FDA Approval Summary: Entrectinib for the Treatment of NTRK gene Fusion Solid Tumors

The FDA-approved entrectinib on August 15, 2019, for the treatment of adult and pediatric patients 12 years of age and older with solid tumors that have a neurotrophic tyrosine receptor kinase (NTRK) gene fusion without a known acquired resistance mutation, are metastatic or where surgical resection is likely to result in severe morbidity, and have progressed following treatment or have no satisfactory alternative therapy. This tissue agnostic indication relies on the common oncogenic effects of NTRK gene fusions across tumors and demonstration of a treatment effect of large magnitude and durability across a variety of primary cancer types. This article summarizes the data submitted, issues identified during review, and the basis for approval.

During review of this submission, an estimated 1.7 million patients where predicted to be diagnosed with cancer and more than 600,000 people to die of cancer in the United States (1). In 2018, the annual incidence of new cancers that are NTRK gene fusion–driven (NTRK fusion) tumors was estimated at 1,500–5,000 cases in the United States (2). The presence of NTRK fusion is pathognomonic or very frequent in a few rare cancers, specifically secretory carcinoma of the breast (3, 4), infantile fibrosarcoma (5), and congenital mesoblastic nephroma (6). However, the frequency of NTRK fusions in the most common solid tumors in the United States, that is, lung, breast, colon cancers, is very low (<1%). There are limited data regarding the prognostic relevance of NTRK fusions in common solid tumors.

Prior to the approval of entrectinib, pembrolizumab and larotrectinib received FDA approval for a biomarker-based indication regardless of the site of origin (i.e., a tissue-agnostic indication). Pembrolizumab was approved for the treatment of adult and pediatric patients with unresectable or metastatic, microsatellite instability-high (MSI-H) or mismatch repair deficient solid tumors that have progressed following prior treatment and who have no satisfactory alternative treatment options, based on demonstration of an overall response rate (ORR) of 39.6% [95% confidence interval (CI): 31.7–47.9], including a 7% complete response (CR) rate, in 147 patients with various MSI-H solid tumors; 78% of responses were durable for ≥ 6 months. Larotrectinib was approved for the treatment of adult and pediatric patients with solid tumors that have an NTRK gene fusion without a known acquired resistance mutation and are metastatic or where surgical resection is likely to result in severe morbidity and who have no satisfactory alternative treatments or that have progressed following treatment based on demonstration of an ORR of 75% (95% CI: 61–85), including a 22% CR rate, in 55 patients with various NTRK-positive solid tumors; 63% of responses were durable for ≥ 6 months.

The entrectinib commercial formulation is an immediate-release hard capsule for oral administration, manufactured with standard excipients using conventional equipment and manufacturing processes. The commercial formulation (F06) demonstrated bioequivalence to the developmental formulation (F2A) administered in the three clinical efficacy trials. Two dose-proportional capsule strengths (100 mg and 200 mg entrectinib) are available in the U.S. market.

Entrectinib is a multikinase inhibitor. Both entrectinib and its major metabolite, designated M5, inhibit the tropomyosin receptor tyrosine kinases (TRK) TRKA, TRKB, and TRKC encoded by the NTRK1, NTRK2, and NTRK3 genes, respectively, as well as the proto-oncogene tyrosine-protein kinase ROS1 (ROS1) and the anaplastic lymphoma kinase (ALK) with IC₅₀ values of 0.1 to 2 nmol/L. Entrectinib also inhibits JAK2 and TNK2 with IC₅₀ values > 5 nmol/L. Entrectinib demonstrated in vitro and in vivo inhibition of cancer cell lines derived from multiple tumor types harboring NTRK, ROS1, or ALK fusions. Following either intravenous or oral dosing, entrectinib was detectable in brain tissue in multiple animal species (mice, rats, and dogs) at exposures as high as 60%–100% those in plasma, depending on duration of treatment. In vivo antitumor activity occurred in mice implanted intracranially with TRKA- or ALK-driven tumor cell lines.

