
==== Front
JTO Clin Res Rep
JTO Clin Res Rep
JTO Clinical and Research Reports
2666-3643
Elsevier

S2666-3643(24)00073-0
10.1016/j.jtocrr.2024.100703
100703
Original Article
ALKTERNATE: A Pilot Study Alternating Lorlatinib With Crizotinib in ALK-Positive NSCLC With Prior ALK Inhibitor Resistance
Itchins Malinda BMedSci, M.B.B.S., FRACP, PhD malinda.itchins@sydney.edu.au
abc∗
Liang Shirley BSc a
Brown Chris MBiostats, BSc d
Barnes Tristan BSc (Med), M.B.B.S., FRACP e
Marx Gavin BSc, M.B.B.S., FRACP fg
Chin Venessa M.B.B.S., FRACP, PhD hij
Kao Steven BHB, MBChB, PhD, FRACP ck
Yip Po Yee MBChB, FRACP, PhD lm
Mersiades Antony J. BMedSc, M.B.B.S., FRACP, MMed (Clin. Epi) de
Nagrial Adnan M.B.B.S., FRACP, PhD nop
Bray Victoria M.B.B.S., FRACP, PhD q
Peters Geoffrey BPharm, M.B.B.S., FRACP rs
Parakh Sagun BSc, MBChB, FRACP, PhD tu
Garg Kavita PhD v
Li Bob T. MD, PhD, MPH w
McKay Matthew PhD x
O'Byrne Kenneth M.B.B.S., FRACP, FRCPA, MD y
John Thomas M.B.B.S., FRACP, PhD zaa
Gill Anthony J. MD, FRCPA ab
Molloy Mark P. PhD bx
Solomon Benjamin J. M.B.B.S., FRACP, PhD zaa
Pavlakis Nick BSc, M.B.B.S., MMed (Clin. Epi), PhD, FRACP ab
a Royal North Shore Hospital, St Leonards, Australia
b Northern Clinical School, University of Sydney, St Leonards, Australia
c Chris O'Brien Lifehouse, Camperdown, Australia
d NHMRC Clinical Trials Centre, University of Sydney, Camperdown, Australia
e Northern Beaches Hospital, Frenchs Forest, Australia
f Sydney Adventist Hospital, Wahroonga, Australia
g Australian National University, Sydney, Australia
h The Kinghorn Cancer Centre, St Vincent’s Hospital Sydney, Darlinghurst, Australia
i The Garvan Institute of Medical Research, Darlinghurst, Australia
j University of New South Wales, Darlinghurst, Australia
k Sydney Medical School, University of Sydney, Camperdown, Australia
l Macarthur Cancer Therapy Centre, Campbelltown Hospital, Campbelltown, Australia
m School of Medicine, Western Sydney University, Campbelltown, Australia
n Crown Princess Mary Cancer Centre, Westmead Hospital, Westmead, Australia
o Blacktown Hospital, Blacktown, Australia
p Westmead Clinical School, University of Sydney, Westmead, Australia
q Liverpool Hospital, Liverpool, Australia
r Canberra Hospital, Canberra, Australia
s Australian National University, Canberra, Australia
t Olivia Newton-John Cancer Research Institute, Austin Hospital, Heidelberg, Australia
u School of Cancer Medicine, La Trobe University, Bundoora, Australia
v Resolution Bioscience, Kirkland, Washington
w Memorial Sloan Kettering Cancer Center, New York, New York
x Kolling Institute, University of Sydney, St Leonards, Australia
y Princess Alexandra Hospital, Woolloongabba, Australia
z Peter MacCallum Cancer Centre, Melbourne, Australia
aa University of Melbourne, Melbourne, Australia
∗ Corresponding author. Address for correspondence: Malinda Itchins, BMedSci, M.B.B.S., FRACP, PhD, Department of Medical Oncology, Royal North Shore Hospital, Reserve Road, St Leonards, NSW 2065, Australia. malinda.itchins@sydney.edu.au
08 7 2024
9 2024
08 7 2024
5 9 1007034 3 2024
9 6 2024
29 6 2024
© 2024 The Authors
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Introduction

ALK-positive lung cancers represent a molecularly diverse disease. With drug exposure, driving selection pressure, and resistance pathways, disease relapse will emerge. There is compelling rationale to investigate novel treatment strategies, informed by dynamic circulating tumor DNA (ctDNA) monitoring.

Methods

The single-arm, pilot study ALKTERNATE investigated fixed alternating cycles of lorlatinib intercalated with crizotinib in individuals resistant to second-generation ALK inhibitors. Dynamic ctDNA explored the correlation with disease response and disease recurrence and defined disease resistance. The primary outcome was time-to-treatment failure, a composite of tolerability, feasibility, and efficacy. Secondary outcomes included standard survival measures, toxicity, pharmacokinetic analysis, and patient-reported outcomes. Tertiary outcomes were proteogenomic analyses of tissue and plasma.

Results

A total of 15 individuals were enrolled; three encountered primary resistance to lorlatinib induction. There were 12 participants who received alternating therapy, and this approach revealed safety, feasibility, and effectiveness. Patient-reported outcomes were maintained or improved on therapy, and toxicity was consistent with previous reports. The pharmacokinetic measures were similar to the single-arm drug experience. Median time-to-treatment failure was 10 months; overall survival was 23 months. ctDNA profiles indicated inferior survival in those with preexistent TP53 mutations and those without clear or cleared ctDNA at trial induction. The study defined a vastly heterogeneous population with an abundance of ALK coexisting with non-ALK resistance variants.

Conclusions

ALKTERNATE revealed feasibility with a novel alternating ALK inhibitor strategy in ALK-positive NSCLC. Results support progressing inquiry into this approach and propose a flexible design with drug(s) selected and alternating time frames, informed by real-time plasma profiling. Moving this concept to treatment naive may also optimize impact.

