Background

Non-small cell lung cancer (NSCLC) is the leading cause of cancer-related mortality worldwide with most patients diagnosed at an advanced, incurable stage [1–3]. Over the past years, advanced molecular diagnostics have revealed distinct driver molecular alterations in about 15–20% of these tumors, which also constitute particular therapeutic susceptibilities. Targeted treatment with tyrosine kinase inhibitors (TKI) has consistently demonstrated superior efficacy and tolerability compared to conventional chemotherapy in eligible patients and is meanwhile the standard first-line option.

ALK (anaplastic lymphoma kinase)-positive tumors comprise approximately 5% of metastatic NSCLC, mainly adenocarcinomas of younger, non-smoker patients [4, 5], and are unique in many aspects: They typically have a lower number of genetic alterations compared to other NSCLC subtypes, as partly reflected by a mean tumor mutational burden (TMB) uniquely below 3 mut/Mb [6, 7]. ALK+ NSCLC patients require the most sophisticated management at present, i.e., high-level expertise and close cooperation between medical oncology, interventional pneumology, radiology, thoracic surgery, radiation oncology, as well as molecular pathology over several years, and enjoy the best outcome. Even the first-generation compound crizotinib results in longer overall survival (OS) for ALK+ NSCLC patients compared to their TKI-treated EGFR+ counterparts (for example > 46 months in the PROFILE 1014 study vs. 39 and 32 months in the two arms of the FLAURA trial), while sequential ALK TKI, particularly with upfront administration of second-generation inhibitors such as alectinib, confer a median OS over five years, which is certainly one of the greatest successes in modern thoracic oncology [6, 8–10]. First-line administration of the newer second-generation inhibitor brigatinib has produced comparable results to alectinib in the randomized phase III ALTA-1 L trial [11, 12], and both substances are currently initial treatments of choice [13, 14]. An important advantage of both compounds is enhanced intracranial efficacy, with brain overall response rates (iORR) of approximately 80% and rates of intracranial progression < 10% per year, compared to 30–50% and approximately 20% with the first-generation inhibitor crizotinib, respectively. Very recently, interim results from the CROWN trial, comparing lorlatinib against crizotinib in the first line, have been reported [15]. With a median follow-up of 18 months, data are still immature, but the hazard ratio (HR) of 0.28 for progression-free survival (PFS) in favor of the third-generation TKI lorlatinib indicates that this compound will probably also play a role as upfront therapy in the future. At the same time, the generally favorable outcome of ALK+ NSCLC facilitates identification of special patient subsets with earlier treatment failure and higher risk of death. Currently, high-risk ALK+ disease comprises three main groups: Tumors with the EML4-ALK fusion variant 3 (V3), whose shorter structure lowers TKI sensitivity and increases metastatic potential [16–20], tumors with TP53 mutations, either at initial diagnosis [21, 22] or at the time of TKI failure [23], and “double-positive patients”, who have the worst outcome [22], as also recently demonstrated in the context of the randomized phase 3 ALTA-1 L trial [24]. Importantly, repeat molecular profiling based on tissue or liquid rebiopsies at the time of disease progression can identify actionable resistance mutations and guide individualized selection of next-line targeted therapies with higher clinical benefit for the patient [25, 26, 27].

The aim of this trial is to optimize management of ALK+ NSCLC by analyzing the efficacy of brigatinib and other ALK TKI in conjunction with deep clinical and molecular patient phenoty** across two treatment lines.

Methods/design

Study design

This is a prospective, randomized, open-label, multicenter phase II trial comparing brigatinib vs. other second-generation ALK inhibitors as first-line therapy for ALK+ NSCLC in conjunction with deep patient profiling across two treatment lines.

Study setting

The ABP trial will recruit patients from currently 24 participating centers across Germany over a period of 36 months. Recruitment started in June 2020. A full list of sites is available at clinicaltrials.gov (NCT04318938).

