The battle against ALK resistance: successes and setbacks

Abstract

Chromosomal translocations and single point mutations involving the Anaplastic Lymphoma Kinase (ALK) gene have been described in several human tumors, including anaplastic large cell lymphoma (ALCL), non-small cell lung cancer (NCSLC), inflammatory myofibroblastic tumor (IMT) and neuroblastoma. Cancer cells are “addicted” to ALK constitutive activation and are highly sensitive to the treatment with small-molecule inhibitors. Crizotinib, an oral ALK inhibitor, has proved to provide dramatic clinical benefit in patients with NSCLC harboring ALK rearrangements. Nonetheless, acquired drug resistance inevitably develops and leads to tumor progression and relapse. Different mechanisms of crizotinib acquired drug resistance, resembling those reported for other tyrosine kinase targeted therapies, have been recently reported both in cell lines and in patients. Thus, the identification of the molecular mechanisms of crizotinib resistance will be strictly required in order to pursue the appropriate therapeutic options for patients with ALK-rearranged tumors.

Keywords::

• ALK resistance
• ALK-rearranged tumors
• crizotinib
• targeted therapy

1. Introduction

The deregulated activation of kinases is recognized as one of the major tumorigenic events in human cancer. Remarkably, the majority of kinases related to human cancer are so called “driver oncogenes” since cancer cells depend on their constitutive activation for proliferation and survival [1]. Notably, constitutively active receptor tyrosine kinases (RTKs) and non-receptor tyrosine kinases are excellent targets for the development of specific inhibitors of the kinase activity and for the design of modern “patient specific” and/or “tumor restricted” therapies [2]. In fact, many clinical trials reported the efficacy of tyrosine kinase targeted therapies in a subset of molecularly defined cancer patients, such as imatinib in Bcr-Abl expressing Chronic Myelogenous Leukemia (CML), gefitinib/erlotinib in EGFR mutated Non Small Cell Lung Cancer (NSCLC), sorafenib or vemurafenib in B-Raf mutated melanomas and more recently crizotinib in anaplastic lymphoma kinase (ALK)-translocated NSCLC patients [3,4]. Based on these therapeutic options, the molecular screening is more and more mandatory to enroll patients into the appropriate targeted therapeutic regimen based on the genetic profile.

In NSCLC, the most recently discovered “druggable” kinase is ALK, originally described in Anaplastic Large Cell Lymphoma (ALCL) [5]. ALK is a receptor tyrosine kinase that is activated by fusions or point mutations, or overexpressed in a variety of human cancer, including ALCL, NSCLC, inflammatory myofibroblastic tumor (IMT) and neuroblastoma. Many partner genes have been described to fuse to the ALK gene in different tumors (such as NPM in ALCL, EML4 in NSCLC and TPM3 or TPM4 in IMT). ALK fusion partners mediate a ligand-independent oligomerization of ALK that results in the constitutive activation of the kinase domain. Constitutive activation of ALK drives tumorigenesis in vitro and in vivo [5]. Remarkably, ALK expressing tumors are sensitive to the treatment with small molecule inhibitors. Among these, crizotinib is a potent ATP-competitive inhibitor of c-Met and ALK, which was developed before the identification of EML4-ALK translocation in NSCLC. In Phase-I clinical trial in ALK positive advanced NSCLC patients, crizotinib showed a remarkable tumor regression in nearly 60% of patients [6]. In addition, crizotinib showed promising results for the treatment of IMT and ALCL patients and preliminary studies have demonstrated that mutated ALK is a suitable molecular target in neuroblastoma [7,8]. Nevertheless, despite the high rate of response, some NSCLC patients did not show any clinical benefits in terms of OS or PFS [9]. Also, initially responder patients inevitably became refractory to crizotinib, with pattern of development of resistance similar to that described for other TK inhibitors. As reported for other RTKs, mechanisms of escape to crizotinib have been already described in ALK positive NSCLC patients that develop progressive disease despite continuous treatment (Table 1). In general, three common mechanisms of acquired drug resistance to RTK inhibition have been described: i) secondary aminoacid mutations in the RTK-drug interaction domains; ii) gene amplification or overexpression of the target RTK; iii) activation by various genetic mechanisms of alternative compensatory signaling pathways, including other RTKs or downstream molecules. In EGFR mutated NSCLC or Bcr-Abl expressing CML, typically drug resistance is due to second point mutations in the gene that encode for the drug target (T790M for EGFR and T315I for Bcr-Abl, respectively). These mutations commonly occur at the gatekeeper site within the kinase domain and alter the affinity of binding to the target of the small molecule leading to cancer cell insensitivity. Recently, some groups reported the finding of L1196M single point mutation in patients with ALK positive lung cancer who developed acquired resistance to crizotinib. Aminoacid L1196 corresponds to the “gatekeeper” site in ALK and mutations at this site alter drug binding and decrease the efficacy of the inhibitor [10]. Other mutations in resistant ALK positive NSCLC patients have been described (Table 1), but the mechanisms by which they induce drug resistance needs to be further clarified. Nonetheless, it seems that these ALK mutations vary for their resistance to crizotinib, or for their sensitivity to second-generation ALK inhibitors (such as CH5425802, ASP3026, and AP26113), likely because they differently modify the kinase domain of ALK (Table 1) [11-14]. Interestingly, it has been recently reported that in neuroblastoma the two recurrent gain-of-fuction mutations, R1275Q and F1174L, display a different sensitivity to crizotinib in vitro and in xenograft models, the first being totally responder to crizotinib, the latter being completely resistant [15]. Intriguingly, a crizotinib-sensitive patient with IMT harboring a RANBP2-ALK fusion developed a secondary mutation at the F1174 site [16]. Thus, F1174L can function as activating mutation and as well as a TKI-resistant mutation. The resistance of F1174L was overcome in vitro by increasing the dose of crizotinib. In addition to point mutations, other mechanisms for resistance to ALK inhibition have been described, such as ALK amplification or the activation of compensatory pathways, i.e., EGFR and c-KIT signaling [11,17,18]. A gain in ALK gene copy number has been implicated in crizotinib resistance and was found in some crizotinib resistant patients independently of the mutational status of ALK [17,18]. Increased EGFR phosphorylation and activation was found in patients both before crizotinib treatment and at the time of disease progression and was principally caused by EGFR activating mutations. However, in some samples the combination of crizotinb treatment and gefitinib or erlotinib failed to induce apoptosis of resistant cells [11,13]. Thus, other genetic lesions might be considered or might coexist in the same tumor cells or in the same tumor sample. Gene amplification of c-KIT in crizotinib resistant NSCLC occurs in a similar fashion to Met amplifications found in gefitinib resistant NSCLC. Lastly, there is evidence that the presence of stromal cells secreting SCF is fundamental to confer a fully resistant phenotype to crizotinib that is completely reversed by treatment with imatinib (a small molecule inhibitor of c-KIT), suggesting that a combination of crizotinib with imatinib might overcome this type of resistance [11].

