Can We Extend Beyond 2 Years of Adjuvant ALK TKI Established by ALINA and ELEVATE in Resected Early Stage ALK+ NSCLC a Disease with High Predilection for CNS Metastasis with Novel Clinical Trial Designs?

Abstract

The development and approval of three generations of ALK tyrosine kinase inhibitors (TKIs) since the identification of anaplastic lymphoma kinase fusions (ALK+) in NSCLC has led to improved survival especially with lorlatinib, the only third-generation ALK TKI in the metastatic incurable stage. The establishment of 3 years of adjuvant EGFR TKI in resected stage IB-IIIA EGFR+ NSCLC has led to improvement in overall survival (OS) when compared to placebo. Logically, the use of adjuvant ALK TKI in resected early-stage ALK+ NSCLC has been investigated. But instead of three years of adjuvant TKI, in line with EGFR+ and RET+ NSCLC, the use of ALK TKI was investigated for only 2 years in both ALINA and ELEVATE trials. Unsurprisingly, alectinib (ALINA trial) and ensartinib (ELEVATE trial), respectively, has led to improved disease-free survival (DFS) so far with OS follow up on-going. Nonetheless, this is a disease that requires constant tumor suppression to prevent or delay CNS metastasis, as exemplified by the CNS data from the 7-year CROWN update. In addition, to reviewing the results of ALINA and ELEVATE, we propose two alternate novel clinical designs to build upon the 2-years of adjuvant 2G ALK TKIs.

Introduction

Since the discovery of anaplastic lymphoma kinase ALK fusion (ALK+) in non-small-cell lung cancer (NSCLC) in 2007,^1,2 11 Phase 3 randomized trials in the first-line (1L) setting have led to six ALK-TKIs being approved globally: crizotinib,³ ceritinib,⁴ alectinib^5–7 brigatinib,⁸ ensartinib,⁹ and lorlatinib,¹⁰ and three additional ALK TKIs in China, (iruplinalkib [WX-0593],¹¹ envonalkib,¹² and dirozalkib [XZP-3621]).¹³ These nine ALK TKIs can be divided into three generations: first-generation (1G) (crizotinib) and loraltinib third-generation and the rest second-generation. Recently, the 7-year update of the CROWN trial was reported and the median progression-free survival (PFS) has still not been reached at year 7, with a landmark progression-free survival (PFS) rate of 55% at 7 years for the overall ALK+ NSCLC patients treated with lorlatinib,¹⁴ firmly establishing that lorlatinib is the first-line choice of advanced ALK+ NSCLC.^15–17 Despite this tremendous achievement, metastatic ALK+ NSCLC still remains incurable at this stage despite the extension of PFS to likely >100 months, while unprecedented, is still considered non-curative. In order to offer the potential of a cure, ALK+ NSCLC patients need to be diagnosed at an early stage, followed by surgery and adjuvant treatment if necessary.

For patients with resected early stage (stage IB-IIIA) epidermal growth factor receptor mutated (EGFR+) NSCLC, 3 years of adjuvant osimertinib based on the ADAURA trial led to improved OS over placebo.¹⁸ Similarly, 3 years of adjuvant aumolertinib also improved disease-free survival (DFS) over placebo in stage II–IIIA EGFR+ NSCLC with on-going OS follow-up.¹⁹ Adding to the importance of adjuvant TKI treatment, 3 years of adjuvant RET TKI, selpercatinib, also achieved significant improvement in DFS.²⁰

