∙3–2∙6)

NR (10∙0–NR)

NR (11∙1–NR)

6∙9 (5∙6–12∙5)

6∙9 (4·2–7∙0)

1∙4 (1∙4–2∙7) NR (4∙1–NR) 7∙0 (5·6–8·3)

1∙4 (1∙4–2∙9) NR (5∙5–NR) 7∙2 (5·6–9·8)

1∙4 (1∙3–2∙7) NR (11∙1–NR) 6∙9 (5·6–8∙3)

Intracranial responses Number of patients||

3

23

9

49

81 16 (20%)

Best overall intracranial response Complete response†

0

5 (22%)

1 (11%)

10 (20%)

Partial response†

2 (67%)

15 (65%)

4 (44%)

16 (33%)

35 (43%)

Stable disease

1 (33%)

3 (13%)

0

17 (35%)

20 (25%)

Objective progression

0

0

3 (33%)

4 (8%)

7 (9%)

Indeterminate

0

0

1 (11%)

2 (4%)

3 (4%)

Patients with confirmed intracranial objective response (%; 95% CI)‡ Median time to first intracranial response, months (IQR) Median duration of intracranial response, months (95% CI)§

2 (66∙7%; 9∙4–99∙2) 2∙0 (1∙2–2∙7) NR (NR–NR)

20 (87∙0%; 66∙4–97∙2) 1∙4 (1∙3–1∙4) NR (8∙4–NR)

5 (55∙6%; 21∙2–86∙3) 1∙4 (1∙4–2∙6) NR (4∙1–NR)

26 (53∙1%; 38∙3–67∙5)

51 (63∙0%; 51∙5–73∙4)

1∙4 (1∙3–3∙1)

1∙4 (1∙3–2∙7)

14∙5 (6∙9–14∙5)

14∙5 (8∙4–14∙5)

Data are n (%) unless specified otherwise. ALK=anaplastic lymphoma kinase. EXP=expansion cohort. NR=not reported. TKI=tyrosine kinase inhibitor. *Lines of therapy; if the same TKI was given twice, it was counted as two previous lines of treatment. †Confirmed response. ‡Using exact method based on binomial distribution. §Using Brookmeyer and Crowley method. ¶Estimates are based on the reverse Kaplan-Meier method with 95% CIs based on the Brookmeyer and Crowley method. ||Number of patients with at least one measurable CNS lesion at baseline.

Table 2: Overall and intracranial responses by independent central review

entrectinib [Roche]). The dose of previous ALK tyrosine kinase inhibitors was not captured. Of these 198 patients, objective responses were observed in 93 patients (47∙0%; 95% CI 39∙9–54∙2), with four complete responses and 89 partial responses; 58 (29%) patients had stable disease (table 2; appendix p 13). At the time of analysis, 63 (68%) of the 93 confirmed responses were ongoing and the median duration of response had not been reached (95% CI 11·1–NR). The estimated median duration of follow-up for response was 6∙9 months (IQR 5·6–8·3; table 2). Among patients in EXP2–5 with CNS lesions at baseline per ICR, the median number was five lesions (both target and non-target) per patient, IQR 2–7. Of the 81 patients with measurable baseline CNS lesions per ICR, intracranial responses were observed in 51 patients (63∙0%; 95% CI 51∙5–73∙4; table 2, appendix p 14). The estimated

median intracranial duration of response for patients with measurable baseline CNS lesions was 14∙5 months (95% CI 8∙4–14∙5), with 37 (73%) of those 51 responses ongoing at the time of analysis. Both overall and intra­ cranial responses to lorlatinib were rapid in onset, with a median time to overall first tumour response of 1∙4 months (IQR 1·3–2·7) and a median time to first intracranial response of 1·4 months (IQR 1·3–2·7; table 2). Although the proportion of patients with intra­ cranial response was higher than the proportion with extracranial response, the overall proportion with objective response was not solely driven by intracranial responses (appendix p 6). Median progression-free survival was 7∙3 months (95% CI 5∙6–11∙0) and 97 (49%) of 198 patients were censored at the time of analysis (figure 4). 23 (37·1%; 95% CI 25∙2–50∙3) of 62 patients who received alectinib as the last previous ALK tyrosine kinase

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

7

Articles

Change in tumour size from baseline (%)

A 80 70 60 50 40 30 20 10 0 –10 –20 –30 –40 –50 –60 –70 –80 –90 –100

* *

Change in tumour size from baseline (%)

*

**

* *

**

* *

* *

*

****

* *

*

** *

**

*

*

D

C 80 70 60 50 40 30 20 10 0 –10 –20 –30 –40 –50 –60 –70 –80 –90 –100

B

Best overall response Complete response Partial response Stable disease/no response Objective progression Indeterminate * Off treatment or PD occurred

*

*

* * * * *

* ***

* *

* *

*

* *

*

*

*

*

***** *** ** ** ******* *** ****** * *** *** ** *

**** **

**** **

*** * * ** *

* ** *

Figure 2: Best percentage change in tumour size from baseline (A) EXP1: treatment-naive patients. (B) EXP2–3A: previous crizotinib with or without 1–2 chemotherapy regimens. (C) EXP3B: previous non-crizotinib ALK TKI with or without chemotherapy. (D) EXP4–5: two or more previous ALK TKIs with or without chemotherapy. Each bar represents an individual patient’s maximum reduction in target lesion size, as assessed by ICR. Patients with at least one on-study target lesion assessment as per ICR were included. If any procedure was different and not interchangeable from the procedure at screening, the percentage change from baseline could not be calculated and is not displayed. Colours indicate patients’ best overall responses per Response Evaluation Criteria in Solid Tumors version 1.1 according to ICR. Complete response was defined as the disappearance of all target lesions; when nodal disease was included in target lesions, reversion to normal node size (<10 mm) prevented the percent change from baseline from reaching –100%. Some patients with a total change from baseline of –100% are shown as having partial responses because of the presence of non-target lesions in the summary. ALK=anaplastic lymphoma kinase. EXP=expansion cohort. ICR=independent central review. PD=progressive disease. TKI=tyrosine kinase inhibitor.

