Phase 1 Study of JNJ-64619178, a Protein Arginine Methyltransferase 5 Inhibitor, in Advanced Solid Tumors

Protein arginine methyltransferases (PRMT) catalyze the transfer of methyl groups to the arginine residues of protein substrates, including histones, spliceosomes, and transcription factors; and thus, are involved in numerous biological processes (1). Nine PRMTs have been identified in humans and are classified into three groups with overlapping, diverse functions. Type I PRMTs regulate asymmetric dimethylarginine, type II's regulate symmetric dimethylarginine (sDMA), and type III's regulate monomethylarginine. PRMT5, in a complex with methylosome protein 50 (MEP50), catalyzes the transfer of a methyl group from the methyl donor S-adenosylmethionine (SAM) to arginine residues on protein targets (2, 3). In mammals, PRMT5 represents the main type II arginine methyltransferase for sDMA of both histone and non-histone proteins (4, 5).

PRMT5 is overexpressed in several human cancers and has been shown to have oncogenic properties via epigenetic mechanisms (6–9). PRMT5 plays a regulatory role in the progression of human cancers by promoting the proliferation, invasion, and migration of cancer cells (10). Overexpression of PRMT5 has been demonstrated to be a poor prognostic indicator across many human cancer types (11–16). Recent studies have shown that inhibition of PRMT5 has antitumor effects in malignant cell lines and animal models of cancer across histologically diverse tumor types (17, 18). Thus, inhibition of PRMT5 is an attractive clinical target for the treatment of malignancies. PRMT5 has been shown to be a synthetic lethal target in tumors with deletion of methylthioadenosine phosphorylase (MTAP) leading to aberrant accumulation of the metabolite methylthioadenosine (MTA) in tumors, which acts as a competitive inhibitor to SAM (19–21). Additional predictors of PRMT5 inhibitor sensitivity have been reported in preclinical studies, including splicing factor mutations, DNA damage repair alterations, wild-type TP53 status, and the ratio of the PRMT5-interacting proteins pICIn and RioK1 (22–26).

In the past several years, PRMT5 inhibitors have entered the clinic in phase 1 studies, including the protein substrate-competitive inhibitors GSK3326595 and PRT543 and the SAM-competitive inhibitor PF-06939999 (27–29). More recently, trials are underway with MTA-cooperative inhibitors designed to preferentially inhibit the PRMT5–MTA complex in MTAP-deficient cells, sparing MTAP wild-type cells (30).

JNJ-64619178 is a selective, orally bioavailable inhibitor of PRMT5 that is being developed for the treatment of advanced malignant solid tumors and lymphomas (31). JNJ-64619178 occupies the SAM and the protein substrate pockets of PRMT5 resulting in extended trapping of PRMT5/MEP50 in a catalytically inactive state. JNJ-64619178 demonstrates slow off-rate binding kinetics resulting in prolonged, pseudo-irreversible inhibition of protein arginine dimethylation. This mechanism of action enables potent and long-lasting target engagement and dose-dependent cell death, as demonstrated in multiple preclinical cancer models. Pharmacological inhibition of PRMT5 by JNJ-64619178 resulted in an increase in alternatively spliced events, which correlated with drug sensitivity, in a panel of histologically diverse cancer cell lines. Most tumor types showed a sensitive subset of cell lines, without predominantly sensitive histologies. Splicing factor mutations in acute myelogenous leukemia patient samples and engineered cell lines correlated with increased sensitivity to JNJ-64619178; however, splicing factor mutations were not strongly associated with sensitivity of solid tumor xenografts. Consistent with its high-affinity binding to both the SAM and substrate pockets of PRMT5, JNJ-64619178 did not show enhanced selectivity for cell lines with MTAP deletion. Nor did in vitro cell line sensitivity significantly correlates with other preclinical predictors described above.

Preclinical in vivo antitumor efficacy in NCI-H1048 lung cancer xenografts correlated with >90% reduction in symmetrically dimethylated SmD1/3 ribonucleoproteins in tumors as well as >50% reduction of plasma sDMA. Human efficacious plasma exposures (C_trough) in the range of 6–18 ng/mL were calculated on the basis of multiple lung cancer xenograft models (32). The sustained pharmacodynamic (PD) effect on PRMT5 and its substrates provided the possibility to explore intermittent dosing to mitigate possible toxicities in the clinic. Nonclinical toxicology studies showed myelosuppression and gastrointestinal toxicities that were reversible upon cessation of dosing, consistent with PRMT5 conditional knockout phenotypes reported in the literature (33, 34).