Keywords

ALK
ALKi
NSCLC
Resistance
ctDNA
Lorlatinib
Crizotinib
==== Body
pmcIntroduction

Since first described in 2007,1 the ALK-rearranged NSCLC community has experienced a revolution in personalized ALK inhibitor (ALKi) therapies—first-, second-, and third-generation ALKis now established in the clinic and fourth-generation ALKis in clinical development.2, 3, 4, 5, 6 Despite median survival of up to 7 years with modern agents,7 drug resistance remains inevitable. Acquired resistance to ALKis may manifest through “ALK-dependent” on-target mutations, and “ALK-independent” bypass tract activation,8,9 or rarely histologic transformation.10

Newer generation ALKis promptly superseded first-generation crizotinib as a new standard first-line ALKi. It was shortly thereafter described the array of ALKi disease-resistant phenotypes may be highly variable depending on the unique ALKi exposure(s).8 This phenomenon is in part due to the unique selection pressure applied by each structurally distinct ALKi in this molecularly heterogeneous disease.11,12

Until recently, second-generation ALKis, most frequently alectinib, have been the mainstay first-line therapy globally based on compelling efficacy and superior tolerability.3,13,14

Third-generation lorlatinib was next established as an empirical treatment in ALKi-refractory disease.15 In pretreated individuals, it became apparent those with ALK-resistant mutations derived greater benefit and perhaps those who harbored a variant 3 EML4-ALK fusion performed favorably compared with variant 1.1,16 Furthermore, those who had only been exposed to one-line second-generation ALKi had shorter median progression-free survival (PFS) than those who received multiple prior lines.15

With exposure to more potent on-target, next-generation ALKis, resistance may more frequently manifest with the emergence of bypass tract, “off target” activation, which has been described through multiple pathways, including EGFR, MET, HER2, KRAS, PIK3CA, BRAF, NRAS, and IGF1-R. Most often, these activated resistance pathways co-occur with aberrations in tumor-suppressor gene TP53.8,17, 18, 19, 20, 21 With latter line lorlatinib, in particular, MET dysregulation or compound ALK mutations occur frequently.22 Crizotinib is also an active MET inhibitor. Compound ALK mutations, most often including ALK G1202R, may be resistant to the established ALKis.23

Circulating tumor (ct)DNA has emerged as a valuable minimally invasive tool for temporal genomic analysis in ALK-positive (ALK+) NSCLC.24 This modality provides potential to capture intratumoral, intertumoral, and interindividual heterogeneity in a timely manner. The potential utility in predicting performance on ALKis based on plasma kinetics is reported with ctDNA in ALK+ disease.25, 26, 27

At present, tissue and less commonly blood sampling have been relied on to determine the molecular basis for disease progression, identify phenotypic transformation, and allow selection of the most appropriate therapy.28,29 Treatment at progression however continues for most to be recommended empirically or “blindly” to the individual’s cancer biology.

Given the pace of drug development in ALK+ lung cancer, next-generation ALKis have not been compared head-to-head in randomized studies and are unlikely to ever be so. Lorlatinib has recently emerged as the ALKi with the greatest magnitude of impact first line in terms of PFS and central nervous system (CNS) activity.30 Data describing drug resistance profiles with first-line lorlatinib are awaited. What has been found is that the presence of TP53 co-mutations leads to inferior survival.31

ALKTERNATE hypothesized that alternating ALKi therapy may alter intrinsic ALKi selection pressure and suppress or prevent the expansion of preexistent drug-resistant subclones or the emergence of acquired resistant clones to lorlatinib, thus delaying clinical progression.32 Furthermore, ctDNA could monitor and detect the molecular mechanisms of resistance before clinical and radiological progression. The study design was informed by two case reports, one revealing the role of serial biopsies in a heavily treated ALK+ patient, identifying paradoxical resensitization to crizotinib after lorlatinib resistance,33 and another revealing the utility of ctDNA in resistance profiling over time.24 Data emerging regarding MET dysregulation as a bypass pathway with lorlatinib further informed the rationale to intercalate the ALKi/MET inhibitor crizotinib.22

Materials and Methods

Study Design and Patients

The study design was a proof-of-concept, open-label, single-arm, translational study investigating a fixed schedule of alternating lorlatinib with crizotinib in individuals with ALK+ NSCLC who had received any number of prior lines of therapy for advanced disease, provided they had experienced disease progression on at least one second-generation ALKi. Individuals enrolled required extracranial Response Evaluation Criteria in Solid Tumors (RECIST) version (v.) 1.1 measurable disease34; CNS and leptomeningeal were eligible if asymptomatic, or treated and stable, and magnetic resonance imaging of the brain mandated. The washout period from previous systemic chemotherapy or radiation treatment was 2 weeks and from ALKi 4 days. Eastern Cooperative Oncology Group performance status 0 to 1 was required. Previous crizotinib therapy was allowed provided it was not the most recent line and there was no previous intolerability.

The study was conducted by the Thoracic Oncology Group of Australasia at two Australian tertiary referral centers—Royal North Shore Hospital (RNSH) and Sydney and Peter MacCallum Cancer Centre, Melbourne, Australia. The trial was approved for by the local ethics review board (NSLHD 2019/ETH00389) and registered with Australian New Zealand Clinical Trial Registry (ANZCTR): ACTRN12619000844145. All patients provided written informed consent.

Treatment and Assessments

At enrollment, patients required plasma with or without tissue sampling for biomarker analysis. Plasma samples for ctDNA assessment occurred at baseline, three monthly on study, and at disease progression. Baseline magnetic resonance imaging of the brain and computed tomography-chest, abdomen, pelvis were performed before patients received 100 mg daily of lorlatinib for 3 months, as induction therapy (cycle 1), and then alternating treatment with 1 month crizotinib 250 mg twice daily, and then 2 months lorlatinib 100 mg daily, with dose interruptions and reductions as clinically indicated.

Patients required at least disease control and no progression after 3 months of lorlatinib (cycle 1), to be eligible to enter the “alternating” phase of therapy.

Given anticipated drug-drug pharmacokinetics, before treatment switch, ALKi was ceased for 48 hours, and at cycle 1 to 2a (lorlatinib switch to crizotinib) and cycle 2a to 2b (crizotinib to lorlatinib), blood samples for pharmacokinetic (PK) analysis were collected.

Alternating therapy was continued until disease progression by RECIST v.1.1, intolerability, study withdrawal, or death.

The trial schema is available in Supplementary Appendix a.Figure 1.

