Main

NTRK1/NTRK2/NTRK3 fusions (NTRK) drive oncogenesis through aberrant tropomyosin receptor kinase (TRK) protein signaling1 in up to 0.7% of solid tumors, although prevalence varies greatly by tumor type2,3. NTRK fusions are common in secretory carcinoma of the breast (91%) and in mammary analog secretory carcinoma of the salivary gland (83–90%) among adult patients3; by contrast, the frequency of NTRK fusions in non-small cell lung cancer (NSCLC) is 0.2%4. The use of early-generation TRK TKIs larotrectinib and entrectinib resulted in meaningful clinical benefit in patients with NTRK+ solid tumors5,6; however, durability of responses to these agents is limited by acquired resistance, including on-target NTRK mutations7,8. On-target mutations may occur in the following regions: the solvent front (for example, TRKA G595R, TRKB G639R and TRKC G623R), the xDFG motif (for example, TRKA G667C, TRKB G709C and TRKC G696A) and the gatekeeper residue. These mutations alter the TRK kinase domain conformation, causing interference with TRK inhibitor binding. Bypass resistance or off-target resistance mechanisms have also been described and include activation of non-TRK oncoproteins, such as activating BRAF or KRAS mutations9.

A TRK TKI that can maintain durable response among patients with or without previous TRK TKI treatment and address acquired resistance mutations would provide meaningful benefit to patients with NTRK+ solid tumors. With its compact macrocyclic structure that circumvents steric hindrance10, repotrectinib potently inhibits wild-type TRKA/TRKB/TRKC fusion proteins (half maximal inhibitory concentration (IC50) < 0.2 nmol l−1) and those with resistance mutations (IC50 ≤2.6 nmol l−1)8. Repotrectinib is a next-generation TRK and ROS1 TKI, approved in the United States for treatment of adult patients with ROS1 fusion–positive (ROS1+) locally advanced or metastatic NSCLC and for treatment of adult or pediatric patients (12 years and older) with NTRK+ locally advanced or metastatic solid tumors11,12. Repotrectinib was recently added to the National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology (NCCN Guidelines) as a preferred first-line treatment option for patients with NTRK+ NSCLC13 and as a treatment option in various NTRK+ solid tumors14,15,16, including soft tissue sarcoma17 and cancers of the breast18, rectum19 and colon20.

TRIDENT-1 is an ongoing international, registrational phase 1/2 trial evaluating repotrectinib in adult patients with ROS1+ and NTRK+ tumors21. We report the efficacy outcomes of repotrectinib in patients with NTRK+ solid tumors, including patients with NTRK solvent front mutations, and its safety in all patients regardless of tumor or fusion type.

Results

Patients and treatment

From 27 February 2017 through 15 October 2023, 144 patients with NTRK+ solid tumors were enrolled in TRIDENT-1 (Extended Data Fig. 1). Of the 135 patients in phase 2, 112 (83%) increased dose from 160 mg once daily to 160 mg twice daily at day 15 per protocol. The efficacy population included 120 patients with NTRK+ solid tumors who started treatment with repotrectinib at any dose by 15 February 2023 (phase 1, n = 9; phase 2, n = 111) (Fig. 1). At data cutoff, 42 patients (35%) were still receiving treatment across the NTRK+ cohorts (29 TKI-naive; 13 TKI-pretreated). The median duration of treatment was 33.1 months (range, 0.1–51.2+) for TKI-naive and 7.6 months (range, 0.6–32.0) for TKI-pretreated cohorts. The most common reason for treatment discontinuation was disease progression (n = 41; 34%). Extended Data Fig. 2 shows the duration of treatment. Supplementary Table 1 summarizes subsequent therapies.