This multi-part, first-in-human, Phase 1 study was designed to identify an MTD and a recommended Phase 2 dose (RP2D) and evaluate the safety, pharmacokinetics (PK), PD, and preliminary clinical activity of JNJ-64619178. Part 1 of the study, in patients with advanced solid tumors or relapsed or refractory B-cell non-Hodgkin lymphoma (NHL), is reported here.

This was a multi-part, first-in-human, Phase 1, open-label, multicenter study to evaluate the safety, PK, PD, and preliminary clinical activity of JNJ-64619178 monotherapy administered to adult patients with advanced malignant solid tumors or B-cell NHL who previously received or were ineligible for standard treatment options. All procedures were performed in accordance with the ethical standards of the Institutional Review Board/Independent Ethics Committee and with the 1964 Helsinki Declaration and its later amendments. Patients provided written consent to participate in the study after having been informed about its nature and purpose. The first patient was enrolled on July 24, 2018, enrollment was stopped on June 15, 2021 and the cutoff date was May 29, 2022.

Eligible patients were of ≥18 years of age and had an Eastern Cooperative Oncology Group performance status score of 0 or 1, adequate organ function, corrected QT interval ≤480 ms, and left ventricular ejection fraction within normal range. Patients were required to have: (i) relapsed or refractory non-hematologic solid tumor that is metastatic or unresectable, and previously received or were ineligible for standard treatment options, including appropriate molecularly targeted therapies; or (ii) relapsed or refractory B-cell NHL (diffuse large B-cell lymphoma, follicular lymphoma, or mantle cell lymphoma) previously treated with at least 2 prior lines of systemic therapy with at least 1 line of a standard anti-CD20 plus anthracycline containing systemic chemotherapy regimen. All patients were required to have histological documentation of disease and at least 1 measurable site of disease. Key exclusion criteria included known history of disease involvement of the central nervous system and prior systemic therapy for their cancer within 2 weeks of first dose.

JNJ-64619178 was administrated orally once daily following two schedules. For Schedule A, JNJ-64619178 was administered once a day for 14 consecutive days followed by 7 days of rest on a 21-day cycle. For Schedule B, JNJ-64619178 was administered every day on a 21-day cycle. Dose escalation was initiated at a starting dose of 0.5 mg on Schedule A and doses of 1.0, 1.5, 2.0, 3.0, and 4.0 mg were evaluated on this schedule. On Schedule B, 1.0 and 2.0 mg doses were evaluated. Patients received JNJ-64619178 until disease progression, intolerable toxicity, withdrawal of consent, or investigator decision. Safety was monitored by a study evaluation team.

The primary objective of the study was to determine the MTD and RP2D of JNJ-64619178 based on the assessment of dose-limiting toxicities (DLT) and adverse events (AE). Secondary objectives included PK, PD, and preliminary antitumor activity. Efficacy endpoints included overall response rate (ORR), clinical benefit rate, and duration of response.

Safety assessments included the incidence, severity, and type of AEs, vital sign measurements, clinical laboratory values, and electrocardiograms. The severity of AEs was graded according to the National Cancer Institute Common Terminology Criteria for AEs v. 4.0. Patients were followed for 30 days after the last dose of study drug for AE assessment.

DLTs were assessed during the first cycle (21 days). Non-hematologic toxicities of grade ≥3 except for laboratory abnormalities were considered DLT with exceptions for grade 3 asthenia, fever, or constipation lasting <7 days or grade 3 nausea, vomiting, or diarrhea lasting <7 days despite best supportive care. Laboratory abnormality criteria for DLT included chemistry abnormalities [other than alanine aminotransferase (ALT), aspartate aminotransferase (AST), bilirubin, gamma-glutamyl transpeptidase (GGT), alkaline phosphatase (ALP), lipase or amylase] grade 3 for >7 days without clinical sequelae or >3 days with clinical sequelae despite best supportive care, or grade 4; ALT, AST, or bilirubin grade ≥3 that has not returned to grade 1 or baseline within 3 days, or grade 4, or meeting Hy's law criteria; grade ≥3 lipase or amylase if associated with clinical or radiological evidence of pancreatitis. Isolated grade 3 or 4 GGT or ALP elevation associated with liver metastases was not considered DLT, nor were tumor lysis syndrome-related chemistry abnormalities that recovered to grade ≤2 within 72 hours. Hematological toxicities meeting DLT criteria included febrile neutropenia; grade 4 neutropenia for >7 days; grade 3 platelet count decreased associated with grade ≥2 bleeding or grade 4; grade 4 anemia; or any hematological toxicity of grade 5. For patients who experienced toxicity that met the criteria of DLT, administration of additional study drug was held pending management and resolution of the toxicity.