ALINA and ELEVATE (and ALCHEMIST) Trials

The major difference in the adjuvant ALK TKI trials and the adjuvant EGFR TKI and adjuvant RET TKI trials was the duration of TKI treatment. All the trials including ALINA,²¹ ELEVATE,²² and even the US ALCHEMIST (E4512)²³ which investigated the first-generation crizotinib, only investigated adjuvant ALK TKI for merely 2 years. The patient characteristics and the survival outcome of these three trials are summarized in Table 1. The major difference between ALINA and ELEVATE was that ALINA did not allow adjuvant chemotherapy as it was the control arm while adjuvant chemotherapy was mandatory for higher stage (although patient could decline chemotherapy) and optional for stage IB-IIA per clinical investigators in ELEVATE and adjuvant chemotherapy was stratified for. In addition, the staging system used in ALINA was the seventh edition while ELEVATE used the eighth edition. One difference in the staging system was the “upstaging” of T3N2 disease from stage IIIA to IIIB. However, the most significant difference in the staging system is that the eighth edition considered T1-2N3 also as IIIB while creating the IIIC (T3-4N3) category. Hence, the eighth edition stage IIIB could be either N2 or N3 disease. N3 disease (contralateral or supra-clavicular lymph node metastases) is considered typically unresectable disease regardless of the T categories. Hence, when describing IIIB disease, it was important for trials that used the AJCC eighth staging system (ELEVATE and US ALCHEMIST (E4512)) to specify that only N2 disease was enrolled.

Table 1.

Patient Characteristics, Trial Conduct, and Efficacy Comparison Among E4512, ALINA, and ELEVATE Trials

