New standards of care for ALK-positive disease

The first-generation ALK inhibitor  crizotinib is the current standard option  for patients with newly diagnosed,  advanced ALK-positive NSCLC.  However, patients invariably relapse on  crizotinib treatment, with the central  nervous system (CNS) being one of the  most common and challenging sites of  relapse. The second-generation ALK  inhibitor alectinib is more potent than  crizotinib [1, 2] and shows clinical  activity in crizotinib-resistant NSCLC  [3-6]. Notably, trial data have indicated  significant CNS activity. Alectinib has  become a standard therapy for patients  with crizotinib pre-treated ALK-positive  NSCLC, but research efforts are ongoing  to establish it as a first-line option.

Superiority of alectinib in ALEX

The ALEX trial evaluated alectinib  600 mg twice daily compared to  crizotinib 250 mg twice daily in  untreated patients with advanced or  metastatic ALK-positive NSCLC [7].  ALEX was the first global randomised  phase III study to compare a next- generation ALK TKI with a first- generation ALK TKI in the first-line  setting. In all, 303 patients were  randomised in a 1:1 fashion across 98  sites in 29 countries. Patients with  asymptomatic, treated or untreated  brain metastases were also enrolled.  Approximately 40 % of individuals in  each arm harboured CNS metastases,  which were untreated in 60 % of cases.  Investigator-assessed PFS was defined  as the primary endpoint.

The study met its primary objective.  At the time of the analysis, median PFS  had not been reached in the alectinib- treated arm, while it was 11.1 months in  the crizotinib arm (Figure 1). This  translated into a risk reduction of 53 % (HR, 0.47; p < 0.0001). PFS according to  the Independent Review Committee,  which was a secondary endpoint, was  also significantly longer with alectinib  (25.7 vs. 10.4 months; HR, 0.50; p  < 0.0001). Objective responses occurred  in 83 % versus 76 % of each arm,  although this difference did not reach  statistical significance (p = 0.09).  However, duration of response was  significantly longer with alectinib (not  reached vs. 11.1 months; HR, 0.36).  Median OS had not been reached in  either treatment arm.

Alectinib showed a more favourable  AE profile than crizotinib, with lower  rates of nausea, diarrhoea, vomiting,  peripheral oedema, dysgeusia,  transaminase elevations, and visual  impairment. The AE profile of alectinib  included increased bilirubin levels,  myalgia, anaemia, and weight. Dose  reductions and treatment discon- tinuations were less frequent in the  experimental arm, and the duration of  treatment was longer with alectinib than  with crizotinib.

Findings on intracranial activity

Almost all of the subgroups derived  greater PFS benefit from alectinib than  from crizotinib. This implies that  patients both with and without brain  metastases fared better with the new  ALK TKI. For patients with CNS  metastases at baseline, median PFS was  not reached vs. 7.4 months (HR, 0.40),  and for those without CNS metastases at  baseline, it was not reached vs. 14.8  months (HR, 0.51).

Time to CNS progression in the total  population represented a key secondary  endpoint. According to a competing risk  analysis with CNS progression, non- CNS progression and death as  competing events, the risk of having  CNS progression as the first event was  reduced by as much as 84 % in the  alectinib-treated arm (cause-specific  HR, 0.16). At 12 months, only 9.4 % of  alectinib-treated patients showed CNS  progression, whereas this was 41.4 % in  the crizotinib arm. The CNS ORR, as  opposed to the overall ORR,  demonstrated significant benefit of  alectinib therapy. For patients with  measurable lesions at baseline,  response rates were 81 % vs. 50 % with alectinib and crizotinib, and for those with measurable and non-measurable  CNS lesions, 59 % vs. 26 %. Complete  remissions occurred in 38 % vs. 5 % for  measurable, and 45 % vs. 9 % for non- measurable CNS lesions. The median  duration of response in the brain was  17.3 vs. 5.5 months and not reached vs.  3.7 months, respectively.

As the authors summarise, the large  magnitude of benefit observed with  alectinib suggests that first-line alectinib  will be superior to the sequential  treatment of crizotinib followed by  alectinib. Overall, these results establish  alectinib as the new standard of care for  patients with previously untreated,  advanced, ALK-positive NSCLC.

