The majority of lung cancers are nonâ€“small cell lung cancers (NSCLCs), which are associated with several gene mutations.
Archana Chidambaram, MD, PhD
Roswell Park Cancer Institute
Elm & Carlton Streets
Alex A. Adjei, MD, PhD
Roswell Park Cancer Institute
Elm & Carlton Streets
The majority of lung cancers are nonsmall cell lung cancers (NSCLCs), which are associated with several gene mutations. Targeted therapies have shown great promise compared with cytotoxic chemotherapy in patients with NSCLC whose tumor cells harbor a gene mutation. The echinoderm microtubule associated protein like 4 (EML4) gene was found to be fused with the anaplastic lymphoma kinase (ALK) gene, resulting in a mutant fusion gene product,EML4-ALK, which drives tumorigenesis in a subset of patients with NSCLCs. In less than a decade since this discovery, several drugs have been designed to target this mutant gene. Crizotinib was the first drug approved for use in patients with ALK-rearranged NSCLC; however, resistance patterns have begun to emerge within a year of use of this drug. Newer, more potent ALK inhibitors include ceritinib and alectinib, which have been approved for use in patients who have progressed on crizotinib in the United States and in Japan, respectively. This review article focuses on these next-generation ALK inhibitors, specifically with respect to their pharmacokinetics, spectrum of activity, and ongoing clinical trials. This article also reviews the available literature on other next- generation ALK inhibitors, which are currently in various phases of clinical development.
Lung cancer remains the leading cause of cancer-related mortality, accounting for 26.8% of all cancer deaths. An estimated 158,040 Americans will die from lung cancer in 2015.1This is higher than the mortality attributed to breast (6.8%), colorectal (8.4%), and prostate cancers (4.7%) combined. Therefore, tremendous efforts are being made to discover improved treatment options for patients diagnosed with lung cancer. More than 80% of lung cancers are nonsmall cell lung cancers (NSCLCs). In these tumors, molecular profiling has revealed molecular aberrations that can be specifically targeted to yield significantly improved outcomes.
The success of targeting the BCR-ABL gene in patients with chronic myelogenous leukemia (CML) led to a search for similarly targetable mutations in NSCLCs. Although gefitinib was approved by the US Food and Drug Administration (FDA) as a third-line therapy in patients with NSCLC, the responses to treatment have not been consistent. Retrospective studies of phase II trials demonstrated that responses typically were observed mostly in patients who were female, never-smokers, and in those who had bronchoalveolar carcinoma or adenocarcinoma with bronchoalveolar features.2However, even among these patients, the response was not consistent.
The work of Lynch et al3revealed for the first time the specific activating mutations within the tyrosine kinase domain of epidermal growth factor receptor (EGFR), which correlated with exquisite sensitivity to gefitinib. As a result of this study, patients could now be screened for the presence or absence of these mutations, and a better response to therapy could be predicted. It has been established that in an unselected population, the response rate to EGFR mutation-targeted therapy is roughly 9%,4while among patients with NSCLC who are known to have EGFR mutations this response rate is about 67%.5
Theanaplastic lymphoma kinase(ALK) gene was originally identified in 1994 in anaplastic large-cell lymphoma where the mutation was a translocation from chromosome 2 on to chromosome 5, resulting in a fusion protein to nucleophosmin (NPM).6In 2007, Soda et al7discovered that in lung adenocarcinomas the ALK gene was rearranged and fused to the echinoderm microtubule associated protein like 4 (EML4) promoter. Both theALKandEML4genes are located on the short arm of chromosome 2 and are transcribed in opposite directions (FIGURE 1). However, in a subset of individuals with NSCLC, there is an inversion mutation resulting in the fusion of the N-terminal domain of EML4 to the intracellular kinase domain of ALK.7This creates theEML4-ALKfusion gene, which, when transcribed, results in a constitutively active ALK tyrosine kinase.