	E4512	ALINA	ELEVATE
N (Investigational arm: control)	166 enrolled; 153 centrally confirmed (by FISH) ALK+ (75 crizotinib:78 observation)	257 (130 alectinib: 127 chemotherapy)	274 (137 ensartinib:137 placebo)
Trial design	Open label, initial placebo-controlled modified to observation in control arm	Open label	Placebo-controlled randomized trial
Eligible stage	IIA–IIIB	IB (≥4 cm) –IIIA	IB–IIIB
Treatment	Crizotinib 250 mg BID × 2 years versus Observation × 2 years	Alectinib 600 mg BID × 24 months versus Platinum-based chemotherapy q21 days × 4 cycles	Ensartinib 225 mg QD × 24 months versus placebo × 24 months
Randomization	1:1	1:1	1:1
Primary endpoint	Investigator-assessed Disease-Free Survival (in centrally ALK+ cases)	Investigator-assessed Disease-Free Survival in stage II–IIIA	Investigator-assessed Disease-Free survival (DFS) in stage II–IIIB
Stratification factors	N1 vs N2 Male vs Female Radiation (Yes/No)	Stage IB (>4 cm) vsII vs IIIA Asian vs non-Asian	Stage IB vs II vs III Adjuvant chemotherapy (Yes/No)
Adjuvant chemotherapy	Allowed, not required	No, part of randomization strategy	Mandatory for stage IIB–IIIB Optional for stage IB–IIIA
Peri-operative radiation	Allowed	Not allowed	Pre- and post-radiation not allowed
Methods of detecting ALK fusion	FISH, IHC, NGS	“FDA approved test”	Ventana D5F3 IHC
Surveillance interval	CT chest every 6 months	Every 12 weeks × first 2 years; then every 24 months years 3–5	Every 12 weeks × first 2 years; then every 24 months years 3–5
MRI brain interval	Not specified and not required	Every 12 weeks × first 2 years; then every 24 months years 3–5	Every 48 weeks
Crossover allowed	No	No	No
Conduct of the trial
Enrollment period	August 2014 to May 2024 (9 years and 9 months)	August 2018 to December 2021 (3 years and 4 months)	Not specified from the presentation
Region where trial conducted	USA	Global	China
Median age (years)	62	54	54–55
Median follo- up time (months)	58.3	27.8	24
Stage (%)	Stage II: 36.4 Stage III: 44.2 Unknown: 19.5	Stage IB: 10.1. Stage II: 35.8 Stage IIIA: 54.1	Stage I: 25.1 Stage II: 34.0 Stage III: 40.9
Never smoker (%)	58.8	59.9	81.8
Female (%)	64.7	52.1	63.9
Race	Caucasian (79.1%) Black (7.8%) Asian (5.2%)	Asian (55.6%) Caucasian (41.6%)	Asian/Chinese (100%)
Prior chemotherapy (%)	90.9% received adjuvant chemotherapy; 87% in crizotinib arm 95% in observation arm	0 (not allowed by protocol)	69.7% (total patient) 68.6% in ensartinib arm, 70.8% in placebo arm
Prior radiation	37% (crizotinib) 31 (observation)	None	None
Results
Median DFS (primary endpoint)	Crizotinib: 74.6 months (95%CI: 71.2–NE) Observation: 75.1 months (95%CI: 64.7–NE) HR:1.08 (95%CI: 0.61–1.90); P=0.80	(Stage II–IIIA) Alectinib: NE Chemotherapy: 44.4 (95%CI: 27.8–NE) HR: 0.24 (95%CI: 0.13–0.45) P<0.001	(Stage II–IIIB) Ensartinib: NE (95%CI: NE–NE) Placebo: 24.8 (95%CI: 22.4–NE) HR: 0.20 (95%CI: 0.1–0.38); P=0.001
Median DFS (ITT population)	Same as above	(Stage IB–IIIA) Alectinib: Chemotherapy: 41.3 (28.5–NE) 0.24 (95%CI: 0.13–0.43) P<0.001	(Stage IB–IIIB) Ensartinib: NE (95%CI: NE–NE) Placebo: 24.8 (94%: 22.2–NE) 0.20 (95%CI: 0.10–0.37) P<0.0001
HR of median DFS by stage	Stage II–IIIA (N1): HR: 0·80 (0.38–1.71) Stage IIIA (N2)–IIIB: HR: 1·43 (0.75–2.72)	Stage IB: 0.21 (0.02–1.84) Stage II: 0.24 (0.09–0.65) Stage IIIA: 0.25 (0.12–0.53)	Stage IB: HR <0.0001 (95%CI: <0.0001-NE) Stage II: HR: 0.35 (95%CI: 0.14–0.83) Stage III: HR: 0.13 (95%CI: 0.05–0.33)
HR of CNS DFS		0.22 (95%CI: 0.08–0.58)	0.22 (95%CI: 0.08– 0.60)
Median OS	HR: 0.49 (95%CI: 0.17–1.43); P=0.19	Immature at time of data-cutoff date, not reported	Immature at time of data-cutoff date, not reported
≥G3 TRAE	58% (crizotinib)	29.7% (alectinib), 30.8% (chemotherapy)	27.0% (ensartinib); 6.6% (placebo)
Dose interruptions	58% (crizotinib)	27.3% (alectinib); 18.3% (chemotherapy)	35.0% (ensartinib); 14.6% (placebo)
Dose reductions	22% (crizotinib)	25.8% (alectinib); 10% (chemotherapy)	29.9% (ensartinib); 1.5% (placebo)
Dose discontinuations	Not reported	5.5% (alectinib); 12.5% (chemotherapy)	2.2% (ensartinib); 1.5% (placebo)
Median Relative Dose Intensity (RDI)	Not reported	Alectinib: 99.4% Chemotherapy: 100%	Not reported

Abbreviations: TEAE, treatment emergent adverse events; BID, twice a day, CI, confidence interval; CNS, central nervous system; DFS, disease-free survival; FISH, fluorescence in situ hybridization; G3, grade 3; HR, hazard ratio; NE, not evaluable; QD, once a day; TRAE, treatment-related adverse events.

Both ALINA and ELEVATE achieved their primary endpoints in the designated population (stage II–IIIA/B) and also the overall intention-to-treat (ITT) population (with the addition of stage IB population).