Study End Points and Translational Methods

The primary outcome measure was time-to-treatment failure (TTTF) with alternating ALKi therapy, defined as the time from treatment initiation (post +3 months induction lorlatinib) to treatment discontinuation from any cause. Treatment failure was deemed the appropriate end point given this pilot was assessing feasibility of delivering a novel treatment schedule. Only those who enter alternating therapy were included in the primary outcome analysis. Secondary outcomes included overall survival (OS), PFS on study, and post-progression (PFS2), including crossover to continuous lorlatinib. Objective tumor response was also assessed.

Toxicity was assessed through standard Common Terminology Criteria for Adverse Events (CTCAE) v.4.03 criteria. Patient-reported outcomes were recorded through Functional Assessment of Cancer Therapy Lung v.4 questionnaires. To focus on the tolerability of lorlatinib with this treatment approach, mental health was assessed through the Beck’s Depression Inventory Questionnaire-II and cognitive assessments through the MiniCog quick screening test.

For the tertiary outcome evaluating temporal ctDNA profiles, samples were sequenced with Resolution Bioscience lung ctDx v.8, Resoluton Bioscience, Kirkland, WA, USA. Time points baseline, 6 monthly, and disease progression milestones were analyzed real time, and the remaining time points analyzed retrospectively including 3 months before disease progression on alternating therapy.

For correlative tertiary analysis, tissue core biopsies were obtained if feasible, and formalin fixed, paraffin embedded. Reconfirmation of histologic phenotype and ALK+ immunohistochemistry status were required during screening, with adenocarcinoma-predominant disease required. Tissue was sequenced retrospectively through the Trusight Oncology 500 gene (TSO500) next-generation sequencing assay, in house (Peter MacCallum Cancer Centre).

The PK plasma samples were collected and analyzed, with quantitation at Labcorp Pharmaceutical R&D (Shanghai, People’s Republic of China) using validated methods. Blood samples were collected at 1 hour, 2 hours, 4 hours, 6 hours, and 24 hours on cycle 2 (cycle 1-2a and cycle 2a-2b switches).

For plasma proteome exploration, digests were analyzed in random order by liquid chromatography and tandem mass spectrometry in data-independent acquisition mode using a nano-ultrapressure liquid chromatography system (Ultimate 3000 RSLCnano, Thermo Fisher Scientific) coupled to a Q-Exactive HF-X orbitrap mass spectrometer (Thermo Fisher Scientific).35, 36, 37 Further methodology is available in Supplementary Appendix b.

Statistical Analysis

Study sample size for the pilot was not based on a statistical outcome, with a maximum of 20 patients sought.

Survival, time-to-event Kaplan-Meir plots, and a swimmer plot were calculated and composed using R v.4.1.3 software.

Descriptive analysis of qualitative and quantitative ctDNA profiles are reported, with correlation to survival. Individual patient time point variant allelic frequency (VAF) plots were extracted from the RESOLUTION BIO HUB reporting software.

The PK parameters, including Cmax and AUC0–6, were calculated by noncompartmental analysis methods using the WinNonlin software package (v.8.3.4 Pharsight Corporation, Sunnyvale, CA) for PK analytes crizotinib and lorlatinib.

Differentially abundant proteins were assessed using two-sided t test in Perseus (v.1.6.5.0) with p less than 0.05 and presented with GraphPad Prism v.10 (Dotmatics, Boston, MA, USA).

Study Amendment

The study was approved by the local ethics board at lead site RNSH and opened in July 2019. Delays in accrual occurred with recent approval of first-line use of alectinib in Australia delaying disease progression and the coronavirus disease (COVID19) pandemic. The study completed enrollment in May 2023. The trial was amended to allow for remote assessments by telehealth and local drug delivery during the pandemic.

Results

Patients

There were 21 individuals screened, and 15 were enrolled (Fig. 1). Those not pursuing study included two identified through ctDNA to harbor non–ALK-driven disease, two did not have RECIST v.1 measurable disease, one withdrew consent during screening, and one had persistent elevated bilirubin level.Figure 1 Consort diagram of patient screening, accrual, and progress on the study. ∗Two patients were deemed trial ineligible at the TMC discretion due to baseline ctDNA indicating disease unlikely ALK driven: ALK01-04 plasma, 10 variants detected, no ALK-fusion; ALK01-11 KRAS G12C detected, confirmed on tissue biopsy, no ALK fusion. ˆThree patients with primary resistance to lorlatinib harbored an EML4-ALK V1, two with co-occurring TP53 in the plasma and tissue, and one with MYC CNG present in tissue biopsy (ALK01-17). ctDNA, circulating tumor DNA; TMC, trial management committee.

Three (20%) experienced primary resistance to lorlatinib and were not eligible for alternating therapy. There were 12 patients who proceeded to alternating ALKi therapy.

Median age was 67 years, 60% male, almost 25% Asian, 50% never-smokers, and 50% were Eastern Cooperative Oncology Group of 0. In addition, 40% harbored CNS metastases at trial inclusion with 66% not having received previous local therapy. The median number of previous lines of therapy was two, with almost 50% receiving multiple previous lines of ALKi. Demographics are detailed in Table 1.Table 1 Patient Demographics

Characteristic	N = 15	
Age, y	n (%)	
 Median	67	
 Range	34–77	
Sex		
 Male	9 (60)	
 Female	6 (40)	
Race		
 Asian	4 (27)	
 Non-Asian	11 (73)	
Smoking		
 Never	7 (47)	
 Former	8 (53)	
 Current	0	
ECOG performance status		
 0	7 (47)	
 1	8 (53)	
CNS metastases, no. (%)		
 Yes	6 (40)	
 No	9 (60)	
Prior CNS treatment		
 Stereotactic radiosurgery	1 (17)	
 Whole brain radiotherapy	1 (17)	
 No treatment	4 (66)	
Previous systemic therapies		
 Median (range)	2 (1–5)	
 Second-generation ALK inhibitor only	7 (46)	
 >1 prior ALK inhibitors	7 (46)	
Previous PE/VTE		
 Yes	3 (20)	
Detectable ctDNA at baseline		
 Yes	13 (87)	
 No	2 (13)	
CNS, central nervous system; ctDNA, circulating tumor DNA; ECOG, Eastern Cooperative Oncology Group; PE, Pulmonary embolus; VTE, Venous thromboembolism.

Survival

At data lock on June 30, 2023, median follow-up was 31 months, 83% (N = 10) had experienced “treatment failure” with alternating therapy, and 60% had died.