Fig. 1: CONSORT diagram.
figure 1

aOther treated patients include patients with ROS1+ non-NSCLC, ALK gene fusions and any gene fusions with discordant results between local FISH test and central laboratory test. bAt data cutoff, 42 patients remained on treatment (29 in the TRK TKI-naive cohort and 13 in the TRK TKI-pretreated cohort). The most common reason for discontinued treatment was disease progression (41 patients). cNine NTRK+ patients in phase 1 (five in the TRK TKI-naive cohort and four in the TRK TKI-pretreated cohort). dEfficacy analysis set includes treated patients with at least 6 months of follow-up after the first post-baseline scan.

Full size image

Activity

Efficacy population

The efficacy population included 51 TKI-naive patients and 69 TKI-pretreated patients with sufficient follow-up for analysis (that is, at least 6 months of follow-up for tumor assessment after first post-baseline scan) (Table 1). Among 18 NTRK+ tumor types reported, NSCLC was the most common at 53% and 25% in the TKI-naive and TKI-pretreated cohorts, respectively, followed by thyroid cancer (12% and 10%), salivary gland cancer (10% and 17%) and soft tissue sarcoma (6% and 14%) (Supplementary Fig. 1). A total of 12 and 16 unique cancer types were treated in the TKI-naive and TKI-pretreated cohorts, respectively. Patients with asymptomatic leptomeningeal carcinomatosis were eligible, although none enrolled. The median age was 61 years (range, 25–84) in the TKI-naive cohort and 56 years (range, 18–81) in the TKI-pretreated cohort; approximately half were women (53% and 48%, respectively). Per blinded independent central review (BICR), intracranial disease at baseline was present in 20% and 23% of patients, respectively (Table 1).

Table 1 Characteristics of patients at baseline (efficacy population)a
Full size table

TRK TKI-naive patients

In the TKI-naive cohort, median follow-up was 25.7 months (range, 8.7–74.5). Thirty patients had confirmed response (59%; 95% CI: 44–72), eight (16%) had complete response and 22 (43%) had partial response (Table 2 and Fig. 2a), with median time to response of 1.8 months (1.6–7.3). Responses were observed regardless of tumor type, NTRK gene or fusion partner (Extended Data Fig. 3a). Median DOR was NE, and an estimated 85% of responders (95% CI: 70–99) had response lasting at least 24 months (Fig. 2b). The median PFS was 30.3 months (95% CI: 9.0–NE), and an estimated 60% of patients (95% CI: 46–74) were alive without progression for at least 24 months (Extended Data Fig. 4a). Estimated 24-month overall survival rate was 68% (95% CI: 54-83) (Extended Data Fig. 5a); overall survival was considered immature. Of 18 patients without prior systemic therapy, 61% (95% CI: 36–83) had a response (Supplementary Table 2). Among 27 patients who had NTRK+ NSCLC, 63% (95% CI: 42–81) had a response (Supplementary Table 3). Efficacy outcomes by sex are shown in Supplementary Table 4.

Fig. 2: Efficacy outcomes of repotrectinib in the efficacy population.
figure 2

Shown are the change in the tumor burden (a) and DOR (b) in 51 TKI-naive patients (five patients from phase 1 and 46 patients from phase 2) and the change in the tumor burden (c) and DOR (d) in 69 TKI-pretreated patients (four patients from phase 1 and 65 patients from phase 2). In a and c, the waterfall plots include only patients with baseline and post-baseline target lesion measurements at baseline and during follow-up; dashed lines indicate a reduction of 30% or an increase of 20% from baseline in the tumor size, as assessed according to RECIST version 1.1. Clinical benefit rate (CBR) refers to complete response plus partial response and stable disease. cORR, confirmed objective response rate; CR, complete response; mo, months; PR, partial response.