inhibitor before lorlatinib achieved an objective response, compared with 19 (40∙4%; 26∙4–55∙7) of 47 who received ceritinib and three (37·5%; 8∙5–75∙5) of eight who received brigatinib. Responses by the type of secondgeneration ALK tyrosine kinase inhibitor received before lorlatinib are shown in the appendix (p 7). Of the 59 patients who received previous crizotinib only (with or without previous chemotherapy; EXP2–3A), 41 (69∙5%; 95% CI 56∙1–80∙8) achieved an objective response, with one patient achieving a complete response and 40 patients achieving a partial response (table 2, figure 2B). At the time of analysis, 32 (78%) of the 41 confirmed responses were ongoing and the median duration of response had not been reached (95% CI 11·1–NR). The median duration of follow-up for response was 6∙9 months (IQR 4·2–7·0; table 2). 37 patients had baseline CNS lesions per ICR, with a median of five CNS lesions (both target and non-target) per patient (IQR 3–7). Of 23 patients with measurable baseline CNS lesions per ICR, intracranial responses were observed in 20 (87∙0%; 95% CI 66∙4–97∙2; table 2, figure 3B). Median intracranial duration of response was not reached (95% CI 8·4–NR) at the time of analysis. The median time to first tumour response was 1∙4 months 8

(IQR 1·3–2·6), the median time to first intracranial response was 1·4 months (1·3–1·4), and the median progression-free survival was not reached (95% CI 12∙5 months–NR) at the time of analysis (figure 4). 28 patients received one non-crizotinib ALK tyrosine kinase inhibitor before enrolment (EXP3B). Of these 28 patients, 13 (46%) each received alectinib and ceritinib as the last ALK tyrosine kinase inhibitor before lorlatinib, one (4%) received brigatinib, and one (4%) received another experimental ALK tyrosine kinase inhibitor (entrectinib). Nine patients (32∙1%; 95% CI 15∙9–52∙4) had an objective response, including one complete response and eight partial responses (table 2, figure 2C). At the time of analysis, five (56%) of the nine confirmed responses were ongoing and the median duration of response had not been reached (95% CI 4·1–NR). The median duration of follow-up for response was 7∙0 months (IQR 5·6–8·3; table 2). 13 patients in EXP3B had baseline CNS lesions per ICR, with a median of six CNS lesions per patient (IQR 3–6). Of nine patients with measurable baseline CNS lesions per ICR, intracranial responses were observed in five (55∙6%; 95% CI 21∙2–86∙3; table 2, figure 3C). Median intracranial duration of response was not reached (95% CI 4·1–NR)

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

Articles

Change in tumour size from baseline (%)

A

Best overall response Complete response Partial response Stable disease/no response Objective progession Intermediate * Off treatment or PD occurred

80 70 60 50 40 30 20 10 0 –10 –20 –30 –40 –50 –60 –70 –80 –90 –100

Change in tumour size from baseline (%)

*

*

*

* *

*

*

*

D

C 80 70 60 50 40 30 20 10 0 –10 –20 –30 –40 –50 –60 –70 –80 –90 –100

B

* * **** * * *

*

**

***

*

**

*

****

**

* ** **

**

*

**

Figure 3: Best percentage change in CNS metastatic tumour size from baseline (A) EXP1: treatment-naive patients. (B) EXP2–3A: previous crizotinib with or without 1–2 chemotherapy regimens. (C) EXP3B: previous non-crizotinib ALK TKI with or without chemotherapy. (D) EXP4–5: two or more previous ALK TKIs with or without chemotherapy. Each bar represents an individual patient’s maximum reduction in CNS target lesion size, as assessed by ICR. Patients with at least one on-study target lesion assessment as per ICR were included. If any procedure was different and not interchangeable from the procedure at screening, the percentage change from baseline could not be calculated and is not displayed. Colours indicate patients’ best overall responses per modified Response Evaluation Criteria in Solid Tumors version 1.1 by ICR. Complete response was defined as the disappearance of all target lesions; when nodal disease was included in target lesions, reversion to normal node size (<10 mm) prevented the percent change from baseline from reaching –100%. Some patients with a total change from baseline of –100% are shown as having partial responses because of the presence of non-target lesions in the summary. ALK=anaplastic lymphoma kinase. EXP=expansion cohort. ICR=independent central review. PD=progressive disease. TKI=tyrosine kinase inhibitor.