The median (m) time from diagnosis to trial enrollment was 27 months. The range of previous lines of therapy was one to five. Time on trial ranged from 3 to 42 months. The swimmer plot in Figure 2 illustrates the therapies received and duration on treatment before enrollment, during the study, and after the study, including individuals’ TTTF on alternating treatment.Figure 2 Swimmer plot of participants and treatments received and duration of treatment benefit before and after study commencement (y = 0 is ALKTERNATE trial commencement).

The median TTTF was 11 months (n =12, 8.4 mo– not reached [NR]), 8 months from initiation of alternating therapy (Fig. 3A). PFS1, all-comers, was 6.1 months (n = 15, 5.4 mo–NR) (Fig. 3C). PFS2, those crossed over to continuous lorlatinib, was 3.1 months (n = 9 2.1 mo–NR). Objective response rate was 53%, all partial responses. Four (27%) achieved stable disease as best response and three (20%) progressive disease. Median OS was 23 months (18 mo–NR) and 26 months in alternating (Fig. 3C). Median time from diagnosis to death was 40 months.Figure 3 Kaplan-Meir survival plots. (A) Time-to-treatment failure for alternating therapy; (B) overall survival in all enrolled; (C) progression free survival in all enrolled; (D) progression-free survival based on TP53 trial entry status; and (E) progression-free survival in those with clear ctDNA at trial entry or after one cycle of alternating treatment versus those with persisting detectable ctDNA. ctDNA, circulating tumor DNA. ctDNA, circulating tumor DNA.

In those without baseline TP53 mutations, mPFS was 27.0 months versus 5.5 months in those with p equals 0.025 (Fig. 3D). mPFS in those who had clear plasma (ctDNA) at diagnosis or after 1 cycle of alternating therapy was 24 months versus 5.9 months in those who did not have clear plasma (p = 0.14) (Fig. 3E). There was no difference in PFS in those with ALK+-resistant mutations detected at baseline versus not, nor by ALK fusion variant, clot, CNS status, or previous lines of therapy.

A total of 50% on alternating (n = five of 10) progressed in the CNS, two without previous CNS disease. All with CNS progression were managed with local stereotactic radiotherapy.

Tolerability

Safety

There were no grade 4 or 5 toxicities encountered, 60% experienced a grade 3 treatment-related toxicity, and 40% grade 2 as their highest-grade adverse event (AE). Dose interruption for grade 3 AEs ranged from 1 day to 17 days. Dose reduction was required in four (27%) on lorlatinib due to delirium, mood disturbance, word-finding difficulties, and unconfirmed seizure (later attributed to a hemorrhagic stroke on anticoagulation) and one (7%) on crizotinib due to esophagitis, prompting an early protocol amendment to mandate medication with food (Table 2). The highest-grade and most frequent toxicities are detailed in Table 2. Most toxicities in the study were grade 1; the most common are weight gain, peripheral edema, and hypercholesterolemia. Weight gain ranged from 5% to 25%. In addition, syncope encountered by two patients in the study was potentially attributed to a drug, but it is not confirmed.Table 2 Safety, Tolerability, and Treatment-Related Adverse Events

Safety and Tolerability	N = 15	
Worst-grade toxicity, n (%)		
 2	6 (40)	
 3	9 (60)	
Dose discontinuation	0 (0)	
Dose reduction		
 Lorlatinib	4 (27)a	
 Crizotinib	1 (7)b	
Potential treatment-related adverse events—most frequent adverse events (no. participants)	
Event		Grade		
	1	2	3	
Syncope			2	
Paresthesia	5		1	
Hypercholesterolemia	4	5	1	
Weight gain		3	1	
Dyspnea	3	3	1	
Pneumonia			2	
Esophagitis			1	
Hypertension		3		
Bone pain	4	2		
Rash	4	2		
Upper respiratory tract infection	3	2		
Cough	2	2		
Dysgeusia	1	2		
Urinary tract infection		2		
Peripheral neuropathy	7	1		
Diarrhea	7	1		
Nausea	6	1		
Arthralgia	5	1		
Fatigue	4	1		
Peripheral edema	11			
Dizziness	5			
Transaminitis	4			
Vomiting	3			
Mood lability	3			
Word finding difficulty	3			
Constipation	3			
Flatulence	3			
Anorexia	3			
Abdominal pain	3			
a Dose reduction on lorlatinib due to delirium, mood disturbance, word-finding difficulties, and unconfirmed seizure (later attributed to a hemorrhagic stroke while on direct oral anticoagulation, unprovoked).

b One patient was dose reduced on crizotinib due to esophagitis and re-educated to take with food.

Pharmacokinetics

Plasma AUC0–6, Cmax, and Tmax for both crizotinib and lorlatinib were similar compared with historical single-dose data. Other PK parameters could not be reliably estimated because sampling was limited to 6 hours (the 24-h sample was inconsistently collected pre-dose, which is a study oversight).

Mean and SD plasma concentrations (ng/mL) of crizotinib and lorlatinib from 0 to 6 hours post-drug switch are available in Supplementary Appendix c.Table 1. and d.Figure 2.

Patient-Reported Outcomes

Quality of life was maintained or improved throughout and compared consistently with the adult population reference value.38 Complete results of the Functional Assessment of Cancer Therapy Lung assessments are available in the Supplementary Appendix e.Figure 3.i.–vii.

One patient scored in the depression range on questionnaire assessment, including at baseline. The trial reporting intervals did not allow capture of fluctuations in mood around drug switches after the first 6 months. Anecdotally, multiple individuals volunteered their mental state felt “better,” both mood and cognition, on their month of crizotinib, and this was noted with minor improvements in scoring reported in the first crizotinib switch.

Cognition screening revealed that all but one experienced a degree of cognitive impairment, mild to moderate, which revealed fluctuation on the study and a degree of improvement on crizotinib in the first 6 months when assessed. This observation was neither captured under the standard CTCAE formal reporting nor able to be evaluated with statistical significance.

Biomarker Analysis

Circulating Tumor DNA

A total of 89 samples were analyzed with a reporting success rate of 97%.

There were 13 individuals (87%) who harbored detectable ctDNA at baseline. Two screen ineligibilities were based on the ctDNA profile, one reported a KRAS G12C mutation with no ALK fusion (confirmed then on biopsy); the second patient harbored 12 variants in ctDNA with no ALK fusion, in retrospect the standard-of-care testing deemed falsely positive.

