Aberrations in fibroblast growth factor receptor (FGFR) signaling are an emerging focus for targeted therapy across multiple types of cancer, particularly urothelial carcinoma, gastric cancer, and intrahepatic cholangiocarcinoma.
Aberrations in fibroblast growth factor receptor (FGFR) signaling are an emerging focus for targeted therapy across multiple types of cancer, particularly urothelial carcinoma, gastric cancer, and intrahepatic cholangiocarcinoma. However, recent research findings demonstrating variability in the efficacy of FGFR-targeted therapies across different FGFR aberrations emphasize the importance of patient selection for clinical trials and further studies of combination regimens to be used with FGFR inhibitors.
The fibroblast growth factor (FGF)/FGFR pathway is a tyrosine kinase signaling network involved in cell proliferation, differentiation, apoptosis, and migration.1-4 The FGF family includes 22 known ligands that bind to members of the FGFR family, which consists of 4 highly conserved transmembrane receptor tyrosine kinases (FGFR1/2/3/4) and 1 FGF-binding receptor that does not have an intracellular kinase domain (FGFR5 or FGFRL1). Binding of FGF ligands to FGFRs leads to dimerization and regulation of a cascade of downstream signaling pathways, including MAPK, STAT, PI3K/AKT, and DAG-PKC and IP3-Ca2+ pathways (Figure 1).4-7 Aberrations in FGFR represent a key target for cancer therapy for a subgroup of certain types of malignancies. A study using next-generation sequencing (NGS) on samples from about 5000 patients with various cancers showed FGFR aberrations in 7.1% (Figure 2). Amplifications in FGFR accounted for the majority of these aberrations (66%), followed by activating mutations (26%) and gene rearrangements or fusions (8%).8 Overall, aberrations in FGFR1/2/3/4 were most frequently found in urothelial (31.7%), breast (17.4%), endometrial (11.3%), and ovarian cancers (8.6%).8
Study findings show a variation in the type of aberration and the specific gene within the family across cancer types. Amplification accounts for approximately 89% of all FGFR1 aberrations8 and has been demonstrated in approximately 16% of non—small cell lung cancer (NSCLC),9,10 6% of small cell lung carcinomas,11 5% of hormone receptor (HR)-positive breast cancers,12 and 4% of triple-negative breast cancers.13 FGFR2 amplification was demonstrated in approximately 4% of gastric cancers14 and 4% of triple negative breast cancers.13 However, gastric and breast cancer cell lines with FGFR2 amplifications were particularly sensitive to selective FGFR inhibitors, suggesting that the FGFR2 amplification confers addiction to the FGFR pathway.15,16 FGFR3 amplification is relatively uncommon but was demonstrated in 3% of urothelial cancers.8
Activating mutations in FGFR3 are particularly prevalent in urothelial cancers, occurring in up to 80% of nonmuscle invasive urothelial cell carcinomas, 20% of high-grade invasive urothelial cancers, and 5% of cervical cancers.12 Activating mutations in FGFR2 occur in 12% to 14% of endometrial cancers and have been demonstrated in a small proportion of squamous NSCLCs, gastric cancers, and urothelial cancers.8,12 Activating mutations in FGFR1 and FGFR4 are relatively uncommon and have been observed in pilocytic astrocytoma (FGFR1) and rhabdomyosarcoma (FGFR4).12
FGFR3 translocations/fusions account for 15% to 20% of multiple myelomas and have been observed in glioblastoma and bladder cancer.12 FGFR2 translocations are found in approximately 14% of intrahepatic cholangiocarcinomas and occur occasionally in lung, thyroid, and prostate cancer.12,17,18 FGFR1 translocations are relatively uncommon but have been observed in glioblastoma, breast cancer, squamous cell lung carcinoma, and 8p11 myeloproliferative syndrome.12