The intracellular signaling pathways (FIGURE 1) usually involve the Ras/Raf/MEK/ERK (resulting in uncontrolled cell proliferation) or the JAK-STAT3 or phosphoinositide 3 (PI3)-kinase signaling mechanisms (resulting in cell survival and cytoskeletal changes).9
This chromosomal rearrangement acting as an oncogene was among the first to be discovered in solid tumors and for that reason was considered a fascinating target, dispelling the widely held notion that these rearrangements were restricted to hematologic malignancies. Soda et al10established transgenic mice models that expressedEML4-ALKin the lung alveolar epithelial cells and showed that these mice developed several adenocarcinoma nodules in both lungs. Subsequently, the study researchers also showed that injection of NIH 3T3 fibroblasts expressing theEML4-ALKgene product into athymic mice resulted in tumorigenesis and fatal respiratory failure, which could be prevented by treating these mice with ALK inhibitors. These experiments confirmed that the fusion kinase could serve as a driver mutation when present.10As a result of these findings, the search for ALK inhibitors was initiated with the hope of improving outcomes for patients withALK-mutant NSCLCs.
AE indicates adverse events; ALT, alanine transaminase; AST, aspartate aminotransferase; ORR, objective response rate; OS, overall survival; PFS, progression-free survival
Crizotinib, or PF-02341066, is an orally bioavailable, small-molecule tyrosine kinase inhibitor (TKI) that competes for the adenosine 5’-triphosphate (ATP)-binding sites on tyrosine kinases, which was found to have potent in vitro and in vivo inhibitory activity against c-MET and ALK.11It has a half maximal inhibitory concentration (IC50) of about 0.003 uM.12In a comprehensive kinase assay including more than 120 kinases, crizotinib was found to have a nearly 20-fold higher selectivity for ALK and MET compared with the other kinases.13Cell-based assays by Shaw et al14additionally showed that ROS1 autophosphorylation was also inhibited by crizotinib with a IC50 of 40 nM to 60 nM.
The first phase I study (PROFILE 1001) from August 2008 was conducted by Camidge et al15; 143 patients with ALK-positive advanced NSCLC were assessed and about 60.8% were found to have an objective re- sponse. Median progression-free survival (PFS) was found to be 9.7 months, with the overall survival (OS) estimated at 87.9% at 6 and 12 months. Another phase I trial to study the toxicity profile and efficacy of crizotinib in patients with ALK-rearranged lung cancers was initiated in February 2010.16This study also found an impressive response rate, with about 57% of the patients demonstrating a partial response (PR) or complete response (CR) and 33% patients demonstrating stable disease, with a favorable adverse event (AE) profile.
PROFILE 1005, a phase II trial, enrolled 261 patients with ALK-positive disease who had been treated with at least one line of chemotherapy before enrollment. Even patients with stable brain metastases were deemed eligible for participation in the study. Updated results so far show that treatment with crizotinib leads to a response rate of 60% and a median PFS of 8 months.17
A phase III study (PROFILE 1007) was conducted to compare the responses seen in patients with ALK-rearranged NSCLC treated with crizotinib versus systemic chemotherapy using either pemetrexed or docetaxel.18The results were unequivocally in favor of targeted therapy, with a response rate of about 65% with crizotinib versus 20% with chemotherapy. Moreover, the median PFS was 7.7 months in the crizotinib arm versus 3.0 months in patients receiving chemotherapy. Although no significant improvement was found in OS, the conclusion was that crizotinib was clearly associated with improved response and tolerability compared with standard chemotherapy in patients with ALK-positive NSCLC.
PROFILE 1014,19another multicenter trial, is ongoing to determine the efficacy and safety of crizotinib versus pemetrexed-cisplatin or pemetrexed-carboplatin in patients with previously untreated ALK-positive advanced lung cancers. This phase III study has so far demonstrated significant improvement in PFS (10.9 months vs 7.0 months) and an overall objective response rate (ORR) of 74% in the groups treated with crizotinib versus 45% in those treated with systemic chemotherapy. Median OS had not been reached in either group. An updated report by Mok et al20had shown improvement in the rate of intracranial events with respect to time to intracranial progression in patients who receive crizotinib versus systemic chemotherapy. However, there have been other isolated case reports of patients on crizotinib21,22who have a systemic response but are found to have central nervous system (CNS) disease that either develops anew or progresses.