The three patients (20%) with primary resistance to lorlatinib harbored an EML4-ALK variant 1 fusion, two with co-occurring TP53 mutations and no further cause for resistance in ctDNA.

The ALK-fusion variant was defined in 14 of 15 through ctDNA (Fig. 4A). There were 19 differing ALK-resistant mutations detected throughout, with overlapping variants detected in five, five harboring ALK G1202R, three ALK G1269A, and two ALK F1164L, I1171M, L1196M, and V1180L. The range of ALK mutations in an individual sample was zero to five (Fig. 4B). A spectrum of non-ALK variants were also detected across the cohort (Fig. 4C). One individual harbored a germline TP53 P72A mutation.Figure 4 Qualitative ctDNA and the frequency of variant capture (A) for the ALK fusion variant, (B) ALK mutations detected in the blood and frequency across individuals, (C) non-ALK variants detected in the blood and frequency across individuals, and (D–H) five individual patient temporal ctDNA allelic frequency plots revealing variant quantitatively. CNG is not captured in VAF plots presented, (D). Harbored MET CNG at baseline (6 copies) and PD continuous lorlatinib (four copies). ctDNA, circulating tumor DNA.

Two patients had low-level MET copy number gain (CNG) at trial entry, none at progression on alternating therapy, and one encountered MET CNG on crossover continuous lorlatinib (17 copies). This same individual and one other had a new MET point mutation at alternating progression disease (PD). Four (40% of those with PD) developed compound ALK-resistant mutations at progression on alternating therapy.

Five (42%) had clear plasma at entry or cleared their detectable ctDNA in plasma on treatment (Fig. 4D–H). The remaining on alternating therapy experienced a decrease in VAF with therapy. Two individuals with primary resistance to lorlatinib did not have a decrease in VAF, and the third did not undergo a progression sample collection.

Seven (70%) had an increase in the abundance of detectable preexisting ctDNA variants or new variants detected preceding radiological progression on alternating therapy (Fig. 4A, B, and D–H). Nine (90%) identified new non-ALK variants on alternating treatment and five (50%) new ALK-resistant variants.

The ctDNA profiles and kinetics were unique to each case. Figure 4D to H illustrates the utility of ctDNA profiling in defining the molecular phenotype, monitoring disease response, and determining disease resistance, before radiological disease progression in case examples.

The complete set of ctDNA results is available in Supplementary Appendix f.Table 2.

Tissue Biopsy Evaluation

Ten of 15 individuals (66%) had a tissue biopsy at inclusion and seven of 13 (54%) at disease progression (54%). All were confirmed to remain ALK+ and adenocarcinoma-predominate phenotype.

The tissue genomic sequencing failure rate was 53% due to insufficient sample. Notable findings complementing ctDNA results in those analyzable included an STRN3-NTRK2 fusion, co-occurring with an EML4-ALK (V1) fusion in a resected progressing CNS metastasis, not detected in matched time point ctDNA. A participant with primary systemic progression on lorlatinib harbored EML4-ALK (V1) and MYC CNG (10 copies), not identified in ctDNA.

One individual harbored mixed large cell neuroendocrine histology with adenocarcinoma at entry and, then, at disease progression, new squamous differentiation and an emergent ALK G1202R, not present at baseline tissue. Matched plasma detailed an abundance of additional variants, including ALK G1202R, to be preexistent, ALK V1180L, EGFR amplification, and four TP53 variants, also confirming EML4-ALK variant 3. At progression, the ctDNA VAFs were lower in abundance. The squamous differentiation deemed likely driving disease.

One individual exhibited radiological progression in the left side of chest wall axilla ipsilateral to the site of definitive radiation years prior. Fluorodeoxyglucose-positron emission tomography confirmed avid disease. Core biopsy revealed necrosis and abscess material, with no evidence of malignancy. The individual remained on study treatment for a further 13 months given disease control.

Tissue sample status and next-generation sequencing results are available in Supplementary Appendix g.Table 3.

Plasma Proteome Analysis

Plasma proteome data across the treatment continuum were obtained to explore novel biomarkers. A yield of 379 proteins could be used for quantitative analyses which confirmed feasibility of this method. Proteome signatures mostly clustered by individual with variability mapped over time (Supplementary Appendix h.Fig. 4).

Exploratory analyses were conducted in clinically meaningful groups of interest. The baseline level of coagulation factor IX (FA9) was significantly elevated throughout (p = 0.02), most abundant at progression in those who experienced a shorter TTTF (below median). Coagulation factors vitamin K-dependent protein C and S (PROC, p = 0.02, PROS, p = 0.03) were also predictive of TTTF with higher levels experienced in lower survival. Similarly, in the TP53+ group (significantly lower survival), platelet-activating factor acetylhydrolase (PLA2G7) was more abundant (p = 0.01) (Supplementary Appendix i.Fig. 5.i–ii).

Individuals with an ALK-dependent resistance profile at trial entry identified unique proteins to those with an ALK-independent co-mutational signature. The ALK-dependent plasma proteome was enriched with extracellular matrix (ECM) proteins (p < 0.05) including vinculin (VCL), vitronectin (VTN), osteonenctin (SPARC), periostin (POSTN), collectin (COLLEC11), olfactomedin (OLFM1), and tenascin X (TNXB) (Supplementary Appendix i.Fig. 5.iii–vii) and in TP53+ ECMs nidogen-1 (NID1, p = 0.02) and fibulin-5 (FBLN5, p = 0.02) (Supplementary Appendix i.Fig. 5.viii–xii).

Discussion

This pilot study reveals the feasibility, safety, and activity of alternating ALKi therapy with lorlatinib and crizotinib in an ALKi-refractory population. Crizotinib was used for its potential to maintain and prolong sensitivity to lorlatinib and suppress the activation of MET signaling as a prevalent resistance pathway.

The primary outcome measure being the composite outcome of tolerability, feasibility, and efficacy, TTTF supported this being a deliverable strategy. Survival was comparable with the single-arm lorlatinib registrational data, but not superior. Alternating ALKi therapy did not compromise quality of life with the pattern and frequency of toxicities observed in keeping with those expected. PK concentrations were not compromised with drug switching.