Among members of this pathway, FGFR4 alterations were far less common; in the large NGS analysis, FGFR4 mutations were found in 0.5% of the samples tested.8Given the key role of the FGFR signaling pathway in tumorigenesis for a subset of cancers, FGFR-targeted therapies represent an important area of research. Many nonselective tyrosine kinase inhibitors (TKIs) designed to inhibit VEGFR and/or PDGFR also inhibit FGFR because the kinase domains of these receptor families are phylogenetically related. The nonselective TKIs lenvatinib (Lenvima), regorafenib (Stivarga), ponatinib (Iclusig), and pazopanib (Votrient) are the only FDA-approved agents that have FGFR as one of their targets, although the research on their ability to inhibit FGFR is primarily limited to preclinical models and case studies. Multitargeted TKIs offer the potential to inhibit multiple pathways simultaneously (eg, angiogenesis and proliferation); however, the dosing may not be sufficient to inhibit FGFR and may be limited by drug-related toxicities. Additionally, the degree to which these agents inhibit FGFR is unclear, as evidenced by the difference in the adverse events that occur with most nonselective FGFR inhibitors (eg, hypertension, proteinuria, thrombotic microangiopathy, and hypothyroidism) and selective FGFR inhibitors (eg, hyperphosphatemia, nail and mucosal disorders, and reversible retinal pigmented epithelial detachment).12 Several selective FGFR TKIs have shown promising therapeutic efficacy in preclinical and early-phase trials and may have more manageable off-target effects than nonselective TKIs (Table).Dovitinib
Unlike most other nonselective FGFR TKIs, dovitinib’s activity in FGFR-aberrated tumors has been investigated in clinical trials, rather than in in vitro and preclinical models, although whether its antitumor activity is mediated through FGFR is questionable. Findings from a phase III trial showed that third-line therapy with dovitinib failed to achieve superior progression- free survival (PFS) and overall survival (OS) compared with sorafenib (Nexavar) in patients with metastatic renal cell carcinoma, and baseline levels of FGF2 expression, evaluated by immunoassay, did not predict response, suggesting that the effects of dovitinib may have been mediated through inhibition of other tyrosine kinases.19 Furthermore, findings from a phase II trial demonstrated that although dovitinib was well tolerated, its limited single-agent activity in patients with progressive FGFR3-mutated or FGFR3 wild-type urothelial carcinoma led to early termination of the trial.20
Lucitanib demonstrated strong antiangiogenic activity in tumor xenograft models with FGFR1 amplification.21 Findings from the multicohort phase II FINESSE trial showed that patients with HR-positive, HER2-negative breast cancer with high FGFR1 expression had a 25% objective response rate (ORR) to lucitanib versus 8% in patients with low FGFR1 expression by immunohistochemistry.22 In addition, serum FGF23 levels significantly increased 14 days after initiation of treatment, indicating effective targeting of FGFR1. However, adverse events were common, with grade >3 hypertension occurring in 76%, hypothyroidism in 45%, nausea in 33%, and proteinuria in 32%; the trial was terminated prematurely based on a decision by the study sponsor Servier. Subsequently, Clovis gained full rights to lucitanib and plans to initiate trials investigating lucitanib and rucaparib (Rubraca), a PARP inhibitor, based on promising data of combined inhibition of VEGF, which lucatinib also targets, and PARP.
Lenvatinib, whose molecular targets include FGFR1/2/3/4, VEGFR1/2/3, RET, KIT and PDGFR-α is approved by the FDA for radioiodine-refractory, well-differentiated thyroid carcinoma and in combination with everolimus (Afinitor) for metastatic renal cell carcinoma after VEGF-targeted therapy. Results from the phase III REFLECT trial23 also showed that lenvatinib had noninferior OS compared with sorafenib in patients with untreated nonresectable hepatocellular carcinoma (HCC), and the results of an in vitro study24 showed that lenvatinib demonstrated antitumor activity and reduced tumor microvessel density in a PLC/ PRF/5 xenograft model and 2 HCC patient-derived xenograft models. The authors concluded that lenvatinib may exert antitumor activity in HCC by targeting FGF/FGFR signaling and that further investigation into anti-FGFR activity in clinical trials may be warranted.
Regorafenib inhibits multiple tyrosine kinases, including FGFR, and was approved for metastatic colorectal cancer and advanced gastrointestinal stromal tumors in 2012. However, the mechanism by which regorafenib exerts its antitumor effects and biomarkers that predict response have been unclear until recently.