No statistically significant overall survival (OS) benefit was observed in ALCHEMIST (HR: 0.49; P=0.26) although the data is still immature. At the time of data cutoff, OS data from ALINA and ELEVATE remained immature and were not reported. However, the ELEVATE trial demonstrated that 87.3% in the ensartinib group versus 57.2% in the placebo group achieved 24-month disease-free survival (hazard ratio [HR]: 0.20, P<0.001).²² Moreover, disease recurrence was observed in only 8.8% in the ensartinib group versus 33.6% in placebo, with a reduced risk of CNS disease recurrence and death. With the potency of next-generation ALK TKIs available at the time of progression, OS benefit may never be achieved with the still rather limited number of patients enrolled.

Unanswered Questions

Duration of Adjuvant ALK-TKI

ALCHEMIST, ELEVATE, and ALINA used a 2-year adjuvant treatment window and 2-year primary DFS landmark. This is 1 year shorter than adjuvant EGFR TKI or RET TKI. Already there is data showing after the end of 3 years of adjuvant osimertinib, relapses increased from years 4 to 5 compared to years 1–3.²⁴ ALINA has reported the site of relapse from both alectinib-treated and placebo-treated arms. As expected, distant metastases were much higher in the placebo arm than the alectinib arm (17.3% vs 2.3%). Brain metastases occurred in four patients in the alectinib and 14 patients in the placebo arm.²¹ Similarly, brain was the most common site of recurrence (5 of 137 patients [3.6%] in the ensartinib group and 17 of 137 patients [12.4%] in the placebo group in the ELEVATE trial.²² The HR for CNS progression was 0.22 (95%CI: 0.08–0.58) and 0.22 (95%CI: 0.08–0.60) for ALINA and ELEVATE respectively. Correspondingly, from the 7-year CROWN update, the patients with baseline brain metastasis had stable PFS rate from year 5 to year 7 at 53%,¹⁰ indicating that in order to stabilize advanced ALK+ NSCLC, long-term suppression may be necessary even for microscopic disease, just as in the adjuvant hormonal treatment of estrogen receptor positive breast cancer demonstrating 10 years of adjuvant treatment to be superior to 5 years of adjuvant treatment.²⁵

ctDNA as a Guidance for Length of ALK TKI Treatment?

Translational data from ALINA or ELEVATE which may help increase our knowledge of the role of circulating tumor DNA (ctDNA), has not been published at this time. From the ctDNA subset analysis of ADAURA outcome,²⁶ the one conclusion that could be drawn is the fact that when ctDNA was positive or when minimum residual disease (MRD) was positive prior to adjuvant osimertinib, even if MRD became negative during osimertinib, the chance of relapse was high after stopping osimertinib. Thus, for these patients with MRD positive disease pre-adjuvant osimertinib, the period of osimertinib should be extended. For patients with negative ctDNA or MRD negative prior to osimertinib, there were only occasional relapses after completion of adjuvant osimertinib. Therefore, use of MRD as guidance on the length of adjuvant TKI is hypothesis generating at best.

Role of Adjuvant Chemotherapy

Four cycles of cisplatin-based chemotherapy especially cisplatin/vinorelbine has demonstrated to achieve OS over observation, conducted in an era when knowledge of actionable genomic alterations (AGA+) NSCLC was nonexistent. Given the toxicity of cisplatin-based adjuvant chemotherapy with at least 30–35 mg/m² per week dose density and the delay in starting adjuvant TKI of at least 4 months (2–4 weeks to recover from surgery and 12 weeks of cisplatin-based chemotherapy and 3 weeks to recover from toxicities from chemotherapy). Cisplatin-based chemotherapy is unlikely to have a “CNS protective effect” like next-generation ALK TKIs. The predilection for CNS metastasis is a continuous process essentially beginning when the disease is clinically diagnosed.