Full size image
Table 2 Repotrectinib efficacy outcomes (efficacy population)
Full size table

TRK TKI-pretreated patients

In the TKI-pretreated cohort, 63 patients (91%) received TKI as their most recent prior therapy, of whom 56 (89%) discontinued due to disease progression. At a median follow-up of 21.3 months (range, 8.4–79.2), 33 of 69 patients (48%; 95% CI: 36–60) had a confirmed response, two (3%) had complete response and 31 (45%) had partial response (Table 2 and Fig. 2c). The median time to response was 1.9 months (range, 1.7–3.7), and the median DOR was 9.8 months (95% CI, 7.4–13.0). An estimated 42% of responders (95% CI: 24–59) had responses lasting at least 12 months (Fig. 2d). Responses were observed regardless of tumor type, NTRK gene or fusion partner (Extended Data Fig. 3b). The median PFS was 7.4 months (95% CI: 3.9–9.7), and an estimated 26% of patients (95% CI: 14–37) were alive and without progression for at least 12 months (Extended Data Fig. 4b). The median overall survival was 18.6 months (95% CI: 11.6–25.3); the estimated 12-month overall survival rate was 62% (95% CI: 50–75) (Extended Data Fig. 5b). Efficacy outcomes by sex are shown in Supplementary Table 4. Among TKI-pretreated patients, 46% received prior entrectinib and 52% received larotrectinib. Confirmed response occurred in 18 of 32 patients (56%) who previously received entrectinib and in 14 of 36 patients (39%) who previously received larotrectinib (Supplementary Table 5). A total of 63 patients (91%) received TKI as their most recent prior therapy (Supplementary Table 6). Overall, 17 patients had NTRK+ NSCLC; 53% (95% CI: 28–77) had a response.

Results were also assessed in a subgroup of patients with extended follow-up (at least 19 months from treatment initiation) (Supplementary Tables 7 and 8). Outcomes were similar to the efficacy population.

Intracranial activity

Systemic (intracranial and extracranial) efficacy outcomes in patients with and without baseline intracranial disease confirmed by BICR was observed (Supplementary Table 9). Among patients from phase 2 with measurable intracranial disease at baseline, intracranial response occurred in two of three patients in the TKI-naive cohort and in four of six patients in the TKI-pretreated cohort (Table 2). Range for intracranial DOR was 17.5–24.0 months and 5.5–10.6 months in these cohorts, respectively, and both responders (100%) in the TKI-naive cohort had intracranial response for at least 12 months. Among patients without intracranial disease at baseline, intracranial PFS (defined as time from administration of first dose to development of new brain lesions as first progression assessed by BICR or death) was 87% at 24 months in the TKI-naive cohort and 67% at 12 months in the TKI-pretreated cohort (Extended Data Fig. 6).

Solvent front mutations

Among the 69 TKI-pretreated patients, 30 (43%) had NTRK solvent front mutations at baseline (Supplementary Table 10). Of these, 16 (53%; 95% CI: 34–72) had confirmed response (Fig. 3 and Supplementary Table 11); the median DOR was 8.6 months (95% CI: 5.5–12.9), and the median PFS was 7.4 months (95% CI: 3.9–11.0) (Extended Data Fig. 7). Among the 37 patients (54%) without solvent front mutations, 17 (46%, 95% CI: 30–63) had confirmed response.

Fig. 3: Efficacy outcomes in TRK TKI-pretreated patients with NTRK solvent front mutations.
figure 3

Shown are changes in tumor burden in 30 TRK TKI-pretreated patients with NTRK solvent front mutations. Asterisks indicate that treatment is ongoing, and dashed lines indicate a reduction of 30% or an increase of 20% from baseline in the tumor size, as assessed according to RECIST version 1.1.

Full size image

Repotrectinib resistance

An exploratory analysis was conducted to examine resistance mutations developed by patients with disease progression on repotrectinib. Among 28 patients with paired baseline and end-of-treatment samples, 20 discontinued repotrectinib due to disease progression or death (two TKI-naive, 18 TKI-pretreated). No emergent NTRK mutations were detected in the two TKI-naive patients, and three emergent NTRK3 solvent front mutations were detected in the 18 TKI-pretreated patients (Supplementary Table 12). Notably, five patients who had mutations at baseline had no mutations detected at end of treatment.