These findings suggest that crizotinib may decrease the CNS failure rates compared with systemic chemo- therapy, but patients may also develop CNS progression. Therefore, further studies may be warranted to better identify the differences between the two groups. The initial enthusiasm caused by the dramatic responses to crizotinib therapy has since been tempered by the rising incidence of acquired resistance patterns. Studies have shown that the average time of relapse is 8.9 months to 10.5 months.23,24The known mechanisms of resistance may be divided into two categoriesALK dominant and ALK nondominant.25
As shown inFIGURE 2, the former category accounts for about one-third of the cases with acquired resistance to crizotinib. This category includes secondary gene mutations in theALKkinase domain, ALK or copy number gain or escape of the tumor cells to so-called sanctuary sites such as the CNS where crizotinib has poor penetration. The most common and well characterized among the gene mutations is the substitution of methionine for leucine, or theL1196Mmutation. TheL1196of ALK is located at the bottom of the ATP-binding pocket and considered the gatekeeper position. The presence of an amino acid with a bulky side chain at this crucial position, it is postulated, may render the hydrophobic pocket inaccessible, thereby interfering with the binding of crizotinib.26,27Several other mutations have been characterized, includingG1269A,1151Tins,G1202R,S1260Y,C1156Y,L1152R,F1174C, andD1203N.28
The ALK-independent mechanisms involve activation of other signaling pathways such as the EGFR, heat shock protein 90 (HSP 90), the PI3K/ AKT/mTOR pathways27,29or transformation to sarco- matoid carcinoma.30,31In the remaining third of patients, the mechanism by which tumor cells acquire resistance to crizotinib is unknown.
The poor prognostic significance of ALK, the logical ease with which it can be targeted, and the impressive results seen with crizotinib have all contributed to the frenzied search for other inhibitors with an increased spectrum of activity in terms of potency, bioavailability, and durability of response.Ceritinib (LDK378) is a selective small-molecule ALK inhibitor developed by Zykadia. Similar to crizotinib, ceritinib also competes for the ATP-binding site but has a much higher potency, with an IC50 of 0.00015 uM.11Crizotinib owes its existence to Marsilje et al at Novartis,32whose previous work had shown the high potency of a compound TAE684 in mediating ALK inhibition.33However, TAE684 itself could not be applied to clinical practice because of the toxic metabolites it generates. Hence, as shown inFIGURE 3, it was modified to yield LDK378 or compound 6, as it was then known. In Ba/F3 cells transfected with various kinases, this compound was found to inhibit ALK activation with an IC50 value of 40.7 nM but had IC50 values >100 nM against all other kinases.The time taken to reach maximal concentration after an orally ingested dose was roughly 6 hours. The mean terminal half-life was estimated at 40 hours, suggesting a drug-accumulation effect with increasing doses. The drug was found to be metabolized by cytochrome p450 3A (CYP3A), with potential drug interactions with other drugs that are similarly metabolized.34The drug is primarily eliminated by the liver but requires dose reduction only in moderate or severe hepatic impairment.In vitro cell-based studies35using tumor cell lines that harbored theL1196MandG1269Amutations revealed that ceritinib was able to suppress ALK phosphorylation and its downstream signaling pathways at much lower doses than crizotinib. TheL1196M-mutant cell line MGH045 was implanted in mice, who were treated with vehicle, crizotinib, and ceritinib at 25 mg/kg/day. The group treated with ceritinib demonstrated significantly lower tumor volumes, underlining the efficacy of ceritinib in controlling tumor growth. Ceritinib was also found to have activity against cells expressing other ALK kinase domain mutations, such asG1269A,S1206Y, andI1171T. However, it lacked the capacity to inhibit ALK containing theG1202RandF1174Cmutations.ASCEND 1 was the initial phase I trial conducted by Shaw et al.36Of the 130 patients with ALK-rearrangementpositive disease enrolled by October 2012, 122 had advanced NSCLC and of these, 83 patients had received crizotinib before enrollment in the trial. The remaining eight were non-lung cancer patients, including those with breast cancer, alveolar rhabdomyosarcoma, rectal adenocarcinoma, anaplastic large-cell lymphoma, and inflammatory myofibroblastic tumor. These patients received LDK378 at doses starting from 50 mg/day to 750 mg/day. The primary objective was to determine the maximal tolerated dose (MTD), with secondary objectives of determining the safety and side-effect profile, pharmacokinetic profile, and antitumor activity.