Findings from a preclinical study showed that FGFR2 amplification was the only gene aberration associated with sensitivity to regorafenib.25 This study also showed that regorafenib dose-dependently inhibited phosphorylation of FGFR2 and its downstream signaling molecules in FGFR2-amplified gastric and colorectal cancer cell lines and inhibited tumor growth, proliferation, and FGFR signaling in SNU-16 mouse xenografts. The authors suggested clinical studies investigating regorafenib in patients with FGFR2-amplified cancers are warranted.
Ponatinib is approved for chronic myeloid leukemia and Philadelphia chromosome—positive acute lymphoblastic leukemia; it targets the BCR-ABL gene mutation T315I but has also demonstrated activity against other targets, including FGFR. Ponatinib inhibited FGFR-mediated signaling and cell growth in 14 cell lines representing multiple tumor types and mechanisms of FGFR dysregulation, and oral dosing inhibited tumor growth and signaling in 3 mouse tumor models, suggesting that ponatinib could be a candidate for treatment of cancer FGFR aberrations.26
Pazopanib primarily targets VEGFRs and PDGFRs and is approved for treatment of advanced renal cell cancer and soft tissue sarcoma. In vitro study results showed that treatment with pazopanib led to decreases in cell survival, constitutive FGFR2 signaling, and phosphorylation of downstream effectors in multiple FGFR2-amplified gastric cancer cell lines, suggesting that pazopanib may provide genotype-associated therapeutic benefits beyond its antiangiogenic effects in vascular tumors.27 Additionally, a case study showed that pazopanib led to near-disappearance of brain lesions and therapeutic improvement in lung and liver function in a patient with HR-positive, FGFR1-amplified metastatic breast cancer that was resistant to endocrine therapy, suggesting that pazopanib may have clinical activity against FGFR1 amplifications.28Erdafitinib
Erdafitinib (JNJ42756493) is selective pan-FGFR inhibitor that received a breakthrough therapy designation from the FDA in March 2018 based on results from the phase II BLC2001 trial, which showed that 8-mg/kg continuous dosing led to an ORR of 42% in 59 patients with locally advanced or metastatic urothelial cancer with actionable FGFR2 or FGFR3 mutations or fusions.29 An ongoing open-label phase III trial will compare erdafitinib with vinflunine (Javlor), docetaxel, or pembrolizumab (Keytruda) in patients with advanced urothelial cancer with selected FGFR gene alterations (NCT03390504). A phase IB trial aims to identify the phase II dose and schedule of erdafitinib in combination with an anti—PD-1 antibody (JNJ-63723283) and evaluate the safety and clinical efficacy of erdafitinib with or without JNJ-63723283 in patients with unresectable or metastatic urothelial cancer with selected FGFR gene alterations (NCT03473743). In addition, a phase IB trial is investigating erdafitinib, fulvestrant (Faslodex), and palbociclib (Ibrance) in patients with recurrent or unresectable estrogen receptor—positive, HER2-negative, and FGFR-amplified breast cancer (NCT03238196).
The selective pan-FGFR inhibitor infigratinib also showed promising findings in a phase II trial of patients with previously treated advanced intrahepatic cholangiocarcinoma whose tumors harbored FGFR2 fusions.30 The ORR (confirmed and unconfirmed) was 31.0%, with partial responses (PRs) observed in 25.4% and stable disease (SD) in 57.7%. Although grade ≥3 events occurred in 66.2% of patients, the authors stated in their conference abstract that toxicity was manageable with phosphate binders and routine supportive care. A phase III trial (NCT03773302) started recruitment in 2019 and will compare infigratinib with standard of care treatment (gemcitabine and cisplatin) in the frontline setting for patients with unresectable or metastatic cholangiocarcinoma harboring FGFR2 gene fusions/translocations.
“The next step here is to see whether we can expand this experience in the first-line setting, where this drug can be given upfront without chemotherapy. I believe that the efficacy will be seen in the first-line setting as well,” said lead author Milind Javle, MD, professor, Department of Gastrointestinal Medical Oncology, Division of Cancer Medicine, The University of Texas MD Anderson Cancer Center, in an interview with OncologyLive®.