Of the 18 patients with progressive disease as best overall response, seven had NTRK mutations at screening, all of which were absent at end of treatment (Supplementary Table 13). Additionally, activation of MAPK/PI3K pathways at baseline and the presence of p53 mutations/deletions at baseline were observed in some of these patients. Future analysis of resistance mechanisms may be evaluated based on analyses of circulating tumor DNA (ctDNA) samples collected at baseline, on treatment and after progression.

Safety

Among 565 patients with any tumor or fusion type who received at least one dose of repotrectinib, the most common any-grade TRAEs were dizziness (57%), dysgeusia (50%) and paresthesia (30%) (Table 3 and Supplementary Table 14). Most TRAEs were grade 1 or grade 2 (71%). Grade 3 or higher TRAEs occurred in 162 patients (29%); the most common were anemia (4%), dizziness (3%) and increased blood creatine phosphokinase level (3%).

Table 3 Adverse events in patients who received at least one dose of repotrectiniba
Full size table

Treatment-emergent adverse events (TEAEs) led to dose reduction in 216 patients (38%), dose interruption in 291 patients (52%) and treatment discontinuation in 61 patients (11%) (Table 3). Among patients who escalated to 160 mg twice a day and had subsequent dose reduction due to TEAE within 3 months, an exploratory analysis demonstrated that median DOR in the TKI-naive and TKI-pretreated cohorts was not reached and 13.0 months (95% CI: 3.8–NE), respectively (Supplementary Fig. 2 and Supplementary Methods). The most common TEAEs that led to treatment discontinuation were dyspnea, pneumonitis and muscular weakness (1% each). No patients discontinued treatment due to dizziness. Serious TEAEs occurred in 230 patients (41%), and fatal TEAEs occurred in 35 patients (6%). Fatal TRAEs occurred in two patients, including sudden death in one patient from the TKI-pretreated ROS1+ NSCLC cohort and cardiorespiratory arrest in one patient from the TKI-pretreated NTRK+ solid tumor cohort. Safety outcomes were generally similar among all treated patients with NTRK+ solid tumors; however, dizziness, dysgeusia and ataxia appeared to occur more frequently in TKI-naive patients compared to TKI-pretreated patients (Supplementary Table 15).

Patient-reported outcomes

A total of 111 patients with NTRK+ solid tumors in phase 2 had European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire–Core 30 (EORTC QLQ-C30) assessments (46 TKI-naive and 65 TKI-pretreated), of whom ≥89% completed each assessment through cycle 12 (Extended Data Fig. 8), and 62–97% completed assessments between cycles 13 and 17. Among TKI-naive patients reporting data, mean global health status score at baseline was 67.4 (s.d. 24.2); 65% of patients had a stable score (<10-point increase or decrease from baseline) or improved score (≥10-point increase from baseline) at cycle 12 and cycle 17. Among TKI-pretreated patients reporting data, mean global health status score at baseline was 65.5 (s.d. 21.2); 78% and 69% of patients had a stable or improved score at cycle 12 and cycle 17, respectively.

Discussion

In this phase 1/2 trial, repotrectinib showed clinical activity in patients with a diverse variety of NTRK+ solid tumors, with responses observed regardless of tumor type, NTRK gene, intracranial disease status, prior treatment and NTRK resistance mutation status. Among TKI-naive patients, 59% had confirmed response, and responses were durable, as 85% of responders had responses lasting at least 24 months, and 60% were progression free for at least 24 months.

Other TRK inhibitors also reported activity in NTRK+ solid tumors with similar response rates (larotrectinib, 66%; entrectinib, 62%)5,6. Median PFS with repotrectinib (30.3 months) was similar to larotrectinib (30.8 months), and that of entrectinib was reported as 15.7 months. However, cross-trial comparisons should be interpreted with caution due to differences between trials, such as tumor type distribution and inclusion of pediatric patients. Although data are limited on how response rates vary between tumor types, certain tumors in adults, such as secretory carcinomas of the breast and salivary gland, may respond particularly well to TRK inhibitors compared to other tumor types22,23. Among pediatric patients, infantile fibrosarcoma has also been associated with dramatic responses with TRK inhibitors; for repotrectinib, pediatric patients are being evaluated separately (CARE; NCT04094610)24,25.