Patients who received a dose of at least 400 mg/day were found to have an increase in alanine aminotransferase (ALT) that was dose limiting. Patients receiving doses of 600 mg daily or more developed diarrhea and dehydration, while those receiving 750 mg/ day developed vomiting and nausea more frequently. Overall, the most common AEs were gastrointestinal complaints (82% with nausea, 75% with diarrhea, and 65% with vomiting). Fatigue was listed as a frequent side effect, seen in about 47% of the study population, and an increase in ALT was seen in 35%. Four cases of interstitial lung disease were attributed to the drug, as was one case of asymptomatic QT prolongation. More than half the patients required a dose reduction during the course of treatment.
The early results from this phase I study also revealed an ORR of about 58% for patients who received at least 400 mg/day. The ORR was estimated at 59% for those who received 750 mg/day. The median duration of response (DOR) was 8.2 months and median PFS was estimated at 7.0 months. Among patients who had previously been treated with crizotinib, the ORR was 56% at both dose levels, while the ORR was about 62% for patients who were crizotinib-naiÌˆvereceiving the 400-mg day dosage. The median PFS in these patients was about 6.9 and 10.2 months, respectively.
A subgroup analysis37was later carried out to evaluate the differences in tolerability and efficacy of ceritinib among the Asian patient population versus the Caucasian group. Interestingly, it was found that fewer Asians developed grade 3 or 4 side effects (55% vs 72%, respectively). Moreover, the Asian subgroup also had a higher ORR than the Caucasian group (69% vs 57%, respectively) and a higher me- dian DOR (10.1 months vs 6.9 months, respectively).
Based on these early results, the FDA granted accelerated approval in April 2014, for the use of ceritinib in patients with ALK-rearranged metastatic NSCLC who had progressed through treatment with crizotinib. This was noteworthy because it is one of the few times a drug has received FDA approval based on phase I results alone.38However, this achievement entails some potential pitfalls. The MTD was set at 750 mg/day; however, more than 50% of the patients enrolled in the trial required dose reductions at lower doses of 400 mg and 600 mg. Hence, this suggested that a lower dose that is therapeutically effective and better tolerated may need to be determined. The terminal half-life of 40 hours also argues in favor of dose adjustments, with possibly considering an initial loading dose followed by a maintenance dose that may improve the tolerability profile. Although the study did include patients with stable asymptomatic intracranial metastases, the pharmacokinetic profile did not include any information about the level of the drug in the cerebrospinal fluid (CSF). Also, it remains unknown if the bioavailability of the drug is affected by the intake of food. Finally, because the delayed AE/ toxicity profile remains to be uncovered, more studies examining these factors must be conducted before the drug can be used more widely.
Results from the expansion phase of the ASCEND-I trial continue to be updated. At the 2014 ASCO Annual Meeting, recent findings were presented based on 255 patients,39,40246 of whom had ALK-rearranged NSCLC (including both the group of patients previously treated with an ALK inhibitor, such as crizotinib, and the group of patients who were ALK inhibitor-naiÌˆve). The median ORR for all ALK+ NSCLC was estimated at 61.8%,40the median DOR was 9.7 months, and the median PFS was 9 months. In general, the ALK inhibitor-naiÌˆve group fared significantly better than the group that was previously treated with ALK inhibition (ORR = 72.3% vs 56.4%, respectively; DOR = 17 months vs 8.3 months, respectively; and median PFS of 18.4 months vs 6.9 months, respectively).
Of the 124 patients with stable brain metastases in this cohort,3926 patients were ALK-inhibitor-naiÌˆve, while the remaining 98 had received an ALK inhibitor previously. The ORR in these two groups was 69.2% and 50%, respectively. The median DOR was not estimable in the former group, while it was 6.93 months in the latter group. The median PFS in these two groups was 8.3 months and 6.7 months, respectively, while the intracranial response rates were again significantly higher in the group that was ALK inhibitor-naiÌˆve (75%) versus the group that had received prior ALK inhibition (40%).
Thus, the overall results were highly promising for therapeutic efficacy with continued ALK inhibition in the setting of crizotinib resistance, unlike the case of other TKIs such as EGFR inhibitors where the response rate was around 10% once resistance developed to the first-line TKI.