Data from the phase II trial FIGHT-202 showed an ORR of 40% (19 patients) with pemigatinib, a selective inhibitor of FGFR1/2/3, in the first 47 patients enrolled with pretreated advanced/metastatic or surgically unresectable cholangiocarcinoma whose tumors had FGFR2 translocations.31 An additional 21 patients (44.7%) had SD, leading to a disease control rate of 85% (40 of 47 patients). Median PFS and OS were 9.2 and 15.8 months, respectively, which were a substantial improvements on historical PFS and OS values (3.2 months and 7.2 months, respectively) for patients with biliary tract cancers receiving second-line therapy.31,32 Recruitment has begun for the randomized phase III trial FIGHT-302, which will compare pemigatinib with gemcitabine plus cisplatin in patients with newly diagnosed cholangiocarcinoma and activating FGFR2 translocations (NCT03656536).
Interim analyses of phase II trials conducted by Incyte, the drug developer, showed that pemigatinib induced SD or PR (confirmed or unconfirmed) in 30 of 64 patients with FGFR3-translocated metastatic or surgically unresectable bladder cancer (FIGHT- 201 trial).33 This led to a major cytologic response and clinical or partial clinical responses in the bone marrow, peripheral blood, and extramedullary disease in 8 of 10 patients with FGFR1-translocated myoproliferative neoplasms (FIGHT-203 trial).34
AZD4547 is a small-molecule selective inhibitor of FGFR1/2/3 with antitumor activity demonstrated in several cell lines and tumor xenograft models.35 The results of a phase I open-label study of patients with solid malignancies showed that AZD4547 induced PR or SD in 5 of 15 patients with FGFR1-amplified squamous cell lung cancer36 and 5 of 13 patients with FGFR1 or FGFR2 amplified gastric and gastroesophageal cancer.37 Phase II trial results also showed that the activity of AZD4547 was higher in FGFR2-amplified gastroesophageal cancer than in FGFR1-amplified HER2-negative breast cancer, highlighting the discrepancy in response to FGFR inhibition among different types of FGFR aberrations.38
Results from a phase I trial showed that TAS-120, a highly selective irreversible pan-FGFR inhibitor, led to a confirmed PR in 7 of 28 evaluable patients with cholangiocarcinoma harboring an FGFR2 gene fusion and SD as the best response in 15 patients.39 Based on these preliminary findings, TAS-120 was granted orphan drug status by the FDA in May 2018 for treatment of cholangiocarcinoma, according to the manufacturer Taiho Oncology. The phase I dose-expansion trial (NCT02052778) is currently recruiting for multiple cohorts, including patients with cholangiocarcinoma harboring other FGF/ FGFR aberrations and gliomas.Although several therapies targeting FGFR aberrations appear promising, challenges remain in selecting patients for clinical trials, assessing the variability in dependence on FGF/FGFR signaling for tumorigenesis, and targeting resistance mechanisms to FGFR inhibitors.
Overall, the proportion of patients with FGFR aberrations is relatively low (~7%). Many early trials of FGFR-directed therapy used basket-style recruitment, in which patients with all types of FGFR aberrations were included. Because FGFR aberrations vary in their response to targeted therapy, trials have started to focus on individual aberrations, which further narrows the pool of eligible participants for a particular study. Furthermore, the criteria for FGFR amplification that would benefit from FGFR-directed therapy (eg, gene to centromere ratio) and the effects of clonality on response to therapy are unclear.
Additionally, the difference in the addition to FGFR signaling among FGFR aberrations may affect the response to targeted therapy and introduce the need for a combination therapy in tumors that also rely on alternative pathways for growth and proliferation. In vitro and sequencing studies suggest that FGFR is not a dominant oncogene in FGFR1-amplified cancers,40 whereas FGFR2-amplified cells are highly addicted to FGFR signaling and demonstrate an apoptotic response to FGFR inhibition.15
Finally, development of drug resistance is a persistent challenge for all targeted therapies, including those for FGFR. Gatekeeper mutations in the adenosine triphosphate binding cleft that cause resistance to FGFR-targeted therapy have been identified in preclinical models, and activation of alternative receptor tyrosine kinases have been described as an escape mechanism in FGFR-resistant tumors.41-43 Identifying these mechanisms of drug resistance and ways to target resistance will further help improve efficacy of targeted therapy in FGFR-aberrated cancers.