This proof-of-concept translational study illustrates the potential impact of personalized care with both tissue and liquid biopsy, including the feasibility of ctDNA in routine care. Collating this approach also confirmed the heterogeneity of “ALK+ resistance”. The ctDNA signatures helped inform optimal recruitment and correlated with performance on trial, whereas assessment of kinetics predicted superior performance in the study and altered preceding a relapse in disease. Unlike previous reports, MET was not an identified resistance mechanism in ALKTERNATE, as has been the experience with continuous lorlatinib.22 This may in part be due to intercalated crizotinib exposure. A small series has reported modest clinical activity and general tolerability in adding crizotinib to lorlatinib in those identified to have MET dysregulation.39 In addition, there were no notable unifying non-ALK variants emergent on ALKTERNATE, except for TP53 mutations frequent before the study and dominant at progression. The alternating ALKi method did not diminish in the frequency of ALK-compound resistance mutation found with lorlatinib monotherapy, experienced in 50% in the study.

Close CNS surveillance was performed to monitor for CNS disease progression, but it accounted for 50% of the progression events on alternating therapy. None were found to progress in CNS on a crizotinib cycle and all were asymptomatic. Four patients with CNS PD were successfully treated with stereotactic radiotherapy and all crossed over to continuous lorlatinib with a period of CNS disease control. With continuous lorlatinib latter line, the probability of the first event being CNS progression was consistently lower than the first event being non-CNS progression.40 Managing the CNS is paramount in ALK+ disease, and the crizotinib cycles were designed to be short (1 mo) given inferior CNS penetration. The progression of this study would be alternating in more CNS-penetrant, next-generation ALKis, while also moving this treatment strategy to investigation in the frontline setting to increase impact.

An interest with ALKTERNATE is the potential to mitigate drug AEs, especially the chronic low-grade toxicities given the inbuilt individual drug “holidays.” For example, the cognitive effects of lorlatinib are reported to be experienced in at least 23%.41 The trial design did not enable robust objective assessment of this; however, there was a trend for improvement in cognition and mood in those assessed with drug switch off lorlatinib. It is intriguing that this cognitive impairment was not captured in the CTCAE reporting, as not volunteered by the reporting individual or adequately captured in the survey. Four patients (33%) were dose reduced on lorlatinib in alternating therapy, each with neurologic adverse events. Future optimized objective assessment of the impact of therapy “switches” enabling drug “holidays” as a strategy to mitigate toxicity could add appeal to this strategy. This would be particularly of interest with lorlatinib and chronic low-grade neurocognitive toxicities or weight gain as important examples.

The exploratory plasma proteome analysis found almost 400 proteins quantifiable in the clinically meaningful groups analyzed, which is substantial in the small sample set. Many of the putative biomarkers are suitable for immunoassay in future studies which may move the field beyond genomics. Altered protein profiles in the coagulation pathway, ECM including inflammation and angiogenesis, and apolipoproteins involved in lipid trafficking and proliferation were reported in clinically distinct subgroups. Cancer portends a hypercoagulable state, and indeed ALK+ lung cancers are found to have increased clotting rates, including venous thromboembolism.42,43 Such an approach may enable future prediction of events informed by the proteomic signature.

The main shortcoming of this study is its small sample size with the reported observations being exploratory and hypothesis generating. The initial findings of ALKTERNATE in approaching this novel treatment strategy will serve as a foundation to inform future expansion of this approach.

Recently, lorlatinib has become the standard of care for the frontline treatment of all first-line ALK+ diagnoses, shifting the field from investigating latter line lorlatinib treatment strategies. There is compelling scientific rationale to test bespoke combination escalation and de-escalation drug strategies informed by real-time ctDNA qualitative and quantitative signatures with a proposed approach to this “ALKTERNATE 2.0” presented in Supplementary Appendix j.Figure 6. For example, TP53 mutations are reported in 37% of patients with ALK+ de novo44 and in 73% in the present pretreated study, which has also revealed a clinically and statistically significant inferior survival in those with preexisting TP53 mutations.45 This supports the mounting biological rationale and evidence to expand exploration in a combination approach with ALKi given with anti-angiogenic therapy as there is an up-regulation of vascular epithelial growth factor pathways in TP53-activated pathways, with or without chemotherapy, as monotherapy is simply insufficient.46,47

ALKTERNATE strongly suggests a “one size fits all” universal therapeutic monotherapy approach, even with fixed switch therapy, delivers unpredictable efficacy, and requires ongoing biomarker-informed exploration in pretreated ALK+ disease. In the study, there was an array of unique molecular phenotypes, particularly at disease relapse. To elaborate with case examples, one individual harbored the preexisting ALK fusion with an emergent NTRK fusion in a progressing resected brain metastasis, one encountered histologic squamous transformation, and one compound ALK mutations ALK C1156Y, ALK G1269A, and ALK I1171M. These three unique cases suggest three different optimal next therapeutic pathways, such as an ALKi plus NTRKi, chemotherapy, and a fourth-generation ALKi trial versus ALKi plus chemotherapy, respectively. At present, these people would all be treated with the same therapy under “standard of care”. We support an N = 1 biomarker-informed treatment approach in ALKi drug resistant cases to progress the field in this molecularly distinct group of NSCLCs, and across oncogene-positive disease.48

Conclusion

The ALKTERNATE study revealed a positive signal for safety, preserved QOL, and comparable survival to historical data with fixed alternating ALKi therapy in treatment-refractory disease, while revealing ctDNA to be therapeutically informative and a harbinger of disease progression. Expanded exploration of this novel treatment approach is supported, including in treatment naive. A future proposed study would use real-time ctDNA profiling in the first-line setting to inform the drug selection of alternating brain-penetrant ALKi monotherapies or combination treatment, including non-ALKis, such as chemotherapy, in a dynamic biomarker-driven manner.

CRediT Authorship Contribution Statement

Malinda Itchins: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing - Original Draft, Writing - Review & Editing, Visualization, Supervision, Project administration, Funding acquisition.

Shirley Liang: Validation, Formal analysis, Investigation, Resources, Data curation, Writing - Review & Editing, Visualization, Project administration.

Chris Brown: Software, Validation, Formal analysis, Resources, Data curation, Writing - Review & Editing, Visualization.

Tristan Barnes: Resources.

Gavin Marx: Resources.

Venessa Chin: Resources.

Steven Kao: Resources.