Several trials are evaluating the use of ceritinib in various clinical contexts.30,41The ongoing phase II trial NCT01685138 is a single-arm, multicenter, open-label trial evaluating the efficacy of ceritinib in patients who were either treatment-naiÌˆve (crizotinib or chemotherapy) or had been previously treated with cytotoxic chemotherapy. The initial results from this study were presented at ASCO 2015, and showed an impressive ORR in patients with and without brain metastases (67.6% and 58%, respectively). The median DOR and PFS have been estimated at 10.8 and 11.1, respectively, for patients without brain metastases, while these values are 9.1 months and 10.8 months, respectively, for patients with brain metastases.42Another multicenter, single-arm phase II trial (NCT01685060) seeks to determine the ORR with ceritinib in patients who have previously received chemotherapy and crizotinib. The updated results were presented at ASCO 2015 and showed impressive intracranial disease control rates (IDCR) in patients that matched or surpassed the whole body (WB) response. The most common AEs seen were nausea, diarrhea, and vomiting. The researchers conclude that ceritinib can provide sustained responses in patients treated with crizotinib with or without intracranial metastases.43
ALT indicats alanine aminotranferase; AST, aspartate aminotransferase; CNS, central nervous system; NSCLC, nonsmall cell lung cancer; TRAE, treatment-related AE.
Two phase III trials are evaluating the efficacy of ceritinib in terms of PFS compared with standard chemotherapy agents. Trial NCT01828099 includes patients who are treatment-naiÌˆve and compares ceritinib (arm 1) against combinations of pemetrexed and cisplatin (arm 2) versus pemetrexed and carboplatin (arm 3). This is an interesting study because it may provide information about the utility of ceritinib as a first-line agent in ALK-rearranged NSCLC. The other phase III trial (NCT01828112) includes patients who were previously treated (chemotherapy with the platinum doublet or crizotinib) and has three groups: group 1 receives ceritinib, group 2 receives single-agent pemetrexed, and group 3 receives docetaxel.In an analysis by Friboulet et al,3511 patients developed disease relapse after initially responding to ceritinib. These relapsed tumors were biopsied and 5 of the 11 showed new mutations in the ALK kinase domain at either G1202 or F1174 loci. There remains a paucity of data identifying other resistance mechanisms in these relapsed cases.Alectinib, CH5424802, manufactured by Chugai Pharmaceuticals/Roche Pharmaceuticals, is another orally available, small-molecule inhibitor of ALK with an IC50 of 0.0019 uM.12The half-life (t1/2) was found to be about 8.6 hours and its oral bioavailability in mice was estimated at 70.8%.ALK-rearranged lung cancer cells that have the following mutations conferring resistance to crizotinib are still sensitive to alectinib:L1196M,G1296A,C1156Y,F1174L42,43. In fact, kinase activity assays by Sakomoto et al44showed that this drug mediated substantially potent inhibition of both native ALK and theL1196Mmutant ALK (Ki = 0.83 nM and 1.56 nM, respectively). In contrast, crizotinib had a 10-fold higher Ki value for the mutant ALK compared with wild-type ALK.ALK G1202Rmutation has been reported in a case of acquired resistance to alectinib.45Interestingly, this mutation also confers resistance to crizotinib and ceritinib.A phase I/II study (AF-001JP) was conducted in Japan46that included patients with ALK-rearranged NSCLC from across 13 institutions. The phase I portion of this study included 24 patients, and the primary objective was to determine the dose-limiting toxicities and ascertain the MTD for the phase II portion of the study. Patients on this study received doses ranging from 20 mg twice daily to 300 mg twice daily. No dose-limiting toxicities were noted at the maximal dose; hence, 300 mg twice daily was set as the recommended dose for the subsequent part of the study. The phase II portion of the study recruited 46 patients of whom 2 had CR and 41 had PR. The ORR was an impressive 93.5%. Twelve of the 46 patients had grade 3 treatment-related AEs (neutropenia and increased creatine phosphokinase [CPK] levels) while 5 of the 46 developed serious AEs. However, no grade 4 AEs or deaths were noted. The drug was approved for use in Japan in 2014 for patients with ALK-positive NSCLC that is advanced, recurrent, or unresectable.