at the time of analysis. The median time to first tumour response was 1∙4 months (IQR 1·4–2·7), the median time to first intracranial response was 1·4 months (IQR 1∙4–2∙6), and the median progression-free survival was 5∙5 months (95% CI 2∙7–9∙0; figure 4). A total of 111 patients were previously treated with two or three previous ALK tyrosine kinase inhibitors before study enrolment (EXP4–5). Of these patients, 49 (44%) received alectinib as their last ALK tyrosine kinase inhibitor before starting lorlatinib treatment, 34 (31%) received ceritinib, 18 (16%) received crizotinib, seven (6%) received brigatinib, and three (3%) received another ALK tyrosine kinase inhibitor. Of the 111 patients, responses were observed in 43 (38∙7%; 95% CI 29∙6–48∙5), with two complete responses and 41 partial responses (table 2; figure 2D). At the time of analysis, 26 (60%) of the 43 confirmed responses were ongoing, the median duration of response had not been reached (95% CI 5·5–NR), and the median duration of follow-up for response was 7∙2 months (IQR 5·6–9·8). 83 patients in this cohort had baseline CNS lesions per ICR, with a median of four CNS lesions per patient (IQR 2–7). Of the 49 patients with measurable baseline CNS lesions

per ICR, 26 (53∙1%; 95% CI 38∙3–67∙5) had objective intracranial response, with ten complete responses and 16 partial responses (table 2, figure 3D). The median intracranial duration of response was 14∙5 months (95% CI 6∙9–14∙5). The median time to first tumour response was 1∙4 months (IQR 1∙4–2∙9), the median time to first intracranial tumour response was 1∙4 months (1·3–3·1), and the median progression-free survival was 6∙9 months (95% CI 5∙4–9∙5; figure 4). Of the 184 ALK-positive patients who received at least one previous ALK tyrosine kinase inhibitor and were evaluable for PROs, most had either improved (≥10-point improvement from baseline; 72 [39%] patients) or stable (<10-point change from baseline; 78 [42%] patients) scores in global quality of life (QOL) during treatment. Most patients also had improved or stable scores for each of the functioning domains on the EORTC QLQ-C30 (appendix p 8). Improvements from baseline in more than 25% of patients were seen across a number of key lung cancer symptoms, including fatigue (91 [49%] of 184 patients), insomnia (77 [42%]), appetite loss (72 [39%]), pain (69 [38%]), and dyspnoea (54 [29%]) on the QLQ-C30, and cough (76 [41%]), pain in other

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

9

Articles

EXP1 EXP2–3A EXP3B EXP4–5 Pooled EXP2–5

Number of events (%)

Median PFS, months (95% CI)

7 (23) 21 (36) 18 (64) 62 (56) 101 (51)

NR (11·4–NR) NR (12·5–NR) 5·5 (2·7–9·0) 6·9 (5·4–9·5) 7·3 (5·6–11·0)

Progression-free survival (%)

100 80 60 40 20 0

0

Number at risk (number censored) EXP1 30 (0) EXP2–3A 59 (0) EXP3B 28 (0) EXP4–5 111 (0) Pooled EXP2–5 198 (0)

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

6 (17) 5 (33) 1 (9) 7 (43) 13 (85)

6 (17) 4 (34) 1 (9) 5 (44) 10 (87)

6 (17) 3 (35) 0 (10) 5 (44) 8 (89)

2 (21) 1 (37) ·· ·· 2 (47) 3 (94)

0 (23) 0 (38) ·· ·· 0 (49) 0 (97)

Time since treatment initiation (months) 30 30 29 (0) (0) (0) 59 53 50 (0) (0) (1) 27 18 16 (0) (2) (2) 106 88 72 (4) (7) (11) 192 159 138 (4) (9) (14)

8 28 26 23 20 20 12 (1) (1) (3) (5) (5) (13) (16) 9 49 44 31 28 26 14 (1) (2) (10) (12) (14) (25) (30) 2 7 5 8 15 15 11 (2) (2) (2) (3) (4) (6) (8) 64 56 44 37 33 22 17 (12) (15) (21) (24) (24) (33) (36) 128 115 86 73 66 41 28 (15) (19) (33) (39) (42) (64) (74)

7 (17) 9 (30) 2 (8) 16 (37) 27 (75)

6 (17) 9 (30) 1 (9) 13 (38) 23 (77)

Figure 4: Progression-free survival Kaplan-Meier estimates of progression-free survival in the safety analysis set, as assessed by independent central review. Tick marks on the curves indicate censoring of data. EXP=expansion cohort. NR=not reached. PFS=progression-free survival.

parts (56 [30%]), pain in chest (51 [28%]), and dyspnoea (48 [26%]) on the QLQ-LC13. A worsening from baseline was observed in 47 (26%) of 184 patients on the QLQ-C30 for cognitive functioning and in 66 (36%) of 184 patients on the QLQ-LC13 for peripheral nephropathy. Across all treated ALK-positive or ROS1-positive patients (EXP1–6), 157 (57%) of 275 patients in the safety analysis remained on treatment with a median duration of treatment of 8∙3 months (IQR 5·0–10·7). Overall, the median relative dose intensity was 98∙5% (IQR 86·7–100). The safety profile of lorlatinib observed in phase 2 was similar to that reported in the phase 1 portion of the study. The most common treatment-related adverse events of any grade were hypercholesterolaemia (224 [81%] of 275 patients), hypertriglyceridaemia (166 [60%] patients), oedema (119 [43%] patients), and peripheral neuropathy (82 [30%] patients). Most treatmentrelated adverse events were grade 1 or 2 in severity, with few grade 3 or worse treatment-related adverse events reported (table 3). The most common grade 3–4 treatmentrelated adverse events were hypercholesterolaemia and hypertriglyceridaemia (each in 43 [16%] of 275 patients). Serious treatment-related adverse events occurred in 19 (7%) of 275 patients, the most frequent of which was cognitive effects (two [1%] patients; appendix p 9). In the 262 patients evaluated for weight gain, 80 (31%) had a 10