Po Yee Yip: Resources.

Antony J. Mersiades: Methodology.

Adnan Nagrial: Resources.

Victoria Bray: Resources.

Geoffrey Peters: Resources.

Sagun Parakh: Resources.

Kavita Garg: Formal analysis, Resources.

Bob T. Li: Resources.

Matthew McKay: Resources.

Ken O’Byrne: Methodology.

Anthony J. Gill: Resources.

Mark Molloy: Resources, Writing - review & editing, Visualization.

Tom John: Methodology.

Ben J. Solomon: Conceptualization, Investigation, Resources, Writing - review & editing, Funding acquisition.

Nick Pavlakis: Conceptualization, Methodology, Validation, Investigation, Resources, Writing - review & editing, Supervision, Funding acquisition.

Disclosure

ALKTERNATE received grant funding and provision of drug from Pfizer. The following authors have received personal financial payment from Pfizer for consultancy and/or honoraria: MI, AN, SK, AM, BJS, NP.

Supplementary Data

Supplementary Material

Supplementary Appendix Figure 1

Supplementary Appendix Figure 4

Supplementary Appendix Figure 6

Acknowledgments

Grant funding for study conduct and provision of drug was from Pfizer.

Authorship individual financial disclosures were included in submission documents.

Cite this article as: Itchins M, Liang S, Brown C, et al. ALKTERNATE: a pilot study alternating lorlatinib with crizotinib in ALK-positive NSCLC with prior ALK inhibitor resistance. JTO Clin Res Rep. 2024;5:100703.

Note: To access the supplementary material accompanying this article, visit the online version of the JTO Clinical and Research Reports at www.jtocrr.org and at https://doi.org/10.1016/j.jtocrr.2024.100703.
==== Refs
References