Blood samples were drawn at multiple timepoints on day 1 of the PK run-in phase following a single-dose of JNJ-64619178 and additional samples were collected during the treatment phase at steady-state on cycle 1, day 14. Plasma samples were analyzed to determine concentrations of JNJ-64619178 using a validated LC/MS-MS method. PK parameters of JNJ-64619178 after single dose and at steady-state were determined using non-compartmental analysis with validated software Phoenix WinNonlin (version 8.1, Certara L.P.) and summarized by dose level. Dose proportionality was assessed and accumulation ratios at steady-state versus day 1 were calculated.

Serial plasma concentration of sDMA was used as a PD marker to evaluate JNJ-64619178 engagement of PRMT5 and its inhibitory effect on the formation of di-methylated proteins. Blood samples were drawn at multiple timepoints on day 1 of the PK run-in phase following a single-dose of JNJ-64619178 and additional samples were collected during cycle 1, days 1 and 15 and then day 1 of subsequent cycles. Determination of sDMA in human plasma was done by LC/MS-MS at the Covance Central Laboratory in Indianapolis, IN.

Tumor response was assessed by the investigators according to the RECIST version 1.1 every 6 weeks until week 24, then every 12 weeks or according to the Lugano Classification for NHL (35, 36). ORR is defined as the proportion of patients who had a partial response or better according to the disease-specific response criteria.

Dose escalation was supported by the modified continual reassessment method, which was based on the probability of DLTs by a 2-parameter Bayesian Logistic Regression Model with escalation with overdose control principle (37, 38). The all-treated population, defined as those who received at least 1 dose of JNJ-64619178, was used for the safety and efficacy analyses. Duration of response distribution and PFS were estimated using the Kaplan–Meier method. Duration of stable disease and PFS were calculated from the start of treatment until the date of first progressive disease according to RECIST v1.1 or death due to any cause, whichever occurred first. Patients who were alive without progression were censored at the last disease evaluation before the start of subsequent therapy or at the study cutoff date. Duration of response was calculated from first onset of response until date of first progressive disease according to RECIST v1.1 or death due to any cause, whichever occurred first, with censoring as described above.

The data sharing policy of Janssen Pharmaceutical Companies of Johnson & Johnson is available at https://www.janssen.com/clinical-trials/transparency. As noted on this site, requests for access to the study data can be submitted through Yale Open Data Access (YODA) Project site at http://yoda.yale.edu.

A total of 90 patients received the study drug. The median age was 59.0 years (range, 28–82), and 51.1% were female (Table 1). Patients had received a median of 3 prior lines of systemic therapy (range, 0–11). The most common tumor types were adenoid cystic carcinoma [ACC; 26 (28.9%)], pancreatic cancer (13; 14.4%), prostate cancer 12 (13.3%), and uveal melanoma (8; 8.9%). Increased enrollment of ACC was based on an early response in this study and signals of clinical activity from another PRMT5 inhibitor trial. The enrollment of patients with uveal melanoma was based on the hypothesis that driver mutations in the splicing factor SF3B1, which are prevalent in uveal melanoma, might enrich for responsiveness to PRMT5 inhibition. Only one patient with NHL (mantle cell lymphoma) was enrolled. Representativeness of studied patients is shown in Supplementary Table S1.

The median duration of treatment was 2.0 months (range, 0.23 to 32.26). The median number of treatment cycles was 3.0 cycles (range, 1 to 40). The median dose intensity was 1.0 mg/d (range, 0.3 to 3.2). As of the data cutoff values, 78 patients (86.7%) had discontinued JNJ-64619178, and 12 patients were ongoing. The primary reasons for treatment discontinuation were progressive disease (63 patients; 70.0%) and physician decision (9 patients; 10.0%).

Cohorts evaluated in dose escalation are shown in Fig. 1. The only DLT in the study was thrombocytopenia observed in four patients. Two patients in the 4.0-mg cohort on Schedule A had DLT of grade 4 thrombocytopenia during the DLT period and a third patient had grade 4 thrombocytopenia in Cycle 2. As a result, lower dosing regimens (3.0 mg on Schedule A and 2.0 mg on Schedule B) were explored sequentially. One out of eight patients had a DLT of grade 4 thrombocytopenia in the 3.0-mg cohort on Schedule A, and one out of nine patients in the 2.0-mg cohort on Schedule B had a DLT of grade 3 thrombocytopenia lasting >7 days. None of the events was considered serious. Both patients in the 4.0-mg cohort were managed with treatment interruption followed by dose reduction, and the other patients were managed with treatment interruption.