10–20% increase from their baseline weight and 33 (13%) had an increase of 20% or greater (appendix p 10). Comprehensive guidance on the management of the unique adverse event profile of lorlatinib will be the subject of a separate publication in the near future. Dose interruptions and dose reductions associated with treatment-related adverse events were reported in 83 (30%) and 61 (22%) of 275 patients, respectively. The most common treatment-related cause for dose interruptions and dose modifications was oedema (in 16 [6%] and 18 [7%] of 275 patients, respectively). The incidence of permanent discontinuations due to treatment-related adverse events was low (seven [3%] of 275 patients), and the most common treatment-related adverse event associated with permanent discontinuation was cognitive effects (one patient with confusional state and one with cognitive disorder; appendix p 11). No treatment-related deaths were reported; a listing of all causes of death is available in the appendix (p 12). CNS effects of any cause were reported in 107 (39%) of 275 patients, which included changes in cognitive function (62 [23%] patients), mood (60 [22%]), and speech (23 [8%]), and were generally mild (grade 1 or 2) in severity, transient, intermittent, and reversible after dose modifications (data not shown). Additional safety assessments of cognitive function, mood, and suicidal ideation and behaviour are provided in the appendix (p 2).

Discussion In this phase 1–2 clinical trial, lorlatinib—a novel macrocyclic ALK tyrosine kinase inhibitor with broad coverage of ALK kinase domain mutations and good CNS penetration—demonstrated clinical activity across cohorts of patients with ALK-positive non-small-cell lung cancer who were either treatment naive or had received previous treatment with ALK tyrosine kinase inhibitors. Activity was evident in treatment-naive patients, patients who had received crizotinib as their only previous ALK tyrosine kinase inhibitor, patients who had received a previous second-generation ALK tyrosine kinase inhibitor, and in patients who had received multiple lines of ALK tyrosine kinase inhibitor therapy. Treatment with lorlatinib was associated with intracranial responses in 87% of patients treated with previous crizotinib only and more than half of patients treated with previous second-generation ALK tyrosine kinase inhibitors. The safety profile of lorlatinib was consistent with the preceding phase 1 study, with hypercholesterolaemia and hyper­ triglyceridaemia being the most frequently reported adverse events. In patients with treatment-naive, ALK-positive nonsmall-cell lung cancer, nearly all patients (90%) achieved a confirmed response with lorlatinib treatment, and most patients had ongoing responses at the time of analysis. The proportion of patients with objective response in this small cohort numerically exceeded those reported with second-generation ALK tyrosine kinase inhibitors (alectinib and ceritinib) in the first-line setting.5,6 These

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

Articles

Grade 1–2

Grade 3

Grade 4

Hypercholesterolaemia*

181 (66%)

39 (14%)

4 (1%)

Hypertriglyceridaemia*

123 (45%)

36 (13%)

Oedema*

113 (41%)

6 (2%)

Grade 1–2

Grade 3

Grade 4

(Continued from previous column)

7 (3%)

Pulmonary oedema

2 (1%)

1 (<1%)

0

Acute respiratory failure

0

1 (<1%)

0 1 (<1%)

Hyponatraemia

1 (<1%)

1 (<1%)

0

Peripheral neuropathy*

77 (28%)

5 (2%)

0

Weight increased

45 (16%)

5 (2%)

0

Presyncope

1 (<1%)

1 (<1%)

0

Cognitive effects*

46 (17%)

3 (1%)

0

Respiratory failure

0

2 (1%)

0 0

Mood effects*

39 (14%)

2 (1%)

0

Ascites

0

1 (<1%)

Fatigue*

35 (13%)

1 (<1%)

0

Blood potassium increased

0

0

1 (<1%)

Diarrhoea

28 (10%)

1 (<1%)

0

Diabetes mellitus

0

1 (<1%)

0

Arthralgia

28 (10%)

0

0

Erysipelas

0

1 (<1%)

0

AST increased

27 (10%)

1 (<1%)

0

Gastritis

0

1 (<1%)

0

Dizziness

23 (8%)

2 (1%)

0

Glossitis

0

1 (<1%)

0

ALT increased

22 (8%)

2 (1%)

0

Hydrocephalus

0

1 (<1%)

0

Speech effects*

19 (7%)

1 (<1%)

0

Hypermagnesaemia

0

1 (<1%)

0

Interstitial lung disease

0

1 (<1%)

0

Lipase increased

10 (4%)

7 (3%)

1 (<1%)

Anaemia

13 (5%)

2 (1%)

0

Leukocytosis

0

1 (<1%)

0

Amylase increased

12 (4%)

2 (1%)

0

Mental status changes

0

1 (<1%)

0

Rash

13 (5%)

1 (<1%)

0

Mucocutaneous candidiasis

0

1 (<1%)

0

Vomiting

11 (4%)

1 (<1%)

0

Pancreatitis

0

1 (<1%)

0

Dyspnoea

8 (3%)

1 (<1%)

0

Pneumonia

0

1 (<1%)

0

0

0

1 (<1%)

0

1 (<1%)

0

Hypertension

4 (1%)

4 (1%)

0

Pneumonitis

Ejection fraction decreased

5 (2%)

1 (<1%)

0

Thrombosis

Hyperglycaemia

4 (1%)

2 (1%)

0

Localised oedema

4 (1%)

2 (1%)

0

Hallucination, auditory

4 (1%)

1 (<1%)

0

Abdominal pain

3 (1%)

1 (<1%)

0

Hypophosphataemia

2 (1%)

2 (1%)

0

Hypoxia

1 (<1%)

2 (1%)

0

Night sweats

2 (1%)

1 (<1%)

0

Data are n (%). This table lists treatment-related adverse events reported in at least 10% of patients and all grade 3–4 treatment-related adverse events. No grade 5 treatment-related adverse events were reported. ALT=alanine aminotransferase. AST=aspartate aminotransferase. EXP=expansion cohort. *Cluster term comprising adverse events that represent similar clinical symptoms or syndromes.