1 Soda M. Choi Y.L. Enomoto M. Identification of the transforming EML4-ALK fusion gene in non-small-cell lung cancer Nature 448 2007 561 566 17625570
2 Solomon B.J. Mok T. Kim D.W. First-line crizotinib versus chemotherapy in ALK-positive lung cancer N Engl J Med 371 2014 2167 2177 25470694
3 Peters S. Camidge D.R. Shaw A.T. Alectinib versus crizotinib in untreated ALK-positive non-small-cell lung cancer N Engl J Med 377 2017 829 838 28586279
4 Camidge D.R. Kim H.R. Ahn M.J. Brigatinib versus crizotinib in ALK-Positive non-small-cell lung cancer N Engl J Med 379 2018 2027 2039 30280657
5 Shaw A.T. Bauer T.M. de Marinis F. First-line lorlatinib or crizotinib in advanced ALK-positive lung cancer N Engl J Med 383 2020 2018 2029 33207094
6 Johnson M.L. Ou S.I. Felip E. NVL-655, a selective anaplastic lymphoma kinase (ALK) inhibitor, in patients with advanced ALK-positive solid tumors: the phase I/II ALKOVE-1 study J Thorac Oncol 18 2023 S35 S88
7 Mok T. Camidge D.R. Gadgeel S.M. Updated overall survival final progression free- survival data for patients with treatment-naïve ALK-positive non-small cell lung cancer in the ALEX study Ann Oncol 31 2020 1056 1064 32418886
8 Gainor J.F. Dardaei L. Yoda S. Molecular mechanisms of resistance to first- and second-generation ALK inhibitors in ALK-rearranged lung cancer Cancer Discov 6 2016 1118 1133 27432227
9 Lovly C.M. Shaw A.T. Molecular pathways: resistance to kinase inhibitors and implications for therapeutic strategies Clin Cancer Res 20 2014 2249 2256 24789032
10 Balla A. Khan F. Hampel K.J. Aisner D.L. Sidiropoulos N. Small-cell transformation of ALK-rearranged non-small-cell adenocarcinoma of the lung Cold Spring Harb Mol Case Stud 4 2018 a002394
11 Frankell A.M. Dietzen M. Al Bakir M. The evolution of lung cancer and impact of subclonal selection in TRACERx Nature 616 2023 525 533 37046096
12 Quintanal-Villalonga Á. Chan J.M. Yu H.A. Lineage plasticity in cancer: a shared pathway of therapeutic resistance Nat Rev Clin Oncol 17 2020 360 371 32152485
13 Nakagawa K. Hida T. Nokihara H. Final progression-free survival results from the J-ALEX study of alectinib versus crizotinib in ALK-positive non-small-cell lung cancer Lung Cancer 139 2020 195 199 31812890
14 Zhou C. Kim S.W. Reungwetwattana T. Alectinib versus crizotinib in untreated Asian patients with anaplastic lymphoma kinase-positive non-small-cell lung cancer (ALESIA): a randomised phase 3 study Lancet Respir Med 7 2019 437 446 30981696
15 Solomon B.J. Besse B. Bauer T.M. Lorlatinib in patients with ALK-positive non-small-cell lung cancer: results from a global phase 2 study Lancet Oncol 19 2018 1654 1667 30413378
16 Lin J.J. Zhu V.W. Yoda S. Impact of EML4-ALK variant on resistance mechanisms and clinical outcomes in ALK-positive lung cancer J Clin Oncol 36 2018 1199 1206 29373100
17 Katayama R. Shaw A.T. Khan T.M. Mechanisms of acquired crizotinib resistance in ALK-rearranged lung cancers Sci Transl Med 4 2012 120ra17
18 Tanizaki J. Okamoto I. Okabe T. Activation of HER family signaling as a mechanism of acquired resistance to ALK inhibitors in EML4-ALK–positive non–small cell lung cancer Clin Cancer Res 18 2012 6219 6226 22843788
19 Lovly C.M. McDonald N.T. Chen H. Rationale for co-targeting IGF-1R and ALK in ALK fusion-positive lung cancer Nat Med 20 2014 1027 1034 25173427
20 Shi R. Martins Filho S.N. Li M. BRAF V600E mutation and MET amplification as resistance pathways of the second-generation anaplastic lymphoma kinase (ALK) inhibitor alectinib in lung cancer Lung Cancer 146 2020 78 85 32521388
21 Talwelkar S.S. Mäyränpää M.I. Schüler J. PI3Kβ inhibition enhances ALK-inhibitor sensitivity in ALK-rearranged lung cancer Mol Oncol 17 2023 747 764 36423211
22 Dagogo-Jack I. Yoda S. Lennerz J.K. MET alterations are a recurring and actionable resistance mechanism in ALK-positive lung cancer Clin Cancer Res 26 2020 2535 2545 32086345
23 Yoda S. Lin J.J. Lawrence M.S. Sequential ALK inhibitors can select for lorlatinib-resistant ompound ALK mutations in ALK-positive lung cancer Cancer Discov 8 2018 714 729 29650534
24 McCoach C.E. Blakely C.M. Banks K.C. Clinical utility of cell-free DNA for the detection of ALK fusions and genomic mechanisms of ALK inhibitor resistance in on-Smsall cell lung cancer Clin Cancer Res 24 2018 2758 2770 29599410
25 Angeles A.K. Christopoulos P. Yuan Z. Early identification of disease progression in ALK-rearranged lung cancer using circulating tumor DNA analysis NPJ Precis Oncol 5 2021 100 34876698
26 Horn L. Whisenant J.G. Wakelee H. Monitoring therapeutic response and resistance: analysis of circulating tumor DNA in patients with ALK+ lung cancer J Thorac Oncol 14 2019 1901 1911 31446141
27 Soo R.A. Martini J.-F. van der Wekken A.J. Early ctDNA dynamics and efficacy of lorlatinib in patients with treatment-naïve, advanced, ALK-positive non-small cell lung cancer J Thorac Oncol 18 2023 1568 1580 37295609
28 Rolfo C. Mack P.C. Scagliotti G.V. Liquid biopsy for advanced non-small cell lung cancer (NSCLC): a statement paper from the IASLC J Thorac Oncol 13 2018 1248 1268 29885479
29 National Comprehensive Cancer Network (NCCN) Clinical Practice Guidelines in Oncology: Non-small Cell Lung Cancer. Version 2.2024 2024 National Comprehensive Cancer Network [cited February 18 2024]
30 Solomon B.J. Baur T.M. Mok T. Efficacy and safety of first line lorlatinib versus crizotinib in patients with advanced, ALK-positive non-small-cell-lung cancer: updated analysis of data from phase 3, randomised, open label, CROWN study Lancet Respir Med 11 2023 354 366 36535300
31 Bearz A. Martini J.F. Jassem J. Efficacy of lorlatinib in treatment-naive patients with ALK-positive advanced NSCLC in relation to EML4::ALK variant type and ALK with or without TP53 mutations J Thorac Oncol 18 2023 1581 1593 37541389
32 Nicoś M. Krawczyk P.J.C. Genetic clonality as the hallmark driving evolution of non-small cell lung cancer Cancers (Basel) 14 2022 1813 35406585
33 Shaw A.T. Friboulet L. Leshchiner I. Resensitization to crizotinib by the lorlatinib ALK resistance mutation L1198F N Engl J Med 374 2016 54 61 26698910
34 Eisenhaur E. Therasse P. Bogaerts J. New response evaluation criteria in solid tumours: revised RECIST guideline, version 1.1 Eur J Cancer 45 2009 228 247 19097774
35 O’Rourke M.B. Januszewski A.S. Sullivan D.R. Optimised plasma sample preparation and LC-MS analysis to support large-scale proteomic analysis of clinical trial specimens: application to the fenofibrate intervention and event lowering in diabetes (FIELD) trial Proteomics Clin Appl 17 2023 2200106
36 Li J. Steffen P. Tse B.C. Deep sequencing of early T stage colorectal cancers reveals disruption of homologous recombination repair in microsatellite stable tumours with high mutational burdens Cancers Basel 14 2022 2933 35740599
37 Harney D.J. Hutchison A.T. Hatchwell L. Proteomic analysis of human plasma during intermittent fasting J Proteome Res 18 2019 2228 2240 30892045
38 Mohindra N.A. Peipert J.D. Blum S.I. Shaw J.W. Penrod J.R. Cella D. General population reference values for the functional assessment of cancer therapy-lung and PROMIS-29 Cancer Med 12 2023 12765 12776 37148552
39 Dagogo-Jack I. Kiedrowski L.A. Heist R.S. Efficacy and tolerability of ALK/MET combinations in patients with ALK-rearranged lung cancer with acquired MET amplification: a retrospective analysis JTO Clin Res Rep 4 2023 100534
40 Felip E. Shaw A.T. Bearz A. Intracranial and extracranial efficacy of lorlatinib in patients with ALK-positive non-small-cell lung cancer previously treated with second-generation ALK TKIs Ann Oncol 32 2021 620 630 33639216
41 Bauer T.M. Felip E. Solomon B.J. Clinical management of adverse events associated with lorlatinib Oncologist 24 2019 1103 1110 30890623
42 Alexander M. Solomon B.J. Burbury K. Thromboembolism in ALK-rearranged on-small cell lung cancer Clin Lung Cancer 19 2018 e71 e72 28818428
43 Farge D. Frere C. Connors J.M. 2019 international clinical practice guidelines for the treatment and prophylaxis of venous thromboembolism in patients with cancer Lancet Oncol 20 2019 e566 e581 31492632
44 Camidge D. Kim H. Ahn M. Brigatinib versus crizotinib in anaplastic lymphoma kinase (ALK) inhibitor-naive advanced ALK-positive non-small cell lung cancer: final results of the Phase 3 ALTA-1L Trial J Thorac Oncol 16 2021 2091 2108 34537440
45 Frost N. Christopoulos P. Kauffmann-Guerrero D. Lorlatinib in pretreated ALK-or ROS1-positive lung cancer and impact of TP53 co-mutations: results from the German early access program Ther Adv Med Oncol 13 2021 1 15
46 Wheler J.J. Janku F. Naing A. TP53 alterations correlate with response to VEGF/VEGFR inhibitors: implications for targeted therapeutics Mol Cancer Ther 15 2016 2475 2485 27466356
47 Lin J.J. Schoenfeld A.J. Zhu V.W. Efficacy of platinum/pemetrexed combination chemotherapy in ALK-positive NSCLC refractory to second-generation ALK inhibitors J Thorac Oncol 15 2020 258 265 31669591
48 Gouda M.A. Buschhorn L. Schneeweiss A. Wahida A. Subbiah V. N-of-1 trials in cancer drug development Cancer Discov 13 2023 1301 3109 37070849