The doses of 3.0 mg on Schedule A and 2.0 mg on Schedule B were identified as the MTDs. However, doses of 2.0 or 3.0 mg on Schedule A and 2.0 mg on Schedule B were not well tolerated over time and had a higher rate of treatment interruption and dose reduction compared with doses of 1.0 mg on both schedules (Table 2). Two provisional RP2Ds of 1.5 mg on Schedule A and 1.0 mg on Schedule B were selected to be evaluated in PK/PD cohorts based on: (i) Both doses having been hypothesized to attain efficacious plasma concentration targets calculated on the basis of preclinical models (ii), both doses having achieved robust sDMA reduction, the peripheral PD marker of PRMT5 inhibition (iii), tolerability assessment at 1.0 mg on Schedule B and 1.0 and 2.0 mg on Schedule A, 1.5 mg on Schedule A having been hypothesized to provide the highest dose on an intermittent schedule that would be tolerable for sustained treatment, which may be required for efficacy. The provisional RP2Ds allowed for evaluation of potential tolerability differences with intermittent versus continuous dosing at the same dose intensity.

Overall, 86 (95.6%) patients experienced at least 1 treatment-emergent AE (TEAE; Table 3). The most common TEAEs reported (≥20%) were nausea [42 (46.7%) patients], thrombocytopenia [40 (44.4%) patients], fatigue and anemia [38 (42.2%) patients each], dysgeusia [31 (34.4%) patients], asthenia [23 (25.6%) patients], diarrhea and decreased appetite [22 (24.4%) patients each], and vomiting [18 (20.0%) patients]. These AEs were also the most frequently reported treatment-related AEs.

Fifty (55.6%) patients experienced at least 1 TEAE of grade ≥ 3. The most commonly (≥5% patients) reported TEAEs of grade ≥3 were anemia [17 (18.9%) patients], thrombocytopenia [14 (15.6%) patients], and neutropenia [5 (5.6%) patients]. These AEs were also the most frequent treatment-related AEs (TRAE) of grade ≥ 3: Anemia [16 (17.8%) patients], thrombocytopenia [14 (15.6%) patients] and neutropenia [5 (5.6%) patients]. Thrombocytopenia was the only Grade 4 TRAE [7 (7.8%) patients]. Thrombocytopenia and anemia were the leading causes of treatment interruption [21 (23.3%) and 18 (20.0%) patients, respectively] or dose reduction [9 (10.0%) and 10 (11.1%) patients, respectively]. In the majority of cases treatment interruption for 1 week was sufficient to allow continued dosing. Twenty-six (28.9%) patients experienced serious AEs (SAE) and 2 (2.2%) patients experienced study drug-related SAEs (grade 3 nausea and grade 3 diarrhea in 1 patient each). The SAEs that were reported in more than 1 participant were pneumonia, abdominal pain, diarrhea, and dyspnea [3 (3.3%) patients each]. Eight (8.9%) patients experienced TEAEs leading to study drug discontinuation. The TEAEs leading to study drug discontinuations that were reported in more than 1 participant were weight decreased and pneumonia [2 (2.2%) participants each; Supplementary Table S2]. Forty (44.4%) patients died during the study. The reasons for deaths were progressive disease [37 (41.1%) patients], and AE and cause unknown [2 (2.2%) patients each]. One patient each experienced TEAEs leading to death of euthanasia and pneumonia. No deaths were considered treatment related.

The single-dose serial plasma PK of JNJ-64619178 was evaluated in a total of 42 patients following 0.5, 1.0, 2.0, 3.0 and 4.0 mg dosing. JNJ-64619178 mean plasma PK profiles following single oral-dose are shown in Fig. 2A. The single-dose plasma JNJ-64619178 PK parameters are summarized in Table 4. The median time to reach the maximum plasma concentration (t_max) ranged from 0.9 to 2.5 hours across all doses. The maximum plasma concentration (C_max) of JNJ-64619178 increased in an approximately dose-proportional manner whereas the AUC over 24 hours (AUC_0–24h) increased in a slightly more than dose-proportional manner over the dose range of 0.5 to 4.0 mg. JNJ-64619178 appears to be extensively distributed. The plasma concentration of JNJ-64619178 appears to decrease multiexponentially with a mean terminal half-life ranging from 64.3 to 84.1 hours across all dose levels.