Table 3: Treatment-related adverse events in patients treated with lorlatinib (all cohorts: EXP1–6)

(Table 3 continues in next column)

results indicate that lorlatinib might provide an effective treatment option for patients with newly diagnosed ALKpositive non-small-cell lung cancer and provide rationale for the evaluation of lorlatinib in the first-line treatment of this condition, which is currently underway (the CROWN trial; NCT03052608). In patients previously treated with at least one ALK tyrosine kinase inhibitor (pooled EXP2–5), the proportion with objective response of 47% is notable given that most of this population were heavily pretreated patients whose disease had progressed on up to three previous ALK tyrosine kinase inhibitors, including second-generation ALK tyrosine kinase inhibitors such as alectinib, ceritinib, and brigatinib. Responses were rapid (median time to first tumour response of just 1∙4 months), deep (appendix p 13), and durable. Although a median duration of response had not yet been reached because of the large number of patients with ongoing tumour response and the median duration of follow-up for response of just 6∙9 months at data cutoff, the lower boundary of the duration of response 95% CI was 11∙1 months for the pooled cohort of patients previously treated with at least one ALK tyrosine kinase inhibitor.

The clinical benefit of lorlatinib treatment was also observed in patients pretreated with crizotinib with or without one to two previous chemotherapy regimens (EXP2–3A), among whom almost 70% achieved an objective response. Again, the time to first tumour response was rapid and, although the median duration of response was not reached in this pooled cohort (78∙1% of patients were censored at the time of data cutoff), the lower boundary of the 95% CI was 11∙1 months. Notwithstanding the limitations of cross-trial compari­ sons, this proportion of patients with objective response numerically exceeded previously reported proportions of patients with objective response with other ALK tyrosine kinase inhibitors in this treatment setting (ie, patients treated with only previous crizotinib or with previous crizotinib and chemotherapy), such as ceritinib, alectinib, and brigatinib, for which proportions of patients with objective responses of between 48% and 53% have been reported.8,15–17 Despite the availability of three licensed ALK tyrosine kinase inhibitors in the post-crizotinib setting, disease progression in patients who have been previously treated with two or more ALK tyrosine kinase inhibitors repre­ sents a treatment challenge. Standard therapies including

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

11

Articles

single-agent chemotherapy18,19 or immuno­therapy20,21 have poor efficacy in the second-line setting and beyond. Although platinum–pemetrexed chemotherapy has been reported to have activity in ALK-positive non-small-cell lung cancer,4,6 clinical benefit has not been established after failure of a second-generation ALK tyrosine kinase inhibitor. The proportion of patients who achieved an objective response with lorlatinib treatment after a secondgeneration ALK tyrosine kinase inhibitor (either in EXP3B or EXP4–5) contrasts with the poor responses seen after single-agent chemotherapy. Similar to the pooled group of patients treated with at least one ALK tyrosine kinase inhibitor (EXP2–5), time to first tumour response was rapid (median 1∙4 months for both EXP3B and EXP 4–5), corresponding to the approximate time of the first on-treatment scan. After treatment with a second-generation ALK tyrosine kinase inhibitor when used as the only ALK tyrosine kinase inhibitor with or without previous chemotherapy (EXP3B), lorlatinib led to 32·1% of patients achieving an objective response. In patients previously treated with two or three ALK tyrosine kinase inhibitors with or without previous chemotherapy (pooled EXP4–5), 38∙7% achieved an objective response with lorlatinib treatment. Furthermore, the lower boundaries of the 95% CIs (15∙9% and 29∙6% in EXP3B and EXP4–5, respectively) around the observed pro­ portions of patients with objective response exceeded the proportions of patients with objective response reported for single-agent chemotherapy in the ALUR18,22 (2·9–11·4%) and ASCEND-519 trials (6·9%). The median progression-free survival times in EXP3B (5∙5 months; 95% CI 2∙7–9∙0) and EXP4–5 (6∙9 months; 5∙4–9∙5) also numerically exceeded the median progression-free survival for chemo­therapy in these studies (1·6 months for both trials). However, since no published clinical data are available for chemotherapy in the exact clinical settings of cohorts EXP3B and EXP4–5, direct com­ parisons with single-agent or pemetrexed-based doublet chemotherapy are difficult to make. In the pooled cohort of patients previously treated with two or three ALK tyrosine kinase inhibitors with or without previous chemotherapy (EXP4–5), the median duration of response was not reached at the time of data cutoff because 60% of the patients were censored, mainly because they were still having ongoing tumour responses. However, the lower boundary of the 95% CI was 5∙5 months. Taken together, these data indicate that lorlatinib has the potential to fill an important unmet medical need among these patients whose disease has progressed on a second-generation ALK tyrosine kinase inhibitor. Brain metastases are frequent in patients with ALKpositive non-small-cell lung cancer, and are associated with a poor prognosis, including an increase in physical and neurocognitive morbidities as well as a decrease in progression-free and overall survival.23–25 In this study, 12