Entrectinib is not mutagenic in vitro and did not induce DNA damage in a comet assay in rats; however, it did demonstrate a potential for abnormal chromosome segregation. Developmental studies showed malformations at three times the human exposure and skeletal variations at exposures as low as 0.2 times the human exposure at 600 mg. Juvenile animal studies showed growth delays and impairments in learning and memory, consistent with the drug's primary pharmacology (7, 8).

The pharmacokinetics of entrectinib and its active metabolite M5 are linear and are not dose or time dependent. Steady state is achieved within 1 week for entrectinib and 2 weeks for M5 following daily administration of entrectinib. The elimination half-lives of entrectinib and M5 were estimated to be 20 and 40 hours, respectively. The exposures of entrectinib in pediatric patients ages 12 years or older administered at 600 mg (for BSA > 1.50 m²), 500 mg (for BSA of 1.11–1.50 m²), and 400 mg (for BSA of 0.91–1.10 m²) are comparable with the exposure in adults taking 600 mg daily. There was insufficient pharmacokinetic information on the marketed formulation in the application to determine the entrectinib dose for pediatric patients less than 12 years of age that would provide exposures comparable to the exposures obtained at the recommended dosages for older patients. The pharmacokinetics of entrectinib in the central nervous system (CNS) were not evaluated in clinical studies submitted in the New Drug Application (NDA). However, entrectinib distributed to the CNS in preclinical species models and treatment responses were observed in human patients with CNS tumors or CNS metastases.

Food does not alter the absorption of entrectinib and adjustments of dose is not recommended on the basis of sex, race, mild to moderate renal impairment (creatinine clearance 30 to < 90 mL/minute), or mild hepatic impairment (total bilirubin ≤ 1.5 times upper limit of normal). Entrectinib is primarily metabolized by CYP3A4, with predicted clinically important increases in exposure when coadministered with moderate or strong CYP3A4 inhibitors. The effects of moderate to severe hepatic impairment or severe renal impairment on the pharmacokinetics of entrectinib have not been studied.

Entrectinib exposure-response relationships and the time course of pharmacodynamic responses are unknown. Across clinical trials, 3.1% of 355 patients who received entrectinib at doses ranging from 100 to 2,600 mg daily under fasting or fed conditions (75% received 600 mg orally once daily) and had at least one postbaseline ECG assessment, experienced QTcF interval prolongation of >60 ms after starting entrectinib and 0.6% had a QTc interval > 500 ms (9).

The NDA contained interim efficacy data from three ongoing open-label, single-arm or noncomparative multicohort clinical trials in adults, ALKA-372-001 [GO40783] (EudraCT 2012-000148-88), STARTRK-1 (NCT02097810), and STARTRK-2 (NCT02568267), and interim safety data from these three trials and from the pediatric study STARTRK-NG (NCT02650401). The main design elements are summarized in Table 1. Key eligibility criteria included patients with solid tumors that progressed following effective systemic therapy for their primary cancer, if available, or requiring surgery with significant morbidity. The objectives of all trials were to identify a safe and tolerable dose and estimate the ORR and durations of response per RECIST v1.1. The prespecified sample size for the efficacy population of 56 adult patients enrolled in ALKA-372-001, STARTRK-1, or STARTRK-2 was based on the following assumptions: the true ORR by blinded independent central review (BICR; ORR‐BICR) is 60% and the 95% two‐sided CI around the observed ORR would exclude a lower limit of 30%.

The NDA provided data for 355 entrectinib-treated patients with various cancers that may or may not contain an NTRK1, NTRK2, NTRK3, ROS1, or ALK fusion. At the time of the NDA submission, enrollment was ongoing in Studies STARTRK-1 (NCT02097810), STARTRK-2 (NCT02568267), and STARTRK-NG (NCT02650401; ref. 10) and follow-up for safety and efficacy was ongoing in all four trials. The impact of this on the assessment of safety is discussed below.