The steady-state serial plasma PK of JNJ-64619178 was evaluated in 68 patients following 0.5, 1.0, 1.5, 2.0, 3.0, and 4.0 mg oral dosing on cycle 1, day 8 (0.5 mg) or day 14 (1.0, 1.5, 2.0, 3.0, and 4.0 mg). The steady-state JNJ-64619178 mean plasma PK profiles following multiple once daily oral doses are shown in Fig. 2B. Across all doses evaluated, the observed steady-state JNJ-64619178 plasma exposures were comparable with or above the calculated efficacious plasma concentration range based on preclinical models. JNJ-64619178 PK parameters after multiple once daily doses are summarized in Table 5 The t_max ranged from 1.4 to 3.2 hours across all doses. The C_max of JNJ-64619178 increased in an approximately dose-proportional manner whereas AUC_0–24h increased in a slightly more than dose-proportional manner over the dose range of 0.5 to 4.0 mg. Mean AUC_0–24h accumulation ratio between the single dose and multiple doses on day 14 ranged from 4.5 to 11.8 across the 1.0 to 4.0 mg dose range.

Serial plasma concentration of sDMA was used as a PD marker to evaluate JNJ-64619178 engagement of PRMT5 and its inhibitory effect on the formation of di-methylated proteins. Reductions of >70% from baseline were observed with doses as low as 0.5 mg, with maximum observed reduction at cycle 1, day 15 of 83.1% at the 4.0-mg dose (Fig. 2C; Supplementary Table S3). Within the dose range tested, there was a flat exposure–PD relationship. Nadir sDMA concentrations were observed for most patients between cycle 1, day 15 and cycle 2, day 15, and levels generally were maintained near these nadir levels throughout treatment.

The overall ORR according to RECIST v1.1 was 5.6%. A best response of stable disease was observed in 52.2% and progressive disease in 36.7%. Five patients (5.6%) were not evaluable for response. Maximum target lesion size reduction is shown in Fig. 3A. Partial responses were observed in 3 patients with ACC, 1 with ovarian cancer, and 1 with mucoepidermoid carcinoma, of which 4 were confirmed responses. No complete responses were observed. Responses occurred in 3 patients in the 1.0 mg cohort, 1 participant in the 2.0 mg cohort, and 1 participant in the 3.0 mg cohort, all on Schedule A. Of note, the participant at 2.0 mg received the 2.0 mg dose for 5 cycles and then the dose was reduced to 1.0 mg due to AE. One additional patient with prostate cancer had a reduction in the sum of their target lesion diameters ≥30% without meeting criteria for PR. Responses developed slowly in the ACC patients, occurring at 8.1, 11.1, and 16.6 months, whereas responses occurred more rapidly in the patients with ovarian cancer and mucoepidermoid carcinoma (Supplementary Fig. S1A). The median duration of response was 11.3 months (range, 1.5–11.3 months), with one response ongoing at data cutoff value. An exploratory exposure-response analysis based on the limited efficacy data suggested no specific differences or trends in exposure or PD effects between responders and non-responders.

A best response of stable disease lasting ≥6 months was observed in 27 patients (30.0%), of whom 19 had ACC (Supplementary Fig. S1B). Non-ACC patients with stable disease ≥6 months included patients with prostate cancer (n = 3), other salivary gland tumors (n = 2), uveal melanoma, liposarcoma, and pseudopapillary pancreatic cancer (n = 1 each). The median progression-free survival (PFS) for ACC patients was 19.1 months [95% confidence interval (CI), 11.0–27.2 months]. At the data cutoff value, 11 patients with ACC and one patient with pseudopapillary pancreatic cancer continued on treatment. Figure 3B shows the change in the target lesion sum of diameters over time for the patients with ACC. All the patients with ACC with prior systemic therapy for metastatic disease had a longer time on treatment with JNJ-64619178 than with their immediate prior therapy.

In this first-in-human, phase 1 study, we characterized the safety, PK, PD, and preliminary antitumor activity of JNJ-64619178 in patients with relapsed/refractory solid tumors and NHL.