TRIDENT-1 is the first trial to demonstrate clinical efficacy outcomes in a TKI-pretreated population, with nearly half of the TKI-pretreated patients experiencing a confirmed response. Responses occurred in patients regardless of which TKI (larotrectinib or entrectinib) they received previously, and the median PFS was 7.4 months. Approximately half of the TKI-pretreated patients had on-target mutations associated with resistance to current TRK TKIs, with 44% having solvent front mutations at baseline. The response rate in patients with solvent front mutations was 53%, indicating the clinical relevance of the preclinical activity of repotrectinib against these mutations8.

The brain is a common site of disease progression, including in NTRK+ solid tumors26,27. As previously reported, repotrectinib was designed for enhanced intracranial activity and has shown antitumor activity in the brain in a patient-derived intracranial model21. In patients with NTRK+ solid tumors, repotrectinib was active regardless of the presence of baseline intracranial disease in both TKI-naive and TKI-pretreated cohorts, and evidence of durable intracranial responses was observed. Patients treated with repotrectinib who were without baseline intracranial disease showed no documented development of brain lesions.

Repotrectinib-related adverse events were mostly grade 1 or grade 2 (71%); dizziness was most common (57%). As TRKA/TRKB/TRKC receptors are known to be involved in the development and maintenance of the nervous system, neurologic adverse events such as dizziness are expected of TRK inhibition and were also observed in patients treated with larotrectinib and entrectinib28,29. TRAEs rarely led to treatment discontinuation (4%) in TRIDENT-1, and discontinuation because of dizziness was not reported. There did not appear to be a negative impact on DOR in patients with dose reduction due to an adverse event. Serious and fatal TEAEs occurred in 41% and 6% of treated patients, respectively. The safety profile of repotrectinib was consistent between patients with NTRK+ solid tumors and the overall safety population, including patients with ROS1+ NSCLC, as reported previously21.

This trial is limited by its single-arm design and small sample size, owing to the rare patient population with NTRK+ solid tumors. Additionally, NSCLC was the most frequent tumor type in the NTRK+ cohorts (37%), which is in line with real-world data2; however, responses were observed across other tumor types. Enrollment for the NTRK+ cohorts is ongoing, and time-to-event efficacy endpoints and safety will continue to be assessed for long-term outcomes.

In conclusion, repotrectinib, a next-generation TKI, showed durable clinical activity in adult patients with NTRK+ solid tumors, including in patients with previous TKI treatment, with or without NTRK solvent front mutations, across multiple NTRK genes and fusion partners and in patients with or without intracranial disease. Repotrectinib was mainly associated with low-grade adverse events, consistent with previous reports. This evidence supports repotrectinib as a new treatment option for patients with NTRK+ solid tumors.

Looking forward, studies with longer follow-up and larger patient populations will help understand how response durability translates to overall survival as well as ascertain the ideal treatment sequence (for example, via matching-adjusted indirect comparison analyses) for individual tumor types for which the treatment landscapes vary. Translational analyses will be important to further understand on-target and off-target resistance mechanisms to determine appropriate mitigation strategies, such as combination therapies or other novel TRK inhibitors. ctDNA surveillance could help identify early progression and the need for treatment escalation. Other useful studies would be the utility of TRK inhibitors in the early-stage setting and the benefit/risk in pediatric patients (<12 years) with alternative drug formulations.

