Over the past 30 years, from the discovery of a fusion rearrangement to the recent promise of LOXO-292, RET targeting has gone from distant concept toward clinical reality.
Alexander Drilon, MD
Research on the proto-oncogene RET has accelerated exponentially since 1985, when Masahide Takahashi, MD, PhD, and colleagues first reported a fusion rearrangement in lymphoma DNA.1 Early progress was slow. Other investigators noted RET abnormalities—both fusions and mutations—in papillary thyroid cancer (PTC), in multiple endocrine neoplasia type 2 (MEN2), and after radiation exposure, but targeted interventions were still on the distant horizon.2-4
In 2005, the structure of the RET kinase domain was described, and in subsequent years, as small-molecule inhibition gained traction and RET fusions were discovered in patients with lung cancer, research on RET targeting built momentum.5 Initial multikinase inhibitor trials suggested that RET could be a viable target, but “off-target” toxicity limited effective drug design.6
In light of these findings, the focus has shifted to the development of RET-specific inhibitors. In September 2018, the leading prospect in the field, LOXO-292, was granted a breakthrough designation by the FDA for RET fusion—positive metastatic non–small cell lung cancer (NSCLC) and RET-mutant medullary thyroid cancer (MTC).7 Five weeks later, the designation was expanded to RET fusion-positive thyroid cancer.8
RET-specific inhibition appears well tolerated and effective regardless of gatekeeper mutations or RET fusion partners, and researchers are optimistic about the value of RET inhibition across tumor types. How this inhibition will be accomplished—by selective agents, custom-tailored multikinase inhibitors, or in combination with other therapies—has yet to be determined.RET is a transmembrane glycoprotein receptor tyrosine kinase (RTK) encoded by RET, which is located on chromosome 10.9 During embryogenesis, RET aids in development of the enteric nervous system and kidneys; throughout life, in the homeostasis of multiple tissues.10,11
Receptor activation is achieved through indirect ligand binding; glial cell line—derived neurotrophic factor (GDNF) family ligands (GFLs) bind to GDNF family receptor-α coreceptors, resulting in a GFL–GFRα complex that homodimerizes RET.12
With homodimerization, intracellular tyrosine residues are trans-autophosphorylated, signaling adaptors are recruited, and several signaling cascades are activated. RET-associated pathways include RAS/MAPK/ERK, PI3K/AKT, JAK/STAT, and C-gamma, all of which regulate cell proliferation (Figure).13—16With knowledge of associations between RET and downstream cell proliferation pathways, it follows logically that cancer might result from RET dysfunction. Such dysfunction can be retraced to various RET aberrations, which are divided into 2 types: chromosomal rearrangements (fusions) or mutations.1
Chromosomal rearrangements, such as the gene fusion originally described by Takahashi, most commonly lead to overexpression of RET or uncontrolled activation of the RET intracellular domain. Either mechanism can increase downstream activity, thereby contributing to malignancy via angiogenesis, cell survival, or invasion.17
Other rearrangements have also been documented, resulting in gain-of-function of RET or inhibition of tumor suppressor gene partners, but such instances are rare.18 In NSCLC, KIF5B—RET is the most common fusion, whereas CCDC6—RET and NCOA4—RET are most common in PTC.18,19 Each of these genes (KIF5B, CCDC6, NCOA4) is located on chromosome 10 with RET, and rearrangements occur via inversion.1 Rarely, interchromosomal RET rearrangements or translocations occur.17 At a population level, RET aberrations have been identified in less than 2% of diverse cancers, according to a molecular analysis of malignancies present in nearly 5000 patients. These include mutations (38.6%), fusions (30.7%), and amplifications (25%).20 RET rearrangements occur in 1% to 2% of patients with NSCLC (usually adenocarcinomas).19 These cases often share characteristics with patients who have ROS1 or ALK rearrangements; namely, that they tend to be nonsmokers less than 60 years of age.21
With a prevalence rate of 10% to 20%, RET rearrangements are more common in PTC compared with NSCLC, and they are still even more prevalent in patients who develop PTC secondary to radiation exposure, such as those involved in the Chernobyl disaster. In this population, approximately three-quarters exhibited RET rearrangements.2,22