Because of enrollment of only one patient with NHL, conclusions are limited to solid tumors. Dose escalation identified the dose of 3.0 mg on a 14 days on/7 days off schedule and 2.0 mg continuous dosing as the MTDs based on DLT of thrombocytopenia at 4.0 mg. However, dose interruptions and dose reductions due to toxicity were frequent at doses of 2.0 mg and higher, leading to further exploration of lower doses on both schedules. On the basis of the better tolerability over time, evidence of robust target engagement by sDMA reduction and objective responses at the 1.0 mg dose level, two provisional recommended Phase 2 doses (RP2Ds) were selected: 1.5 mg on the intermittent schedule and 1.0 mg the continuous dosing schedule.

The toxicities observed with JNJ-64619178 were predominantly hematologic and gastrointestinal. Fatigue, asthenia, dysgeusia, and decreased appetite were also frequently observed. Thrombocytopenia and anemia showed a trend toward increased frequency and severity with increasing dose of JNJ-64619178. Both provisional RP2Ds were tolerable, and toxicities were manageable with dose interruption or, less frequently, dose reduction. The 1.0-mg dose on the intermittent schedule was also well tolerated with prolonged dosing and demonstrated partial responses and durable disease control and could be considered as a potential RP2D for further study.

Overall, the toxicity profile was consistent with observations from phase 1 dose-escalation reports of three other PRMT5 inhibitors, GSK3326595, PF-06939999, and PRT543 (27–29). In common with JNJ-64619178, the most frequent treatment-related toxicities (>20%) with these compounds included anemia, thrombocytopenia, fatigue, and nausea. Dysgeusia was also common with GSK3326595 and PF-06939999. DLTs for PF-06939999 included thrombocytopenia, anemia, and neutropenia, and DLT for PRT543 included thrombocytopenia and fatigue. Interestingly, alopecia was frequently observed with GSK3326595 (28%) but was not reported as a frequent TRAE for the other PRMT5 inhibitors or JNJ-64619178 (7.8%). The commonly observed toxicities are consistent with the essential role of PRMT5 in sustaining adult hematopoiesis and an impact on rapidly proliferating bone marrow and gastrointestinal mucosal cells, akin to a cytotoxic chemotherapeutic agent. There does not appear to be a marked safety difference among these non-MTA-cooperative PRMT5 inhibitors with different binding mechanisms.

All patients had a decrease in plasma sDMA, a marker of target engagement, with near-maximal reduction already observed at the 0.5-mg dose level. Similar to the preclinical data showing that PRMT5 inhibitor-induced reduction of di-methylated SmD1/3 in cancer cells is a marker of target engagement but not necessarily efficacy, the sDMA reduction may be necessary, but not sufficient, for clinical activity of JNJ-64619178 based on the overall low clinical activity observed. Although limited preclinical data showed a correlation between plasma sDMA reduction and tumor di-methylated SmD1/3 reduction, a limitation of our study is that we did not measure tumor sDMA levels. However, a similar degree of plasma sDMA reduction was associated with near-complete loss of tumor sDMA immunohistochemical staining in a clinical study of GSK3326595, suggesting clinical correlation between these two PD markers of PRMT5 inhibition (27).

JNJ-64619178 demonstrated preliminary antitumor activity, including objective partial responses in patients with ACC (11.5%) and other tumor types, with most of the responders receiving a dose of 1.0 mg. However, the ORRs were low, and prolonged stable disease was mainly observed in patients with ACC. Most patients with ACC had durable disease control, with a median PFS of 19.1 months, and 11 ACC patients (42.3%) were still on treatment at the time of the data cutoff value. Objective responses are rare in ACC with any type of therapy. The response rate, duration of clinical benefit, and tolerability with JNJ-64619178 compare favorably with results from recent single-arm studies of the VEGFR kinase inhibitors lenvatinib and rivoceranib in patients with ACC. Lenvatinib demonstrated an ORR of 15.6% with a median PFS of 17.5 months (95% CI, 7.2 months-not reached; ref. 39). Rivoceranib demonstrated an ORR of 13.6% with a median time to progression of 10.8 months and median PFS of 9.0 months (40). ACC can be an indolent disease, and one must be careful about interpreting prolonged stable disease as proof of clinical benefit in non-randomized, single-arm studies. However, all patients with ACC with prior systemic therapy for metastatic disease had a longer duration of treatment on JNJ-64619178 than on their immediate prior therapy, suggesting a potential effect of JNJ-64619178 treatment rather than the natural history of the disease.