Third-generation agent lorlatinib

Secondary mutations in the ALK  domain can induce resistance to first- generation and second-generation ALK  TKIs, which calls for yet other treatment  options. The potent third-generation  TKI lorlatinib is a selective inhibitor of  ALK and ROS1, with broad-spectrum  effects against most known ALK  resistance mutations, including G1202R  [8, 9]. Also, lorlatinib can cross the  blood–brain barrier to achieve clinically  meaningful CNS activity. Indeed, a  phase I trial showed intracranial efficacy  for lorlatinib, which included deep  responses in patients with measurable  disease [10].

At present, lorlatinib 100 mg/d is  being evaluated in an ongoing phase I/  II study in 220 patients with ALK-positive NSCLC and 40 patients with  ROS1-positive disease. In the phase II  portion of the trial, patients with ALK  rearrangement were divided into five  expansion cohorts according to their  pre-treatment (Figure 2). One cohort  was treatment-naïve (EXP1), while groups EXP2 to EXP5 had received prior crizotinib only (EXP2), prior crizotinib  plus chemotherapy or one other ALK  TKI with or without chemotherapy  (EXP3), two prior ALK TKIs with or  without chemotherapy (EXP4), or three  prior ALK TKIs with or without  chemotherapy (EXP5). Patients with  ROS1 rearrangement are receiving  lorlatinib as any line (EXP6).

The data presented at the ASCO  Congress related to groups EXP2 to  EXP5; i.e., the ALK-positive cohort that  had been treated with at least one ALK  TKI prior to study entry [11]. Together,  EXP2 and EXP3 included 80 patients,  while EXP4 consisted of 70 patients, and  EXP5 of 40 patients. Brain metastases  were present in 55 % to 71 % of cases  across these cohorts at the time of study  entry. The primary endpoint was ORR/  intracranial ORR, according to the  Independent Review Committee.

Meaningful and durable  responses in heavily  pre-treated patients

In the total cohort, ORR was 32.9 %.  Complete responses occurred in 1.2 %,  partial responses in 31.7 %, and disease  stabilisation in 32.9 %. At week 12, the  DCR was 56.1 %. For cohorts EXP2,  EXP3, EXP4 and EXP5, ORRs were  57.1 %, 44.4 %, 25.0 % and 30.8 %,  respectively. The majority of patients  experienced decrease in target lesion  size. There was one complete remission  in the EXP4 cohort.

In addition, lorlatinib therapy evoked  robust and clinically meaningful  intracranial activity, which included  complete intracranial responses,  irrespective of prior lines of therapy.  Target lesions plus non-target lesions  together showed ORR of 48.1 %; for target  lesions only, this was 51.4 %. Complete  responses occurred in 26.9 % and 20.0 % in these two groups. Disease control was  75.0 % at 12 weeks for patients with target  lesions plus non-target lesions. With  regard to the systemic and intracranial  activities, the responses proved durable.  At data cut-off, the longest duration of  treatment was more than 300 days, and  the longest duration of intracranial  response was 7 months.

In the entire group incorporating  cohorts EXP1 to EXP6, the safety  analysis identified hyperlipidaemia as  the most common AE, although this was  successfully managed with lipid- lowering agents. Cognitive effects  occurred in 19.0 % (across all grades)  and were generally mild and rapidly  reversible upon dose modification.  Dose delays became necessary in  29.3 %, and dose reductions in 19.8 %.  Only 3.4 % of patients discontinued  treatment due to AEs.  Based on these data, lorlatinib has  received Breakthrough Therapy  designation from the US Food and Drug  Administration for use in patients with  ALK-positive metastatic NSCLC  previously treated with at least one ALK  TKI. The phase III CROWN study, which  is comparing first-line lorlatinib to  crizotinib, is presently recruiting  patients.