Study objectives and endpoints

The primary objective of this trial is a descriptive comparison of the efficacy of brigatinib vs. other approved second-generation ALK TKI in first-line treatment of locally advanced or metastatic ALK+ NSCLC, as assessed by progression-free survival (PFS1) according to RECIST 1.1 [28]. Secondary objectives are treatment efficacy in the first and second line, assessed by the PFS of second-line therapy (PFS2), time-to-next treatment in the first line (TNT1), second line (TNT2) and until start of the third line (TNT1/2), as well as by the overall survival (OS). Moreover, CNS efficacy of first- and second-line treatment will be evaluated according to RECIST 1.1 regarding iORR, intracranial duration of response (iDOR) and time-to-intracranial progression (iTTP).

Safety in both treatment arms will be evaluated with regard to incidence and severity of adverse events (AE), while quality of life (QoL) will be assessed using the validated questionnaires SF-12 and EORTC-QLQ-BN20 (the latter in case of brain metastases, only).

Exploratory objectives are ty** of ALK fusion variants, assessment of TP53 status, profiling of resistance mechanisms at the failure of first-line treatment, and their relationship with efficacy endpoints. Furthermore, the potential clinical utility of cerebrospinal fluid ctDNA analysis will be assessed for cases with isolated intracranial progression.

Characteristics of participants

A total of 116 adult patients with newly diagnosed ALK+ NSCLC, either locally advanced (stage III) and not suitable for curative treatment, or metastatic (stage IV), will be enrolled into this study after written informed consent. The full list of inclusion and exclusion criteria is given in Table 1. ALK rearrangement will be determined locally by one or more ALK assays that are approved in Germany and/or have been validated within the German Network for Genomic Medicine, e.g. RNA-based NGS, fluorescence in situ hybridization using the ZytoLight SPEC ALK probe (ZytoVision GmbH, Bremerhaven, Germany) and immunohistochemistry using the D5F3 clone (Roche, Mannheim, Germany) [18, 29–31].

Table 1 Inclusion and exclusion criteria
Full size table

In addition, participants must have at least one measurable site of disease as defined by RECIST 1.1, sufficient ECOG performance status (≤ 2), as well as adequate bone marrow, hepatic and renal function. Participants may not have received prior systemic therapy for ALK+ NSCLC, however up to two cycles of chemotherapy as well as stereotactic radiotherapy are allowed to accommodate for the fact that in some cases treatment might be initiated in an urgent manner before availability of molecular results. Further exclusion criteria include pulmonary interstitial disease, pneumonitis, uncontrolled hypertension and cardiovascular disease, significant infections or any other medical condition that may potentially compromise patient safety or hamper study completion.

Study procedures

Patients will be randomized into the experimental (brigatinib) or standard arm in a 1:1 ratio. The variance minimization method will be used for patient allocation, adjusting for the confounders brain metastases (presence vs. absence of brain metastases at baseline) and ECOG performance status (0–1 vs. 2) in order to ensure better comparability between intervention groups. Overall, 58 patients will be enrolled per treatment arm.

Treatment

Patients in standard arm A are treated in the first line with any approved second-generation TKI other than brigatinib (currently alectinib or ceritinib) according to the investigator’s choice, followed by second-line treatment with another approved ALK TKI of the investigator’s choice. Patients in the experimental arm B receive brigatinib in the first line, followed by any other approved ALK TKI in the second line according to the investigator’s choice (Fig. 1). It is recommended that the choice of second-line TKI takes into account the molecular tumor profile (e.g. presence of ALK resistance mutations) at the time of disease progression, which will be analyzed in tissue and liquid rebiopsies as part of the study and made available to the investigators using a “fast-track” procedure in order to support therapeutic decisions. Arm A patients failing standard second-generation TKIs will also have the option to receive brigatinib as second-line treatment, if so chosen by the treating physician, which will be provided by Takeda.