Responses in ACC (n = 2/50, 4%) were also reported in the phase 1 study of GSK3326595 (41). A preclinical study demonstrated downregulation of Myb and Notch signaling with PRMT5 inhibition in non-ACC cell line models (42). ACC is characterized by mutations or fusions in transcription factors, transcriptional co-activators/co-repressors, chromatin remodeling proteins, and histone lysine methyltransferases and demethylases, including MYB, MYBL1, NOTCH 1–4, BCOR, ARID1A, ARID1B, KDM6A, KMT2C, KMT2D, CREBBP, EP300, and RUNX1 (43). Given PRMT5’s many-faceted role in epigenetic regulation, we hypothesized that a genomic subset of ACC might be particularly responsive to PRMT5 inhibition, but we were not able to identify a significant molecular predictor of JNJ-64619178 response given the low number of responses. Overall, despite the low ORR, the tolerability and apparent durable clinical benefit of JNJ-64619178 support further investigation for ACC.

Aside from ACC, the clinical activity observed with JNJ-64619178 was low, despite nearly all patients in this dose-escalation study receiving doses at or above the provisional RP2Ds. This low ORR is consistent with preliminary reports of phase 1 dose-escalation results of several other PRMT5 inhibitors: 5.6% (3/54) for GSK3326595, 7.1% (2/28) for PF-06939999, and 2.0% (1/49) for PRT543 (27–29). We cannot rule out the possibility that higher doses of JNJ-64619178 might have resulted in increased activity, but these doses were not tolerable over multiple cycles, and there was not convincing evidence of target lesion reduction in the higher dose cohorts examined. We note that all responses were observed at lower doses, and we did not see significantly greater target engagement (sDMA reduction) as the dose was escalated. Hematologic toxicity did increase with dose, however, which raises the possibility that plasma sDMA might not be a sensitive enough marker of PRMT5 inhibition.

Because of the heterogeneity of tumor types and paucity of responses, it was not possible to identify a patient selection marker that might enrich for response to JNJ-64619178. PRMT5 regulates a number of important cellular processes and pathways. Our study cohort was not enriched for MTAP-deficient tumors, but the mechanism of JNJ-64619178 binding and preclinical data suggest that JNJ-64619178 potency is not expected to be impacted by increased MTA levels in MTAP-deficient tumors (31). Ongoing studies of MTA-cooperative PRMT5 inhibitors will reveal whether selectivity for MTAP-deficient tumors may allow for higher dose-escalation and increase the therapeutic window for PRMT5 inhibition.

Aberrant splicing in tumors, including those harboring splicing factor mutations, was also hypothesized to confer sensitivity to PRMT5 inhibition. In an exploratory analysis of archival tumor samples, we found no overall association between splicing factor gene mutations and target lesion reduction, although the low prevalence of mutations and degree of target lesion reduction limited the sensitivity of this analysis.

In conclusion, the tolerability and clinical benefit with JNJ-64619178 in ACC support further investigation. However, development of this class of PRMT5 inhibitors for advanced malignant solid tumors more broadly will require identification of robust patient-selection biomarkers or effective combination approaches.