brain metastases were present in about a quarter of treatment-naive patients and two-thirds of previously treated patients. Lorlatinib exhibited substantial intra­ cranial activity, consistent with its ability to cross the blood–brain barrier demonstrated in preclinical models. The proportion of patients with objective intracranial response in the pooled EXP2–5 cohort was 63∙0%, with clinical benefit recorded across all cohorts. This level of intracranial antitumour activity underscores the clinical benefit in this group of patients in whom brain metastases represent a major source of morbidity. The proportion of patients with objective intracranial response to lorlatinib among patients treated with only previous crizotinib or previous crizotinib and chemotherapy (EXP2–3A) was 87∙0% in patients with at least one measurable brain lesion at baseline. Similar to the previous observation made for the overall proportion of patients with objective response and with the caveats of cross-trial comparisons, this proportion of patients with objective intracranial response numerically exceeded previously reported pro­ portions of patients with objective intracranial response of other ALK tyrosine kinase inhibitors in this treatment setting, such as ceritinib, alectinib, and brigatinib, for which proportions of patients with objective intracranial response between 36% and 67% have been reported.16,17,26 Additionally, intracranial responses were recorded among patients in whom a previous brain-penetrable ALK tyrosine kinase inhibitor, such as alectinib, had failed (43∙2% of patients with CNS metastases who received alectinib as last prior second-generation TKI had an objective intracranial response). Lorlatinib treatment showed improvement from baseline in global QOL as indicated by analyses of PROs. Although there was worsening in peripheral neuropathy and cognitive functioning, there were improvements in physical, emotional, social, and role functioning. Improvements were also shown in appetite loss and key lung cancer symptoms such as pain, dyspnoea, cough, and fatigue. Adverse events reported with lorlatinib were generally mild to moderate in severity and manageable through dosing modifications and supportive medical therapy. There were few permanent treatment discontinuations and no treatment-related deaths. Hypercholesterolaemia and hypertriglyceridaemia were the most common adverse events associated with lorlatinib and were manageable with lipid-lowering medical therapy and dose modifications. Early institution of therapy with a statin (eg, pitavastatin, pravastatin, or rosuvastatin, which are not metabolised by cytochrome P450 3A4) with or without a fibrate was effective for most patients. Cognitive effects were generally mild, transient, and rapidly reversible upon dose modifications. Potential limitations of this large, phase 2, nonrandomised study, which included patients treated with ALK tyrosine kinase inhibitors, are that central confir­ mation of ALK positivity was not required and that

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

Articles

pre-treatment biopsies to determine molecular correlates of response to lorlatinib were collected in only half of the patients. To overcome this difficulty, plasma circulating free DNA was collected, allowing analysis of ALK mutation status and outcome, the results of which will be published in a separate report. However, even this molecular analysis is limited by its retrospective, exploratory nature. A further potential limitation is that progressionfree survival was not mature for several of the cohorts given that many patients were still on treatment with ongoing responses. Overall, the results of this trial indicate that lorlatinib has promising antitumour activity in treatment-naive patients with ALK-positive advanced non-small-cell lung cancer and provides an important new treatment option for those patients whose disease has progressed after treatment with crizotinib or second-generation ALK tyrosine kinase inhibitors. Lorlatinib also shows significant intracranial activity in patients with ALK-positive non-small-cell lung cancer who have a high incidence of brain metastases, particularly those who have received previous ALK tyrosine kinase inhibitor therapy. On the basis of the robust anti­ tumour activity demonstrated by lorlatinib, lorlatinib represents a treatment option for patients who have progressed on one or more ALK tyrosine kinase inhibitors. A phase 3 study (NCT03052608) evaluating lorlatinib in comparison with crizotinib as first-line therapy for treatment-naive patients with ALK-positive advanced nonsmall-cell lung cancer is underway. Contributors Data were collected by the investigators and analysed by the sponsor. All authors, including those employed by the funder of the study, contributed to interpretation of the data, and all authors contributed to the development and approval of the manuscript. All authors had full access to all the data in the study and the corresponding author had final responsibility for the decision to submit the publication. Declaration of interests BJS reports personal fees from AstraZeneca, Roche, Merck, Bristol-Myers Squibb, Novartis, Veristrat (Biodesix), Merck Sharp & Dohme, Pfizer, Roche/Genentech, and BeiGene; grants from Pfizer; and non-financial support from Roche, Novartis, and Pfizer. BB reports grants from Roche, Pfizer, Bristol-Myers Squibb/Medarex, Novartis, Pierre Fabre, AstraZeneca, Roche/Genentech, Boehringer Ingelheim, Lilly, SERVIER, Onxeo, Bristol-Myers Squibb, OSE Pharma, and Inivata. TMB reports personal fees from Tennessee Oncology, Sarah Cannon Research Institute, Ignyta, Guardant Health, Loxo, Pfizer, Daiichi Sankyo, and Medpacto; and grants from Ignyta, Pfizer, Daiichi Sankyo, Medpacto, Incyte, Mirati Therapeutics, MedImmune, Abbvie, AstraZeneca, Leap Therapeutics, MabVax, Stemline Therapeutics, Merck, Lilly, GlaxoSmithKline, Novartis, Principa Biopharma, Roche/Genentech, Deciphera, Merrimack, Immunogen, Millennium, Calithera Biosciences, Kolltan Pharmaceuticals, Principa Biopharma, Peleton, Immunocore, Roche, Aileron Therapeutics, Bristol-Myers Squibb, Amgen, Moderna Therapeutics, Sanofi, Boehringer Ingelheim, and Astellas Pharma. EF reports personal fees from AstraZeneca, Boehringer Ingelheim, BMS, Celgene, Eli Lilly, GuardantHealth, MSD, Novartis, Pfizer, Roche, Takeda, and MERCK. RAS reports grants from AstraZeneca; and personal fees from AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Celgene, Merck Sharp & Dohme, Novartis, Pfizer, Roche/Genentech, Taiho Pharmaceutical, and Lilly. DRC reports grants from Takeda; and personal fees from Roche, G1 Therapeutics, Mersana, Takeda, AstraZeneca, Genoptix, Ignyta, Daiichi Sankyo, Hansoh, Lycera, Bio-thera, and Revolution Med. AB reports personal fees from Takeda, Eli-Lilly,