From ALINA, 88/96 (91.7%) patients and 19/23 (82.6%) patients completed four cycles of platinum/pemetrexed and platinum/vinorelbine respectively. Thirteen out of one hundred and twenty (10.8%) patients who started cisplatin had to switch to carboplatin chemotherapy and chemotherapy discontinuation rate was 12.5% due to adverse events.²¹ Correspondingly, 57/89 (64.0%) patients and 58/87 (66.7%) patients completed platinum/pemetrexed in the ensartinib and placebo arm respectively.²² Analyzing these chemotherapy usage data in ELEVATE indicated the full benefit (four cycles) of adjuvant chemotherapy was hard to deliver. Furthermore, given the use of adjuvant chemotherapy is stage dependent and then optional in lower stage, the current breakdown of HRs by adjuvant chemotherapy or not is impossible to analyze until follow-up data of HRs based on stage and adjuvant chemotherapy.

Biologically for ALK+ NSCLC which has the highest among receptor tyrosine kinase (RTK) fusion positive NSCLC, leaving the brain “unprotected” for ~6 months is risky in our opinion given the incidence of brain metastasis is unrelentingly increasing over time,²⁷ although prospective data supporting harm of this 6-month period of “lapse of coverage” is lacking.

In addition, none of the adjuvant TKI trials including EGFR+, ALK+, and RET+ NSCLC to date stratified for receipt of adjuvant chemotherapy or not independent of stage, likely because it was assumed that adjuvant chemotherapy will be given in the vast majority of the patients. Nonetheless, in this new era of targeted therapy, adjuvant chemotherapy may do more harm than good. We have previously argued that adjuvant chemotherapy did not improve DFS from the ADAURA trial.²⁸ Indeed the majority of the ELEVATE patients (~90%) received adjuvant chemotherapy. However, ALINA trial showed that the DFS benefit with adjuvant alectinib did not require adjuvant chemotherapy and is considered the standard of care globally.

2G vs 3G ALK TKI as Adjuvant Therapy? Can We Translate from 7-Year Follow-Up of CROWN

The recent 7-year CROWN follow-up data indicated that the median PFS of lorlartinb still has not been reached and is likely to be projected to be >100 months. As a close approximate, when median PFS will be reached would be around the time PFS rate of 50% is reached. Importantly, in the 7-year follow-up of CROWN, patients with baseline brain metastasis, the 5-year to 7-year PFS rate has stayed stable at 53%. Furthermore, the 7-year PFS rate of lorlatinib from CROWN has surpassed the 7-year OS rate of alectinib from the ALEX trial.²⁹

Potential Designs of Next-Generation Adjuvant ALK TKI Trials to Address Unanswered Questions. Two Proposals

As depicted in Figure 1 in Design A we propose a simple head-to-head randomized, Phase III trial directing comparing 2 years of adjuvant alectinib or ensartinib to 3 years of adjuvant lorlatinib. Two years of adjuvant alectinib (global)/ensartinib (currently in China and likely global approval soon) will serve as the control arm. Three years of adjuvant lorlatinib will be the investigational arm to harmonize with the length of other adjuvant TKI practice. We proposed surveillance period to be every 3 months that include CT chest/abdomen/pelvis and MRI of brain. While some may argue this design will give lorlatinib an unfair advantage given the longer duration, the design is not to compare 2 years to 3 years or alectinib to lorlatinib individually bur as one treatment regimen against the other. Although with proposed surveillance schedule, there is a possibility statistically significant difference in DFS could result during the first 2 years. Further details are described in the figure legend.

Figure 1.

Design A. Hypothetical phase 3 trial to build upon ELEVATE and ALINA. This design is a head-to-head trial of adjuvant alectinib/ensartinib × 2 years (standard of care) versus 3 years of adjuvant lorlatinib. We did not propose the design to specifically compare alectinib or lorlatinib or between 2 years and 3 years of adjuvant ALK TKI. The design is to compare one treatment regimen versus another treatment regimen. The disadvantage of this hypothetical design is that sponsor will have to supply the drugs in the control arm and the cost of 2 years of alectinib/ensartinib will be expensive. Furthermore, even if this trial is positive, it will only switch SOC from 2 years of adjuvant ALK TKI to 3 years of adjuvant ALK TKI which is another important step in the adjuvant ALK TKI usage.

