First-in-human, open-label dose-escalation and dose-expansion study of the safety, pharmacokinetics, and antitumor effects of an oral ALK inhibitor ASP3026 in patients with advanced solid tumors

In recent years, aberrant expression of anaplastic lymphoma kinase (ALK) tyrosine kinase receptor has emerged as a relevant biomarker and therapeutic target for a number of solid tumors [1, 2]. Different types of alterations in the ALK gene have been implicated in human cancer tumorigenesis, and different tumor types have different structural alterations in the ALK gene [3, 4]. In non-small cell lung cancer (NSCLC), echinoderm microtubule-associated protein-like 4 (EML-4) and kinesin family member 5B (KIF5B) account for the majority of ALK gene rearrangement lung cancer [5]. The presence of ALK gene rearrangement defines ~3–13 % of NSCLC [6‐10] that are highly sensitive to the first-generation ALK tyrosine kinase inhibitor (TKI), crizotinib. Crizotinib (250 mg twice daily) has high efficacy in patients with NSCLC harboring this oncogenic kinase, with overall response rates of >50 % [11‐13]. Crizotinib has been established as the standard first-line treatment for patients with advanced ALK-positive NSCLC [13]. The current National Comprehensive Cancer Network (NCCN) guideline recommend crizotinib use in patients with advanced NSCLC harboring ALK gene rearrangement [14]. However, almost all patients develop resistance, typically within 10 months [11, 13, 15, 16].

The most common molecular mechanisms of resistance include amplification of the ALK fusion gene, development of resistance mutations, and activation of alternative or bypass signaling pathways or progression in the CNS [17]. One strategy to overcome crizotinib resistance is to develop potent small molecule TKIs for ALK-rearranged genes and/or specifically target the common resistant mutations, such as gatekeeper mutation L1196M [17].

ASP3026 is a selective, ATP-competitive, second-generation ALK TKI that was identified through a medicinal chemistry campaign designed to obtain compounds with a better pharmacologic profile compared with crizotinib. The kinase selectivity of ASP3026 was evaluated and compared with that of crizotinib against a panel of 86 tyrosine kinases [18]. ASP3026 at 1 μmol/L inhibited 11 tyrosine kinases by more than 50 %, with the highest selectivity for ALK, ROS1, and ACK kinases, showing that the kinase selectivity of ASP3026 differed from crizotinib. ASP3026 was more selective for FRK, YES, ACK, TNK1, and EGFR (L858R), whereas crizotinib had higher selectivity for MET, RON, LCK, JAK2, MUSK, TRKs, TYRO3, AXL, MER, and EPHs [18]. ASP3026 fits within the ATP-binding pocket of both wild-type and L1196M ALK kinase domains and inhibits their kinase activities with IC₅₀ values of 10 and 32 nmol/L, respectively. By contrast, crizotinib fits within the ATP-binding pocket of wild-type ALK kinase domain but not the L1196M ALK kinase domain. Thus, crizotinib displays tenfold weaker activity for the mutated EML4-ALK compared with the wild-type ALK gene [18]. In mice bearing subcutaneous and intracranial xenograft tumors, ASP3026 has potent antitumor activity against both wild-type ALK and EML4-ALK L1196M xenograft tumors compared with crizotinib [18]. ASP3026 also has a higher tissue-to-plasma ratio compared with crizotinib, which could translate into a wide therapeutic margin between efficacious and toxic doses [18]. Preclinical data indicated that ASP3026 may have potential therapeutic effects for patients with crizotinib-resistant ALK-positive NSCLC and potentially for patients with other cancer types of ALK-driven tumors.

Gain-of-function ALK gene alterations have been detected in several types of solid tumors, B cell lymphomas and pediatric tumors [2, 3, 22], for which crizotinib has either established or promising clinical efficacy [17]. Several second- and third-generation ALK inhibitors have also shown either established or promising clinical activity for ALK-positive NSCLC patients who progressed on crizotinib, supporting the development of potent ALK inhibitors as an effective strategy to overcome resistance to crizotinib [17].

We conducted this first-in-human trial to determine the safety, pharmacokinetic, and antitumor effects of a novel second-generation ALK inhibitor ASP3026. Overall, ASP3026 was well tolerated with no treatment-related deaths. The AE profile in the dose-expansion cohort was similar to that reported for crizotinib [11], with gastrointestinal complaints (nausea and vomiting) being some of the most frequently reported. The constellation of AEs in this small sample set was also similar to other agents in this class, such as ceritinib and alectinib, with gastrointestinal AEs commonly reported [16, 23, 24].

ASP3026 had linear PK parameters and demonstrated dose proportionality over the dose range of 25–800 mg once daily. ASP3026 had good oral absorption, with a t _max of approximately 3 h; median half-life was 35 h (range, 22–85 h), confirming that once-daily dosing was suitable.

Of the 16 crizotinib-resistant ALK+ subjects (15 NSCLC and one neuroblastoma) who received 525 mg ASP3026, eight (50 %) achieved PR and seven (44 %) achieved SD at 8 weeks. When restricted to ALK-positive NSCLC patients, the PR and SD rates were 8/15 (53 %) and 6/15 (40 %), respectively. Although caution is warranted due to a lack of head-to-head comparison, this tumor response rate is comparable with those observed in other second-generation ALK inhibitors, such as ceritinib and alectinib, in crizotinib-resistant NSCLC patients [16, 24].

Our study has several other limitations. First, although our study was designed to allow the enrollment of patients with ALK-driven advanced tumors other than NSCLC, only one patient with advanced neuroblastoma, who had a commonly detected ALK F1174L mutation [19‐21] and who progressed on prior crizotinib, was identified during the enrollment period. This is because routine molecular testing for other types of tumors did not, and has not, become standard clinical practice. Secondly, multiplexed genomic testing, such as a targeted resistance mutation panel of the ALK kinase domain and targeted next-generation sequencing, was not required at study entry for determining molecular mechanisms of resistance at disease progression to crizotinib. This is unlikely to affect our result. Although different resistance mutations may confer variable responses to subsequent ALK inhibitor therapy [15, 17, 25], most second-generation ALK inhibitors, such as ceritinib and alectinib, as well as ASP3026, have strong efficacy against both secondary mutations in the ALK tyrosine kinase domain and wild-type ALK gene amplification [16, 24]. Nevertheless, with the increasing use of clinical molecular profiling tests at treatment resistance in patients with advanced malignancies, individualized treatment beyond a second-generation ALK inhibitor should be based on the assessment of molecular mechanism of resistance.