Fig. 1
figure 1

Study design. Eligible patients are randomized into the two treatment arms stratified by presence of brain metastases and ECOG performance status. First-line treatment in arm B is brigatinib, in arm A it is any other second-generation TKI of the investigator’s choice. All subsequent TKI will be chosen by the investigator and should take into account molecular tumor profiles upon progression. After the end of study treatment, patients will be followed-up for toxicity and survival. R: randomization, PD: progressive disease, TKI: tyrosine kinase inhibitor, FU: follow-up

Full size image

All TKI will be taken orally at the recommended daily doses, with predefined dose modifications in case of tolerability problems. Safety will be monitored by physical examinations, ECOG performance status, clinical laboratory profiles and continuous assessments of AE and QoL will further be evaluated using patient questionnaires (SF-12 and, in case of brain metastases, EORTC QLQ-BN20). Patients will be allowed to continue study treatment according to the decision of treating physicians for as long as they derive clinical benefit, i.e. radiologic progression per se does not necessitate a treatment switch.

Radiologic assessments

Baseline tumor evaluation will take place at screening by contrast-enhanced CT of the chest and abdomen. Tumor response will be evaluated according to RECIST 1.1 after two cycles (8 weeks) and every 12 weeks thereafter (Q12W ±7 days) in both the first and second line. Moreover, contrast-enhanced CT/MRI of the brain will be performed at baseline testing (MRI is strongly recommended by the study protocol). If intracranial lesions are detected, brain imaging is recommended with every scheduled tumor response assessment thereafter (intervals may be adapted by the investigators depending on the location and size of detected lesions), otherwise brain surveillance will take place with every second radiologic assessment, i.e. 20 weeks after beginning of first or second line, and every 24 weeks thereafter. Baseline disease assessment for the second therapy line must be performed within 30 days prior to treatment start. Assessments will continue until progression, death or initiation of another antineoplastic therapy after the two study treatment lines according to the standard of care (SOC).

Sampling for molecular profiling

Comprehensive clinical phenoty** and molecular tumor characterization are pivotal to this trial. Tumor biopsies are collected prior to the start of first-line treatment as formalin-fixed paraffin-embedded (FFPE) tissue. Moreover, blood samples will be taken at baseline and with every radiologic assessment during the first and second therapy line.

At the time of progression, “fast-track” NGS-based molecular profiling of liquid (ctDNA) and tissue rebiopsies (preferentially on new or growing lesions) will be offered to guide the choice of next-line therapy. In case of isolated intracranial progression, ctDNA-based NGS profiling using cerebrospinal fluid samples will be offered if the patient gives written consent.

An overview of all study procedures is presented in Table 2.

Table 2 Schedule of Study Assessments and Procedures
Full size table

Study management and data quality assurance will be conducted following the standard operational procedures of the Institut für Klinische Krebsforschung IKF GmbH am Krankenhaus Nordwest (Frankfurt, Germany). An eCRF for data collection will be carefully maintained for each participant, also reporting serious and non-serious AE according to the common criteria for adverse events (CTCAE, version 5.0), and changes of study treatment throughout the trial. Patients will be provided a diary where the study drug intake will be recorded. Missed doses will be captured by the patient diaries and documented in the eCRF.

After the end of the study period, participants will be proactively followed up for survival and/or treatment-related AE until loss-to-follow-up or withdrawal of study consent. Patients unwilling to return to the study site will alternatively be offered a follow-up by telephone every three months. The treating physicians will ensure that patients receive appropriate further therapies according to SOC.

Statistical analysis

Statistical analysis will follow the International Conference on Harmonization (ICH) Guidelines “Structure and Content of Clinical Study Reports” and “Statistical Principles for Clinical Trials”.