The second clinical trial design (Design B) as depicted in Figure 2 try to address the optimal duration of adjuvant ALK TKI and consists of two parts. This design borrowed the design of adjuvant hormonal treatment in estrogen receptor positive (ER+) breast cancer is intended to test the hypothesis of extending the standard of care. Part 1 for patients who had completed 2 years of adjuvant ALK TKI (alectinib or ensartinib) which is standard of care (we anticipate ensartinib will receive adjuvant indication given the publication of ELEVATE in New England Journal of Medicine) and without documented disease progression, they are randomized in a 1:1 ratio to receive three years of lorlatinib or placebo. Patients who were randomized to receive placebo will be able to receive lorlatinib at the time of progression. Surveillance scans will be every 3 months that include CT chest/abdomen/pelvis and MRI of the brain.

Figure 2.

Design B is an alternate hypothetical phase 3 design to build upon ELEVATE and ALINA. The design contains two parts. Part 1. This part of the design if positive could potentially extend the adjuvant ALK TKI period from 2 years to 5 years. Eligible patients who had completed 2 years of adjuvant ALK TKI (either alectinib or ensartinib once it is approved for adjuvant indication in multiple regions) and no documented progression will be eligible to be randomized in a 1:1 ratio to 3 years of lorlatinib or 3 years of placebo. Patients in the placebo arm can cross over to received lorlatinib if there is documented confirmed disease progression. The primary endpoints for both designs will be DFS as assessed by investigators. Major secondary endpoints will include OS and toxicities. Part 2. This part of the trial design seeks to identify the optimal duration of adjuvant ALK TKI. The part 2 of Design B will depend on whether part 1 is positive. Part 2 is a randomized discontinuation design. For patients who completed additional 3 years of adjuvant lorlatinib without documented progression, they are randomized to continue lorlatinib until disease progression or discontinue lorlatinib at progression. The primary endpoint will be DFS2. Surveillance period will be every 6 months.

If part 1 of the trial is positive so we anticipate will be within 2 years due to high CNS trophism of ALK+ NSCLC then part 2 will proceed. For patients enrolled onto the additional 3 years of adjuvant lorlatinib and without documented disease progression they will be randomized to continue lorlatinib indefinitely or observation in a randomized discontinuation design. The primary endpoint is investigator-assessed DFS. This trial is more complicated, but it may provide the optimal duration of adjuvant ALK TKI treatment. Finally, Design B may also be more economic as the sponsor will not have to provide 2 years alectinib and only 3 years of placebo.

Conclusion

ALINA and ELEVATE trials provide the evidence that 2 years of adjuvant ALK TKI improved DFS. The adjuvant treatment was only investigated for 2 years which may be insufficient. We propose two next-generation adjuvant ALK TKI trial designs to address many of the questions raised in this editorial. While Pfizer is unlikely to sponsor such as trial at this late life cycle of lorlatinib, we hope these designs, specially Design B may be able to be implemented as investigator-initiated trials sponsored by cooperative groups or international oncology societies.

Disclosure

Dr Misako Nagasaka reports Consulting fees from Caris Life Sciences; Honoraria from AstraZeneca, Genentech, Boehringer Ingelheim, BMS, Regeneron, Bayer, Eli Lilly and Company, Johnson and Johnson, Takeda, Daiichi Sankyo, Pfizer; Meeting/travel support from Nuvation Bio, outside the submitted work. Dr Sai-Hong Ou reports Consulting fees from Abbvie; Honoraria from LIANGYIHUI Healthcare, MedJ; Data safety/advisory board participation from Abbvie; Stock or stock options from MBrace Therapeutics, BlossomHill Therapeutics, Nuvation Bio, Nuvalent, outside the submitted work. The authors report no conflicts of interest in this work.