In contrast with the rearrangements found in NSCLC and PTC, proto-oncogene activation in MTC is most commonly due to an RET mutation.23 These can be somatic or germline and are often substitution mutations of the intracellular or extracellular domains; in the latter case, cysteine residues may be replaced with other amino acids, ultimately resulting in ligand-independent dimerization and activation.24 RET germline mutations are responsible for MEN2, a syndrome that leads to about 25% of MTC cases.15,25
Specifically, certain mutations can be associated with subsets of this disease, such as substitutions at the V804 codon, which lead to familial MTC.17 In the other 75% of MTC cases, which are sporadic, 43% to 71% have a RET mutation.20 M918T mutations are most common, followed by Cys634 point mutations, deletions, or small insertions.17,26
Somatic RET mutations also occur in about 20% of patients with sporadic pheochromocytoma.27 Recently, both similar and novel mutations have been found in a variety of other cancer types, including hepatoma, gastrointestinal stromal tumor, meningioma, atypical lung carcinoid, paraganglioma, and Merkel cell, endometrial, colorectal, and breast carcinoma.20 This list may continue to expand with increasing usage of next-generation sequencing.1Multikinase Inhibitors
Based on the prevalence of RET aberrations in NSCLC and thyroid cancer, the ever-expanding list of other RET-mutated cancers, and the upstream location of RET itself, RET targeting has potential for pan-cancer application. However, the orientation of RET and its roles in organ development and homeostasis have given rise to safety concerns for inhibition.1
Preclinical Ret-knockout models and diseases with inactivating RET mutations have provided some hints at safety issues, as documented by a plethora of studies, but the relevance of their findings is diminished by recent clinical trials with RET-specific inhibitors. In short, despite the potentially catastrophic consequences of RET inhibition in early development, resulting in tissue agenesis and fatal physiologic derangements, inhibition from late childhood onward appears relatively safe.1
A variety of RET inhibitors have been approved as multikinase agents, in which RET is not the primary target.1 These include sunitinib (Sutent), cabozantinib (Cabometyx), and lenvatinib (Lenvima), with various indications. Although some are labeled for thyroid cancer, none are specifically approved for RET-aberrant subtypes.
Efficacy is comparable across agents, but mechanisms of action vary.28 For example, sunitinib is a type I inhibitor that attaches to the adenosine triphosphate binding pocket of activated RET, cabozantinib is a type II inhibitor, binding the inactive form; others have unknown mechanisms.6,29 Several of these agents have shown activity against RET-rearranged models, resulting in suppression of downstream pathways and decreased cell proliferation.29,30,31 Similarly, treatment of mutant models results in decremental tumor growth, both through RET wild-type targeting and activity against specific mutant isoforms.1
Clinical trials have echoed these findings. For example, in the phase III EXAM trial evaluating cabozantinib, patients with RET M918T—mutated MTC had longer progression-free survival with the drug compared placebo than did those whose RET mutation status was not known.33 Cabozantinib has also shown activity in RET fusion-positive NSCLC, with response rates ranging from 20% to 50% in patients who have received 2 or more lines of therapy.28,34 These are, however, essentially proof-of-concept trials, because results do not match the efficacy of multikinase inhibitors against other major oncogenic drivers such as EGFR or ROS1.28 In part, this is because off-target activity results in dose-limiting toxicities before substantial RET inhibition can occur.1
Other agents are required to achieve significant RET inhibition without toxicity. Combination therapies may be needed, since some studies suggest that RET inhibition is most effective with additional downstream targeting. Similarly, custom multikinase agents that are more selective for RET but still maintain some downstream inhibition could potentially accomplish the same goal. Although the possibilities of multitarget agents remain open, the research spotlight is currently on RET-selective inhibitors. Of these, three leading agents warrant a closer look:The foremost RET-selective inhibitor, LOX-292, was recently granted a breakthrough designation by the FDA for RET fusion-positive NSCLC, RET fusion-positive thyroid cancer, and RET-mutated MTC.7,8 These designations were given following the notable responses among patients treated with LOXO-292 during the phase I/II LIBRETTO-001 trial (NCT03157128). Patients were heavily pretreated, with a median 3 lines of prior therapy.