The NTRK efficacy population comprised the first sequentially enrolled 54 adult entrectinib-treated patients enrolled in ALKA-372-001 (n = 1), STARTRK-1 (n = 2), or STARTRK-2 (n = 51) with a solid tumor harboring NTRK1, NTRK2, or NTRK3 gene fusions as determined by a central laboratory (Pharos) or a local Clinical Laboratory Improvement Amendments–certified or equivalently accredited diagnostic laboratory using a nucleic acid–based diagnostic testing method; no evidence of a second driver mutation (e.g., EGFR, ROS1, or ALK); sufficient follow-up to assess durability of responses (≥6 months after first dose of entrectinib); and no prior exposure to a TRK inhibitor. All patients in this subgroup received entrectinib 600 mg orally, once daily, until disease progression or unacceptable toxicity. Genentech provided 5 months' of additional follow-up to characterize durability of responses during the NDA review.

The baseline demographic and prognostic characteristics for the pooled efficacy population were: median age of 57 years (range: 21–83); female predominance (59%); White (80%), Asian (13%), and Hispanic or Latino (7%); and Eastern Cooperative Oncology Group performance status 0 (43%) or 1 (46%). The majority (96%) had metastatic disease, including 22% with CNS metastases; the remaining 4% had locally advanced, unresectable disease. The most common NTRK fusions involved NTRK3 (57%) or NTRK1 (41%), with only one patient's cancer (2%) harboring an NTRK2 fusion. All patients received prior treatment that included surgery (78%), radiotherapy (67%), or systemic therapy (89%).

The ORR-BICR was 57% (95% CI: 43–71), which included a 7% complete response rate. The duration of response ranged from 2.8 months to 26.0+ months, with 68% of responses lasting ≥ 6 months. Among 4 patients with measurable CNS metastases at baseline identified by the BICR and no radiation to the brain within 2 months of study entry, 3 patients achieved an objective response in intracranial lesions. Health-related quality-of-life data were collected only in STARTRK-2; insufficient numbers of patients completed disease-specific questionnaires (e.g., QLQ-LC13, QLQ-CRC29) to assess patient-reported outcomes.

The safety of entrectinib was evaluated in a pooled population of 355 pediatric and adult entrectinib-treated patients enrolled in the four studies. The baseline demographic and prognostic characteristics of the safety population were: median age 55 years (range: 4–86 years); 5% (n = 17) were 18 years or younger; 55% were female; and 66% were White, 23% were Asian, and 5% were Black; 3% were Hispanic/Latino. The most common primary cancer types were lung (56%), sarcoma (8%), and colon (5%). ROS1 fusions were present in 42% and NTRK fusions were present in 20% of tumors. In adults, the entrectinib daily dose ranged from 200 to 2,600 mg, with the majority of adults (75%) receiving entrectinib 600 mg orally daily. Among the 17 pediatric patients, the entrectinib daily dose ranged from 250 to 750 mg/m².

Nearly all (99%) patients experienced at least one adverse event (AE), 60% experienced a grade 3 or 4 AE, and 39% experienced a serious adverse reaction. The most common grade 3–4 AEs were pulmonary infections, weight gain, dyspnea, fatigue/asthenia, cognitive disorders, syncope, pulmonary embolism, hypoxia, pleural effusion, hypotension, diarrhea, and urinary tract infections. Serious risks of entrectinib include congestive heart failure (CHF), CNS adverse reactions, skeletal fractures, hyperuricemia, hepatotoxicity, QT prolongation, and vision disorders.

FDA concluded that the demonstrated effect on ORR [ORR 57% (95% CI: 43–71)] and durability of response [68% of the 31 responders with duration of response (DOR) ≥ 6 months and 45% with DOR ≥12 months] provided substantial evidence of the effectiveness of entrectinib for the treatment of adult and pediatric patients 12 years of age and older with advanced solid tumors that harbor a NTRK gene fusion and either progressed following treatment or have no satisfactory alternative treatments. FDA also determined that the benefits of entrectinib in this patient population outweighed the serious risks of entrectinib, which include CHF, neurocognitive impairment, mood disorders and other CNS effects, skeletal fractures, hepatotoxicity, QT prolongation, ocular toxicity, and hyperuricemia requiring medical intervention.