M. Vieito reports nonfinancial support from Roche, Debiopharm, BMS, Novartis, Enterome, Mundipharma, Hutchinson Pharma, Servier, and Incyte and personal fees from Roche outside the submitted work. V. Moreno reports personal fees from BMS, Bayer, Roche, Basilea, Janssen, AstraZeneca, and Affirmed outside the submitted work; in addition, V. Moreno reports being a Principal Investigator and receiving institutional funding from AbbVie, AceaBio, Adaptimmune, ADC Therapeutics, Aduro, Agenus, Amcure, Amgen, Astellas, AstraZeneca, Bayer, Beigene, BioInvent International AB, BMS, Boehringer-Ingelheim, Boston, Celgene, Daichii Sankyo, Debiopharm, Eisai, e-Therapeutics, Exelisis, Forma Therapeutics, Genmab, GSK, Harpoon, Hutchison, Immutep, Incyte, Inovio, Iovance, Janssen, Kyowa Kirin, Lilly, Loxo, MedSir, Menarini, Merck, Merus, Millennium, MSD, Nanobiotix, Nektar, Novartis, Odonate Therapeutics, Pfizer, PharmaMar, Principia, PsiOxus, Puma, Regeneron, Rigontec, Roche, Sanofi, Sierra Oncology, Synthon, Taiho, Takeda, Tesaro, Transgene, Turning Point Therapeutics, and Upshersmith. A. Spreafico reports other support from Janssen Oncology/Johnson & Johnson during the conduct of the study as well as other support from Novartis, Merck, Symphogen, AstraZeneca/Medimmune, Bayer, Surface Oncology, Janssen Oncology/Johnson & Johnson, Roche, Regeneron, Alkermes, Array Biopharma/Pfizer, GSK, Treadwell, Amgen, ALX Oncology, Genentech, Seagen, and Servier and personal fees and other support from BMS outside the submitted work. I. Brana reports grants from Janssen Oncology, Cellex Foundation, and La Caixa Foundation during the conduct of the study; as well as grants from AstraZeneca, Bicycle Therapeutics, Celgene, Dragonfly, Hookipa, GlaxoSmithKline, Gliknik, ISA Pharmaceuticals, Kura, Merck Serono, Merck Sharp & Dohme (MSD), Nanobiotics, Novartis, Northern Biologics, Odonate Therapeutics, Regeneron, Pfizer, Sanofi, PharmaMar, Seattle Genetics, Shattuck Labs, and VCN Biosciences and grants and personal fees from Boehringer Ingelheim, Bristol-Myers Squibb, Immutep, Achilles Therapeutics, Cancer Expert Now, and Pierre Fabre outside the submitted work. J.S. Wang reports other support from Janssen during the conduct of the study as well as other support from AstraZeneca and Eisai outside the submitted work; in addition, J.S. Wang reports research funding to institution only from 7,8 Pharma, Accutar, Adagene, Artios Pharma, Astellas, AstraZeneca, Bayer Healthcare, Beigene, Bicycle Therapeutics, BioNTech, BioTheryX, Biosplice, Blueprint, Boehringer Ingelheim, Celgene/BMS, Clovis Pharma, Cullinan, Cyteir, Daiichi Sankyo, Erasca, Forty Seven, Genentech/Roche, GlaxoSmithKline, H3 Biomedicine, Hotspot, Hutchinson MediPharma, IGM Biosciences, Immuno-Gen, ImmunoOnc, Jazz Pharma, Klus Pharma, Kymab, LSK BioPartners, MabSpace, Macrogenics, Medikine, Moderna, NGMBio, Novartis, Nurix, ORIC, Olema Therapeutics, Phoenix Molecular Designs, Pionyr, Prelude Therapeutics, PureTech Health, Pyxis, Qilu Pugent Sound, Relay Therapeutics, Revolution Medicines, Ribon Sanofi, StingThera, Syndax, Teneobio, TopAlliance, Treadwell Therapeutics, Xencor, and Zymeworks. A.R. Hansen reports other support from Janssen during the conduct of the study as well as personal fees and other support from Merck, Eisai, and GSK and other support from BMS, Roche, AstraZeneca, Boehringer Ingelheim, Astellas, and Genentech outside the submitted work. L. Lenox reports other support from Janssen outside the submitted work. R.J. Brown reports other support from Janssen R&D during the conduct of the study as well as other support from Johnson & Johnson outside the submitted work. A. Kalota reports personal fees from Janssen Pharmaceuticals during the conduct of the study. F. Pastore reports employment by Janssen (Clinical Research Hematology, Global Research and Development). B. Patel reports other support from J&J during the conduct of the study as well as other support from J&J outside the submitted work. J. Lauring reports personal fees from Janssen Research and Development outside the submitted work; in addition, J. Lauring reports a patent 11571437 issued. M.R. Patel reports other support from Janssen during the conduct of the study. No disclosures were reported by the other authors.

M. Vieito: Conceptualization, methodology, writing–original draft, writing–review and editing. V. Moreno: Writing–original draft. A. Spreafico: Writing–original draft. I. Brana: Writing–original draft. J.S. Wang: Writing–original draft. M. Preis: Writing–original draft. T. Hernández: Writing–original draft. S. Genta: Writing–original draft. A.R. Hansen: Writing–original draft. B. Doger: Writing–original draft. V. Galvao: Writing–original draft. L. Lenox: Writing–original draft. R.J. Brown: Writing–original draft. A. Kalota: Conceptualization, resources, data curation, formal analysis. J. Mehta: Writing–original draft. F. Pastore: Writing–original draft. B. Patel: Writing–original draft. P. Mistry: Writing–original draft. J. Gu: Writing–original draft. J. Lauring: Writing–original draft. M.R. Patel: Writing–original draft.

The authors would like to thank the patients and their families, the investigators and site staff, and the Janssen team for supporting this study. Writing and editorial support was provided by Colleen Elliott, PhD, of CME Science Writers, LLC and funded by Janssen Pharmaceuticals. Support for clinical data outputs was provided by Jeffrey Haslam and Veena Puttur Chandra (Janssen R&D).

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.