ALEX-1, a phase I/II trial, was conducted in the United States by Gadgeel et al47to ascertain the MTD of alectinib because the Japanese study showed no AEs at the maximal studied dose of 300 mg twice daily. In the US study, 47 patients received increasing doses of alectinib, from 300 mg twice daily to 900 mg twice daily. The most common AEs were fatigue, myalgias, and peripheral edema. Dose-limiting toxicities were seen in two patients receiving the 900-mg twice-daily dosing and included grade 3 headache and grade 3 neutropenia. At the data cutoff point at a median of 126 days follow-up, about 44 patients were evaluable and the ORR was 55%, which was dramatically lower than the response seen with the Japanese study. More impressively, however, 21 of the 47 patients were diagnosed with stable CNS metastases at baseline, and of these, 6 had a CR following treatment with alectinib, 5 had a PR, and 8 had stable disease. The ratio of alectinib concentration in the CSF to serum was 0.75, suggesting a very high level of CNS penetration. Based on this study, the dose of alectinib for the phase II portion (ongoing) was recommended at 600 mg twice daily. At the 2015 ASCO Annual Meeting, data presented from the phase II portion of the study showed a particularly impressive CNS response to alectinib.48
Subsequently, some reports have described the use of alectinib on a compassionate basis for leptomeningeal disease that developed in patients treated with crizotinib and ceritinib.44Of the four patients who were treated with 600 mg twice daily of alectinib, three developed significant clinical and radiologic improvement, while the fourth patient had stable disease for 4 months prior to progressing.However, patients treated with these next-generation ALK inhibitors also develop resistance to the drugs. With alectinib, Katayama et al50found a novelV1180Lgatekeeper mutation and another novelI1171Tmutation.Interestingly, these muta- tions were associated with continued sensitivity to ceritinib and other next-generation ALK inhibitors (TABLE 3) but were resistant to crizotinib. Studies have also shown that activating the MET signal pathway predisposes to resistance to alectinib. Kogita et al demonstrated that the MET ligand hepatocyte growth factor (HGF) enables the H3122 and H2228 ALK-positive NSCLC cell lines to develop resistance to alectinib; however, the cells still retain their sensitivity to crizotinib.51Over the past decade, significant progress has been made in targeted therapy in NSCLC, especially with the use of ALK inhibitors for ALK-rearranged NSCLC. While these have followed the patterns seen with the use of other TKIs, such as those that inhibitEGFRandKRAS, the dramatic initial response has been followed by emergence of resistance patterns within a year. There have been heartening differences with significant responses seen with the next-generation ALK inhibitors, in contrast to the second-line of therapy inEGFRand other mutant pathways. However, this also raises interesting questions about the most optimal sequence in which to employ this arsenal of inhibitor drugs.
Multiple treatment algorithms have been proposed to answer these questions, some of which are being attempted in clinical trials. One school of thought suggests using crizotinib up front for ALK- rearranged NSCLC. On disease relapse, rebiopsy of the tumor specimen and use of ceritinib (if sensitive) is suggested. Given the PFS with crizotinib of 8.2 months and median PFS of crizotinib-exposed patients on ceritinib at 7.8 months, the combination is expected to yield a median PFS of 17.4 months and a median OS of 49.4 months.53
Alternatively, if the rebiopsy reveals a mutant-cell population that would be resistant to ceritinib (containing theC1156Ymutation, for example) or if CNS relapse was the more notable feature, alectinib may be a more reasonable second-line therapy. In case of a third case of crizotinib resistance, activation of alternate signaling pathways such as EGFR-, insulin-like growth factor 1 receptor (IGF-1R), or KRAS is noted. In these cases, instead of using another ALK-TKI, it may be more prudent to use an anti-EGFR/ anti-IGF-1R drug, or systemic chemotherapy. Other non-TKI drugs are also being investigated, including HSP 90 inhibitors and immunotherapeutic agents.
It is important to determine what the consequences are of using these TKIs in any given order. It has been reported that the incidence of leptomeningeal metastases is increasing among patients with ALK-rearranged NSCLC. Given the rising use of crizotinib and even ceritinib in these patients, could there be a link between these events? If so, how would this influence our choice of initial and subsequent lines of therapy for ALK-positive tumors?
Consequently, much potential exists for studies using different combinations to produce the best outcome.