At the 2018 American Society of Clinical Oncology Annual Meeting, investigtors reported an objective response rate (ORR) of 77% (95% CI, 58%-90%) with LOXO-292 for patients with RET fusion—positive (NSCLC and 45% for patients with RET-mutated MTC. By the April 2018 data cutoff, 82 patients had been treated with LOXO-292 across 7 cohorts.
In the most recently reported data, treatment with LOXO-292 resulted in the following ORRs: RET-mutated MTC, 59%; RET fusion-positive NSCLC, 68%; and RET fusion-positive thyroid cancer, 78%.34,35, 36
Additionally, all patients (n = 4) with intracranial lesions responded to therapy. LOXO-292 appeared safe and well tolerated, with very few patients discontinuing therapy; most adverse events (AEs) were limited to grade 1 and unrelated to treatment. About 5% of patients had grade 3 treatment-related AEs, including elevated liver enzymes, diarrhea, thrombocytopenia, and tumor lysis syndrome, but all resolved with dose interruption.
In an interview with OncologyLive®, lead investigator Alexander E. Drilon, MD, a medical oncologist and clinical director of the Early Drug Development Service at Memorial Sloan Kettering Cancer Center in New York, New York, reflected on the LIBRETTO-001 findings. “Just to put the [results] in perspective: With the older and ‘dirtier’ multikinase inhibitors for RET fusion—positive cancers, we see a 30% response rate with much more toxicity. Here, we are seeing a drug that is achieving outcomes that are similar to what we would expect for targeted therapy in EGFR, ALK, and ROS1,” Drilon said.Results from the phase I ARROW trial (NCT03037385) demonstrated slightly lower response rates than LOXO292.37 During the dose-escalation study, 51 patients were treated. Those with RET-mutated MTC had a best ORR of 40%, and the RET fusion-positive NSCLC group had a best ORR of 50%. Treatment with BLU-667 resulted in 3 dose-limiting toxicities. Grade 3 AEs included hypertension and neutropenia.
Reporting these findings at the American Association for Cancer Research Annual Meeting in April 2018, lead author Vivek Subbiah, MD, of The University of Texas MD Anderson Cancer Center in Houston, said, “ARROW dose escalation data validate BLU-667 as a promising precision therapy for RET-altered cancers.” He added that the responses were noteworthy considering the low doses given to most patients.In 2017, at the European Society for Medical Oncology Annual Congress, investigators reported results from a phase I/IB clinical trial (NCT01877811) in which the RET inhibitor RXDX-105 was given to 152 patients.16 The presenters highlighted a subgroup of 22 patients with RET fusion—positive NSCLC. Six out of 8 patients (75%) with non–KIF5B-RET fusions had partial responses. Of the 14 NSCLC patients with KIF5B-RET fusions, 3 achieved stable disease. Early data also pointed to some potential issues with toxicity. Grade 3 or higher rash occurred in 12% of patients, and nephrotoxicity was documented. Additionally, 1 patient had a serious AE that did not resolve upon dose interruption.
Presenting the findings, coauthor Lyudmila A. Bazhenova, MD, of the University of California, San Diego, Health Moores Cancer Center in La Jolla, said, “RXDX-105 has demonstrated a manageable safety profile in patients with advanced or metastatic solid tumors.”Over the past 30 years, from the discovery of a fusion rearrangement to the recent promise of LOXO-292, RET targeting has gone from distant concept toward clinical reality. RET rearrangements and mutations are represented across multiple tumor types, so emerging agents may serve as pan-cancer therapies. Whether RET-selective inhibitors are used alone or in combination with other agents remains to be seen, as does first-line efficacy. Regardless, the results from LIBRETTO-001 prompt optimism for the future of RET inhibition.