Pfizer, MSD, BMJ, Novartis, and Roche. SMG reports grants from Pfizer, Roche/Genentech, Bristol-Myers Squibb, Clovis Oncology, Merck, Incyte, Millennium, AstraZeneca/MedImmune, Halozyme, Acerta Pharma, ACEA Biosciences, Janssen Oncology, Novartis, Five Prime Therapeutics, and OncoMed; personal fees from Pfizer, Roche/Genentech, Ariad, AstraZeneca, Bristol-Myers Squibb; and non-financial support from Roche/Genentech and ARIAD/Takeda. GJR reports personal fees from Genentech; grants from Novartis, Roche/Genentech, Millennium, GlaxoSmithKline, Pfizer, Infinity Pharmaceuticals, and ARIAD; and nonfinancial support from Merck Sharp & Dohme. EHT reports grants from Bristol-Myers Squibb; personal fees from MSD and Boehringer Ingelheim; and non-financial support from MSD and Boehringer Ingelheim. TS reports grants from Astellas Pharma, AstraZeneca, Chugai Pharmaceutical, Daiichi Sankyo, Eli Lilly Japan, Kissei Pharmaceutical, MSD, Nippon Boehringer Ingelheim, Pfizer Japan, YakultHonsha, Bayer Yakuhin, Eisai, Merck Serono, Novartis Pharma, and Verastem; and personal fees from Astellas Pharma, AstraZeneca, Chugai Pharmaceutical, Daiichi Sankyo, Eli Lilly Japan, Kissei Pharmaceutical, MSD, Nippon Boehringer Ingelheim, Pfizer Japan, YakultHonsha, Bristol-Myers Squibb, Kyowa Hakko Kirin, Mochida Pharmaceutical, Nippon Kayaku, Ono Pharmaceutical, Roche Singapore, Sanofi, Showa Yakuhin Kako, Showa Yakuhin Kako, and Takeda Pharmaceutical. LPJ reports personal fees from Pfizer; and stocks from Pfizer. JSC reports personal fees from Pfizer. AA reports personal fees from Pfizer; and stocks from Pfizer. J-FM reports personal fees from Pfizer; and stocks from Pfizer. JC reports personal fees from Pfizer; and salary from Pfizer. GP reports personal fees from Pfizer; and stocks from Pfizer. HT reports personal fees from Pfizer; and stocks from Pfizer. S-HIO reports grants from Pfizer, AstraZeneca, Roche Pharma AG, AstraZeneca/MedImmune, Clovis Oncology, ARIAD, Ignyta, Peregrine Pharmaceuticals, GlaxoSmithKline, Astellas Pharma, and Chugai Pharma; and personal fees from Pfizer, Roche/Genentech, Novartis, AstraZeneca, Takeda, Foundation Medicine, Genentech, Roche Pharma AG, and ARIAD/Takeda. ATS reports grants from Pfizer; and personal fees from Pfizer, Novartis, Genentech, Roche, Ariad, Ignyta, Blueprint Medicines, Daiichi Sankyo, EMD Serono, Taiho Pharmaceutical, KSQ Therapeutics, Natera, Foundation Medicine, Takeda, and Loxo. RC and C-CL declare no competing interests. Data sharing Upon request, and subject to certain criteria, conditions and exceptions, Pfizer will provide access to individual de-identified participant data from Pfizer-sponsored global interventional clinical studies conducted for medicines, vaccines and medical devices for indications that have been approved in the USA or EU, or in programmes that have been terminated (ie, development for all indications has been discontinued). Pfizer will also consider requests for the protocol, data dictionary, and statistical analysis plan. Data may be requested from Pfizer trials 24 months after study completion. The de-identified participant data will be made available to researchers whose proposals meet the research criteria and other conditions, and for which an exception does not apply, via a secure portal. To gain access, data requestors must enter into a data access agreement with Pfizer.

For more on Pfizer’s data sharing policy see https://www.pfizer.com/science/ clinical-trials/trial-data-andresults

Acknowledgments This study was supported by Pfizer. We thank the participating patients and their families, the investigators, sub-investigators, research nurses, study coordinators, and operations staff. Medical writing support was provided by Jade Drummond and Brian Szente of inScience Communications, Springer Healthcare (Chester, UK, and Philadelphia, PA, USA), and was funded by Pfizer. References 1 Koivunen JP, Mermel C, Zejnullahu K, et al. EML4-ALK fusion gene and efficacy of an ALK kinase inhibitor in lung cancer. Clin Cancer Res 2008; 14: 4275–83. 2 Rikova K, Guo A, Zeng Q, et al. Global survey of phosphotyrosine signaling identifies oncogenic kinases in lung cancer. Cell 2007; 131: 1190–203. 3 Takeuchi K, Soda M, Togashi Y, et al. RET, ROS1 and ALK fusions in lung cancer. Nat Med 2012; 18: 378–81. 4 Solomon BJ, Mok T, Kim DW, et al. First-line crizotinib versus chemotherapy in ALK-positive lung cancer. N Engl J Med 2014; 371: 2167–77.