The FDA assessment of benefit in the entrectinib application posed challenges similar to those encountered in the tissue agnostic approvals for pembrolizumab (11) and larotrectinib (12) in that there were limited numbers of patients comprising the efficacy population and the population was heterogeneous with respect to factors other than the tumor biomarker. For all three approvals, the possibility of differential antitumor activity (ORR and response duration) across different tumor types existed on the basis of the site of tumor origin, extent of prior treatment, disease burden, or other factors such as fusion partner or the presence of additional unknown oncogenic driver mutations (11). Thus, there was inherent uncertainty regarding the generalizability of similar antitumor activity with entrectinib across all tumor types despite the presence of NTRK fusions. This risk was considered low for entrectinib based on the strong biologic rationale, supportive preclinical data, the observed magnitude and durability of responses (ORR-BICR of 57%; 68% of responses durable ≥ 6 months), which is similar to results supporting the pembrolizumab and larotrectinib tissue agnostic approvals; and the consistency of response observed (see Table 2) across primary cancers harboring a variety of NTRK fusion partners. The application was further supported by evidence from previous approvals with other drugs in that a large absolute magnitude of improvement in progression-free survival or a large response rate of prolonged durability provided compelling evidence of benefit over time (13–16). The potential risk of inappropriate generalization of the treatment effect to all cancers harboring an NTRK fusion was mitigated by the indication, which is limited to the treatment of patients who have no satisfactory alternative treatment options or whose cancer has progressed following treatment. Therefore, patients will not be forgoing effective therapies if treated with entrectinib. To address these uncertainties and potential risks, Genentech must conduct a study to verify and confirm the benefit of entrectinib through more precise estimation of ORR and response duration in additional patients including an adequate number of patients with common solid tumors (e.g., colorectal cancer, gynecologic cancers, and melanoma) harboring NTRK fusions where the ORR in this tumor type has not been precisely estimated as well to gain additional efficacy data in NTRK fusion pediatric solid tumors and primary central nervous system malignancies. This information may also be supplemented with real-world data.

Similar to larotrectinib and unlike pembrolizumab, the safety population supporting the initial approval of entrectinib is limited because of the rarity of NTRK fusion solid tumors, such that serious risks of the drug occurring at an incidence of less than 1% may not have been observed. In addition, effects on growth and development in pediatric patients have not been fully characterized. Therefore, Genentech must conduct postapproval clinical studies to further assess for serious risks of cardiac toxicity (i.e., cCHF), skeletal fractures, and signals of adverse long-term effects on growth and neurodevelopment in pediatric patients ages 12 years and older.

Finally, FDA considered the implications of the absence of FDA-approved tests to identify patients with NTRK-fusion positive solid tumors at the time of approval. Although FDA generally expects that companion in vitro diagnostic device(s) will be approved contemporaneously with the drug, FDA guidance (17) states that “if the benefits from the use of the therapeutic product are so pronounced as to outweigh the risks from the lack of an approved or cleared in vitro diagnostic (IVD) companion diagnostic device, FDA does not intend to delay approval of changes to the labeling of the therapeutic product until the IVD companion diagnostic device is approved or cleared.” Patient eligibility for enrollment was predominantly based on NGS-based tests for NTRK fusions available in the community as laboratory-developed tests. FDA determined that despite the potential risk of false-positive results, given the efficacy observed and the indication for use in patients where no satisfactory alternatives exist, the potential benefits outweighed the risks and approval should not be delayed until a companion diagnostic test is FDA approved for this use. Instead, FDA and Genentech agreed to develop and submit companion diagnostic tests for detection of NTRK fusions as postmarketing commitments.

FDA acknowledges that additional unanswered questions exist with regard to use of entrectinib in this population. Such questions include the optimal duration of therapy, appropriate dose for pediatric patients less than 12 years of age, and whether the addition of other drugs to entrectinib would improve outcomes. Development of site-agnostic indications requires forethought regarding both the drug and the method of selection of patients (e.g., companion diagnostic device). Furthermore, collaboration between stakeholders (government, industry, academia, patients) will be necessary to define these potential new indications.

P. Keegan reports other from Top Alliance Biosciences outside the submitted work (after completing her contribution to the submitted work, she left FDA and now works at Top Alliance Biosciences). No disclosures were reported by the other authors.

The Editor handling the peer review and decision-making process for this article has no relevant employment associations to disclose.