Analysis of EML4-ALK variants

Ou et al. investigated the association of  ALK resistance mutations with specific  variants of the EML4-ALK rearrangement  [12]. Samples from 634 patients with  ALK-positive NSCLC collected in the  FoundationCORE database were  analysed. The most common variants  were EML4-ALK v1 and EML4-ALK  v3a/b, each of which was found in 32 %  of cases. EML4-ALK v2 occurred in 8 %,  other EML4-ALK variants in 12 %, and  non-EML4-ALK rearrangements in 16 %.  The presence of known ALK resistance  mutations was significantly associated  with v3 as compared to v1 (p = 0.0002).  G1202R was the most frequent ALK  resistance mutation in this dataset. This  mutation also showed significant  association with v3 compared to all  non-v3 variants (p = 0.0004). Drop-out,  switching, and evolution of multiple ALK  resistance mutations occurred over the  course of sequential ALK inhibitor  treatment.

The authors concluded that the use of  tissue-based and blood-based next  generation sequencing allows for  detection of specific ALKfusion variants  and increases the understanding of the  biology of ALK-positive NSCLC. In  addition, it might have value to predict  potential mechanisms of resistance and  inform the selection of ALK inhibitor  therapy. Non-ALK mechanisms of  acquired resistance should be  considered, especially in tumours with  ALKrearrangements that are non- variant 3. For instance, MET kinase  domain duplication was identified as a  novel mechanism of acquired resistance  after crizotinib and ceritinib treatment in  a patient harbouring EML4-ALK v1.

References:

1. Sakamoto et al., CH5424802, a selective ALK  inhibitor capable of blocking the resistant  gatekeeper mutant. Cancer Cell 2011; 19: 679-  690
2. Kodama et al., Selective ALK inhibitor  alectinib with potent antitumor activity in models  of crizotinib resistance. Cancer Lett 2014; 351:  215-221
3. Ou et al., Alectinib in Crizotinib-Refractory  ALK-Rearranged Non-Small-Cell Lung Cancer:  A Phase II Global Study. J Clin Oncol 2016; 34:  661-668
4. Shaw et al., Alectinib in ALK-positive,  crizotinib-resistant, non-small-cell lung cancer: a  single-group, multicentre, phase 2 trial. Lancet  Oncol 2016; 17: 234-242
5. Yang JC et al., Pooled efficacy and safety  data from two phase II studies (NP28673 and  NP28761) of alectinib in ALK+ non-small-cell  lung cancer (NSCLC). WCLC 2016, abstract  P3.02a-016  REFERENCES
6. Gadgeel et al., Pooled Analysis of CNS  response to alectinib in two studies of  pretreated patients with ALK-positive non-small- cell lung cancer. J Clin Oncol 2016; 34: 4079-  4085
7. Shaw AT et al., Alectinib vs. crizotinib in  treatment-naïve advanced ALK+ NSCLC:  primary results of the global phase III ALEX  study. ASCO 2017, abstract LBA9008
8. Gainor JF et al., Molecular Mechanisms of  Resistance to First- and Second-Generation  ALK Inhibitors in ALK-Rearranged Lung Cancer.  Cancer Discov 2016; 6: 1118-1133
9. Johnson TW et al., Discovery of (10R)-7-  amino-12-fluoro-2,10,16-trimethyl-15-oxo-  10,15,16,17-tetrahydro-2H-8,4-(metheno)  pyrazolo[4,3-h][2,5,11]-benzoxadiazacyclotetra- decine-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: 4720-4744
10. Solomon BJ et al., Safety and efficacy of  lorlatinib (PF-06463922) from the dose- escalation component of a study in patients with  advanced ALK+ or ROS1+ non-small cell lung  cancer (NSCLC). J Clin Oncol 2016; 34 (suppl;  abstr 9009)
11. Shaw AT et al., Efficacy and safety of  lorlatinib in patients (pts) with ALK+ non-small  cell lung cancer (NSCLC) with one or more prior  ALK tyrosine kinase inhibitor (TKI): A phase I/II  study. ASCO 2017, abstract 9006
12. Ou S-H I et al., Association of ALK  resistance mutations by EML4-ALK variant (v3  vs. non-v3) in ALK+ non-small cell lung cancer  (NSCLC). ASCO 2017, abstract 9010