Methods

Trial design and treatment

From the phase 1 component of TRIDENT-1 (NCT03093116), the recommended phase 2 dose (RP2D) of repotrectinib was determined as 160 mg once daily for 14 days, followed by 160 mg twice daily21. Phase 2 was conducted at 152 sites across 19 countries. Patients enrolled in six cohorts by tumor molecular characteristics and treatment history. Two cohorts consisted of patients with NTRK+ solid tumors and included patients without previous TRK TKI treatment (TKI-naive cohort) and patients treated with one or two previous TRK TKIs (TKI-pretreated cohort). The remaining four cohorts included patients with ROS1+ NSCLC and were previously described21. TRIDENT-1 trial design is provided in Extended Data Fig. 1; study protocol details were previously reported21.

Trial endpoints and assessments

The phase 2 primary endpoint was confirmed objective response (complete response or partial response) as assessed by BICR according to Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1. Key secondary endpoints included DOR; clinical benefit; PFS; overall survival; intracranial response in patients with measurable intracranial disease at baseline, as assessed by BICR according to modified RECIST (mRECIST) version 1.1 (ref. 30); safety as assessed with Common Terminology Criteria for Adverse Events version 4.03; and patient-reported outcomes as assessed with the EORTC QLQ-C30. Exploratory endpoints included potential prognostic utility of genomic alterations, emergence of repotrectinib resistance mutations and confirmed response by patient subgroup (demographic and baseline risk factors). Sex of patients was recorded by each study site.

Patients

Eligible patients had tumors harboring an NTRK fusion and were at least 18 years of age in phase 1 or at least 12 years of age in phase 2. NTRK fusion was determined by local tissue-based testing with retrospective confirmation at a centralized diagnostic laboratory or determined locally by fluorescence in situ hybridization (FISH) with prospective confirmation by a central diagnostic laboratory before enrollment. Patients with asymptomatic central nervous system (CNS) metastases (treated or untreated) were allowed to enroll. Efficacy population included all patients from NTRK+ cohorts pooled from phase 1 and phase 2 with at least one measurable target lesion according to RECIST version 1.1 and prospectively confirmed by BICR, who started treatment with repotrectinib at any dose by 15 February 2023, with data cutoff of 15 October 2023. Biomarker assay methods used to identify NTRK resistance mutations are in the Supplementary Methods. Safety population included all patients treated prior to data cutoff of 15 October 2023 who received any dose of repotrectinib in phase 1 or phase 2, regardless of tumor or fusion type.

Trial oversight

This trial was sponsored and designed by Turning Point Therapeutics, a wholly owned subsidiary of Bristol Myers Squibb. Input was provided by investigators with agreement to keep all trial aspects and outcomes confidential. The trial was conducted following US Food and Drug Administration regulations and the International Council for Harmonisation E6 guideline for Good Clinical Practice. Appropriate health authorities and institutional committees reviewed the protocol (Supplementary Methods)21. All patients provided written informed consent. The clinical safety committee (phase 1), the data and safety monitoring committee (phase 2) and Turning Point Therapeutics provided clinical trial oversight. All authors participated in drafting the manuscript, and medical writing support was funded by the sponsor. Authors verified data accuracy and comprehensiveness and fidelity of the trial to the protocol.

Statistical analysis

For both overall and intracranial response data, the proportion of patients with confirmed response was reported with 95% CIs, calculated using the two-sided 95% Clopper–Pearson method. Time-to-event endpoints were reported using the Kaplan–Meier method, with 95% CIs determined using the Greenwood variance estimate. CIs should not be used in place of hypothesis testing; CI widths were not adjusted for multiple comparisons. Additional details are provided in the Supplementary Methods and were reported previously21.

The protocol-specified target sample sizes for the TKI-naive and TKI-pretreated cohorts were 55 and 40, respectively (see the Supplementary Methods for details). The target sample size of 55 patients for the TKI-naive cohort was not achieved within the reported cutoff date. We report herein patient populations of 51 and 69 for the TKI-naive and TKI-pretreated cohorts, respectively, with at least 6 months of follow-up for tumor assessment after first post-baseline scan, which is consistent with the registrational dataset agreed upon with regulatory agencies.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.