Table 1. ALK crizotinib resistance.

Mutation	ALK fusion	Sensitivity to ALK inhibitors	Disease	Author	Ref.
L1152R	EML4-ALK E6:A20	ND	NSCLC	Sasaki T, et al. 2011	[13]
C1156Y	EML4-ALK E13:A20	ND	NSCLC	Choi YL et al. 2010 Katayama R et al. 2011	[10,12]
F1174L	RANBP2-ALK	High dosage Crizotinib	IMT Neuroblastoma	Bresler SC et al. 2011 Sasaki T et al. 2010	[15,16]
L1196M	EML4-ALK E13:A20	AP26113, CH5424802	NSCLC	Katayama R et al. 2011 Sakamoto H et al. 2011	[12,14]
G1202R	ND	CH5424802, ASP3026	NSCLC	Katayama R et al. 2012	[11]
S1206Y	ND	CH5424802, ASP3026	NSCLC	Katayama R et al. 2012	[11]
G1269A	EML4-ALK	ND	NSCLC	Doebele et al. 2012	[17]
F1174C	EML4-ALK	ND	NSCLC	Doebele et al. 2012	[18]
D1203N	EML4-ALK	ND	NSCLC	Doebele et al. 2012	[18]
1151Tins	ND	CH5424802	NSCLC	Katayama R et al. 2012	[11]

ND: Not determined.

2. Expert opinion

The existence of multiple different resistance mutations or different mechanisms of acquired resistance in ALK patients has important clinical implications. First, direct sequencing of multiple exons or a multiplexed assay will be necessary to look for different specific resistance mutations. Second, the concomitant presence of different mutations in the same patient has to be considered, increasing the complexity and the difficulty of detecting each mutation. Third, the identification of alternative compensatory pathways might be difficult and expensive in each tumor specimen. Nevertheless, the identification of additional mutations or amplifications of other RTKs in crizotinb resistant patients will be strictly required in order to find the appropriate therapeutic options for these patients.

The genetic heterogeneity of tumors implies that newly activated pathways or different mechanisms of resistance may coexist within the same tumor, and that targeting one single molecule might be ineffective. Indeed, recent studies reported that the variable responses of NSCLC patients to ALK inhibitors might largely depend on the activation of alternative pathways in which wild type or mutated RTKs can compensate for the loss of ALK signaling, and support cell growth and survival. As a result, an emerging paradigm is the potential use of combinations of different agents, administered together or in alternating usage, to overcome intrinsic or acquired drug resistance and to optimize the cure of ALK positive patients. These options might include higher doses of crizotinib, new more potent ALK inhibitors or the combination with other RTK inhibitors, such as EGFR, c-KIT or c-Met inhibitors. In addition to ALK inhibition by small molecules, the inhibitor of the heat shock protein-90 (IPI-504), a known ALK interactor, has shown a therapeutic effect in terms of tumor regression and stabilization of the disease in mice and in ALK+ NSCLC patients, respectively [19-21].

Finally, common molecules downstream of several RTKs, such as PI3K/MEK [19], could be associated as broader therapeutic targets than single ALK inhibitors. Therefore, multi-target drugs with a broad spectrum of action, which would not result in unacceptable toxicities, might be more appropriate than highly restricted inhibitors to improve the therapeutic efficacy in patients with ALK positive malignancies.

Acknowledgements

The work was supported by grants from the Associazione Italiana per la Ricerca sul cancro (AIRC), from the International Association for Cancer Research (AICR) and grant FP7 ERC-2009- StG (Proposal No. 242965 – “Lunely”) to R.C.

Declaration of interest

The authors state no conflict of interest and have received no payment in preparation of this manuscript.