Sample size and study duration

The sample size of n = 116 is primarily determined by considerations of enrollment feasibility. The planned study duration was based on published data from the ALTA-1 L [12] and ALEX [32, 33] phase III trials. In these calculations, the HR for PFS of first-line alectinib vs. crizotinib was set to 0.45 (i.e. the average of 0.47 [32] and 0.43 [33], the HR for PFS of first-line brigatinib vs. crizotinib was assumed identical to that of alectinib vs. crizotinib (based on data of the ALTA-1 L trial with HR = 0.49, 95% confidence intervall (CI) 0.33–0.74 [12]), and the median PFS under first-line crizotinib was considered to be 11 months (i.e. the average of 11.1 [32] and 10.9 months [33]). Consequently, assuming an exponential distribution of PFS, the expected median PFS under first-line alectinib or brigatinib was estimated as 24.4 months (11/0.45), and the proposed duration of the ABP trial was based on an expected total follow-up time of 32 months for the last patient, i.e. an expected first-line PFS of 24.4 months plus an expected second-line PFS of approximately 7 months (based on the median PFS of 5.5–6.9 months [95% CI 2.9–9.5] under lorlatinib after failure of second-generation ALK inhibitors in the EXP3B/4/5 cohorts of a phase II trial [34]). The recruitment period of 36 months was proposed based on the expected number of newly diagnosed ALK+ patients in the centers expected to participate.

To quantify the potential degree of evidence regarding PFS1 that can be gained with a total of 116 patients, we calculated the number of expected events d, the expected 95% CI for the median PFS of alectinib and brigatinib in the first line (assumed to be equal, as explained above), and the expected 95% CI for the HR of PFS under first-line brigatinib vs. alectinib in the ABP trial, given a constant accrual over a time of 36 months, a follow-up time of 32 months for the last patient, and exponentially distributed PFS times. Under these assumptions, the expected number of PFS events is d = 87, the expected 95% CI of the median PFS in the first line is [16.6–34.2 months] (both arms), and the expected 95% CI of the HR for PFS in the first line [0.66–1.52]. The number of events d was calculated using Schoenfeld’s formula [35] and ADDPLAN v6.1 software. The CI for the median PFS was calculated via bootstrap** using 1,000,000 datasets simulated in R v3.3.3 (http://r-project.org) and a fixed random number seed to yield stable and reproducible results, and the CI for the HR was calculated using the (approximate) formula exp. (±1.96√4/d) [36].

Methods of statistical analysis

A Cox proportional hazards model will be used to assess the primary endpoint PFS1. As covariates, the model includes the factor “treatment group” and is adjusted for the presence of brain metastases (yes vs. no) and ECOG (0–1 vs. 2) at baseline. The treatment groups will be compared at a two-sided α of 0.05, and a 95% CI for the HR determined. Furthermore, Kaplan-Meier curves will be provided. Primary analysis will be based on all randomized patients (intention-to-treat population), while sensitivity analyses will be conducted for the per-protocol set (patients without major protocol violations) and for predefined subgroups of secondary and exploratory endpoints.

Analyses of secondary endpoints will be descriptive and will include the calculation of appropriate summary measures of the empirical distributions. AE and SAE will be summarized by relative and absolute frequency and severity grade based on CTCAE V5.0. Summary tables will provide the number and percentage of patients with AE and the 95% CI for the event rates.

A solely descriptive interim evaluation to assess the primary (PFS of first-line treatment) and secondary (iORR, iDOR, iTTP, TNT of first line treatment, PFS of second-line treatment, TNT of second line treatment, OS, QoL, frequency of SAE) endpoints will be performed when 25 patients in each arm have a follow-up of at least 12 months.

All analyses will be carried out using SAS (SAS Institute, Cary, NC) version 9.4 or higher.

Trial status

As of June 21, 2021, 25 study centers have been initiated. The first patient was enrolled on June 18, 2020.

Discussion

The ABP trial has been designed to provide insights and generate hypotheses that will facilitate optimization of management for patients with ALK+ NSCLC. Besides supporting therapeutic decisions for enrolled patients by providing “fast-track” molecular results to the treating physicians, the ABP trial also aims to deepen understanding of the underlying biology and facilitate development of a framework for individualized management of ALK+ NSCLC according to molecular properties. One special focus is the relationship of resistance mechanisms acquired under treatment with the therapeutic compound used and baseline tumor characteristics. Patients with low molecular risk and the perspective of a “chronic disease” will be distinguished from “high-risk” cases, and molecular features of the latter will be utilized to elaborate improved non-invasive methods of disease monitoring and preclinical modeling in order to advance therapeutic strategies.