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1

13

Articles

5 6

7 8 9 10 11 12

13

14

15

14

Peters S, Camidge DR, Shaw AT, et al. Alectinib versus crizotinib in untreated ALK-positive non-small-cell lung cancer. N Engl J Med 2017; 377: 829–38. Soria JC, Tan DSW, Chiari R, et al. First-line ceritinib versus platinum-based chemotherapy in advanced ALK-rearranged non-small-cell lung cancer (ASCEND-4): a randomised, open-label, phase 3 study. Lancet 2017; 389: 917–29. Lin JJ, Riely GJ, Shaw AT. Targeting ALK: precision medicine takes on drug resistance. Cancer Discov 2017; 7: 137–55. Shaw AT, Gandhi L, Gadgeel S, et al. Alectinib in ALK-positive, crizotinib-resistant, non-small-cell lung cancer: a single-group, multicentre, phase 2 trial. Lancet Oncol 2016; 17: 234–42. Shaw AT, Kim DW, Mehra R, et al. Ceritinib in ALK-rearranged non-small-cell lung cancer. N Engl J Med 2014; 370: 1189–97. Costa DB, Shaw AT, Ou SH, et al. Clinical experience with crizotinib in patients with advanced ALK-rearranged non-small-cell lung cancer and brain metastases. J Clin Oncol 2015; 33: 1881–88. Gainor JF, Dardaei L, Yoda S, et al. Molecular mechanisms of resistance to first- and second-generation ALK inhibitors in ALK-rearranged lung cancer. Cancer Discov 2016; 6: 1118–33. Johnson TW, Richardson PF, Bailey S, et al. Discovery of (10R)-7-amino-12-fluoro-2,10,16-trimethyl-15-oxo-10,15,16,17tetrahydro-2H-8,4-(m etheno)pyrazolo[4,3-h] [2,5,11]-benzoxadiazacyclotetradecine-3-carbonitrile (PF-06463922), a macrocyclic inhibitor of anaplastic lymphoma kinase (ALK) and c-ros oncogene 1 (ROS1) with preclinical brain exposure and broad-spectrum potency against ALK-resistant mutations. J Med Chem 2014; 57: 4720–44. Shaw AT, Felip E, Bauer TM, et al. Lorlatinib in non-small-cell lung cancer with ALK or ROS1 rearrangement: an international, multicentre, open-label, single-arm first-in-man phase 1 trial. Lancet Oncol 2017; 18: 1590–99. Long GV, Trefzer U, Davies MA, et al. Dabrafenib in patients with Val600Glu or Val600Lys BRAF-mutant melanoma metastatic to the brain (BREAK-MB): a multicentre, open-label, phase 2 trial. Lancet Oncol 2012; 13: 1087–95. Crino L, Ahn MJ, De Marinis F, et al. Multicenter phase II study of whole-body and intracranial activity with ceritinib in patients with ALK-rearranged non-small-cell lung cancer previously treated with chemotherapy and crizotinib: results from ASCEND-2. J Clin Oncol 2016; 34: 2866–73.

16 Kim DW, Tiseo M, Ahn MJ, et al. Brigatinib in patients with crizotinib-refractory anaplastic lymphoma kinase-positive non-small-cell lung cancer: a randomized, multicenter phase II trial. J Clin Oncol 2017; 35: 2490–98. 17 Ou SH, Ahn JS, De Petris L, et al. Alectinib in crizotinib-refractory ALK-rearranged non-small-cell lung cancer: a phase II global study. J Clin Oncol 2016; 34: 661–68. 18 Novello S, Mazières J, Oh IJ, et al. Alectinib versus chemotherapy in crizotinib-pretreated anaplastic lymphoma kinase (ALK)-positive non-small-cell lung cancer: results from the phase III ALUR study. Ann Oncol 2018; 29: 1409–16. 19 Shaw AT, Kim TM, Crino L, et al. Ceritinib versus chemotherapy in patients with ALK-rearranged non-small-cell lung cancer previously given chemotherapy and crizotinib (ASCEND-5): a randomised, controlled, open-label, phase 3 trial. Lancet Oncol 2017; 18: 874–86. 20 Bylicki O, Paleiron N, Margery J, et al. Targeting the PD-1/PD-L1 immune checkpoint in EGFR-mutated or ALK-translocated non-small-cell lung cancer. Target Oncol 2017; 12: 563–69. 21 Gainor JF, Shaw AT, Sequist LV, et al. EGFR mutations and ALK rearrangements are associated with low response rates to PD-1 pathway blockade in non-small cell lung cancer: a retrospective analysis. Clin Cancer Res 2016; 22: 4585–93. 22 Novello S, Mazieres J, Oh I, et al. Primary results from the phase III ALUR study of alectinib versus chemotherapy in previously treated ALK+ non-small-cell lung cancer (NSCLC). ESMO 2017 Congress; Madrid, Spain; Sept 11, 2017. Abstr 1299O_PR. 23 Berger LA, Riesenberg H, Bokemeyer C, Atanackovic D. CNS metastases in non-small-cell lung cancer: current role of EGFR-TKI therapy and future perspectives. Lung Cancer 2013; 80: 242–48. 24 Solomon BJ, Cappuzzo F, Felip E, et al. Intracranial efficacy of crizotinib versus chemotherapy in patients with advanced ALK-positive non-small-cell lung cancer: results from PROFILE 1014. J Clin Oncol 2016; 34: 2858–65. 25 Yoshida T, Oya Y, Tanaka K, et al. Clinical impact of crizotinib on central nervous system progression in ALK-positive non-small lung cancer. Lung Cancer 2016; 97: 43–47. 26 Kim DW, Mehra R, Tan DS, et al. Activity and safety of ceritinib in patients with ALK-rearranged non-small-cell lung cancer (ASCEND-1): updated results from the multicentre, open-label, phase 1 trial. Lancet Oncol 2016; 17: 452–63.

www.thelancet.com/oncology Published online November 6, 2018 http://dx.doi.org/10.1016/S1470-2045(18)30649-1