Purpose:

STM 434 is a soluble receptor ligand trap targeting activin A, a protein in the TGFβ family that plays important roles in growth, differentiation, and cancer cachexia. This study evaluated the safety, antitumor activity, and metabolic effects of STM 434 in a first-in-human, multicenter, phase I clinical trial (NCT02262455).

Patients and Methods:

Patients with advanced solid tumors were enrolled in 8 dose cohorts ranging from 0.25 mg/kg every 4 weeks to 8 mg/kg every 2 weeks via a 3 + 3 dose-escalation design. The primary endpoint was maximum tolerated dose (MTD). Secondary endpoints included safety, pharmacokinetics, and response. As activin A is implicated in metabolism and muscle function, changes in key metabolic parameters, including lean body mass and 6-minute walk test, were serially measured.

Results:

Thirty-two patients were treated on study. The most common treatment-related adverse events were fatigue (41%) and mucocutaneous bleeding complications including epistaxis (34%) and gingival bleeding (22%), likely related to off-target inhibition of bone morphogenetic protein 9 (BMP9). STM 434 treatment resulted in the expected follicle-stimulating hormone level decreases in most patients and in metabolic parameter changes, including an increase in total lean body mass and 6-minute walk test distance. No responses were observed in the 30 evaluable patients, but the stable disease rate in patients with granulosa cell ovarian cancer was 10 of 12 (80%).

Conclusions:

Although no direct antitumor efficacy was documented, potentially clinically meaningful dose-related metabolic effects, including treatment of cancer cachexia, were observed that support further exploration of activin A inhibitors that limit BMP9 blockade.

See related commentary by Bonilla and Oza, p. 5432

This article is featured in Highlights of This Issue, p. 5429

Translational Relevance

This is the first-in-human clinical trial of STM 434, a soluble receptor ligand trap targeting activin A, which plays an important role in growth, differentiation, and cancer cachexia. Although treatment with STM 434 did not result in clinical response, predicted metabolic effects of activin A inhibition, including increase in total lean body mass and 6-minute walk test, were observed. This provides a rationale for further exploration of activin A inhibitors as metabolic modulators in other clinical settings, including improving muscle wasting and cancer cachexia.

Activins are members of the TGFβ superfamily, which also includes TGFβ, inhibins, nodals, bone morphogenetic proteins (BMP), myostatin, and growth and differentiation factors (GDF). The activin family, in particular, is believed to play an important role in regulating growth of both normal tissues and tumors (1, 2). Activin A, which signals through the high-affinity activin type 2B receptor (ActR2B or ACVR2B), leads to downstream SMAD signaling and modulation of gene transcription. Activin A expression is tightly regulated through numerous mechanisms, including by endogenous inhibitors inhibin and follistatin, which inhibit activin A by blocking access to the activin receptor–binding site (1).

Overexpression of activin A has been associated with shortened survival in ovarian, colon, gastric, breast, non–small cell lung cancers, as well as neuroblastoma (3). Likewise, functional experiments have shown that activin A promotes ovarian tumor proliferation and invasion (4). Experimental data also support a role for activin signaling in mediating cancer stem cell survival (5, 6).

Beyond exerting a direct role in modulating cancer growth, activins and other ligands of ActR2B, including myostatin, are also believed to play an important role in skeletal muscle homeostasis (7–9). In animal models, overexpression of activin A or deficiency of its counterregulator, inhibin, induces a profound cachexia syndrome (10). Similarly, perturbations in activin expression are believed to play a key role in the wasting syndromes observed in several human cancers (11).

Activin A overexpression has been best studied in serous, clear cell, and sex cord stromal (granulosa cell) ovarian cancers. In these tumor types, malignant epithelial cells are believed to be the primary source of activin A production. Secreted activin A can be detected in both the serum and body fluid of affected patients (12–14). Consistent with the hypothesis that activin biology is important in gynecologic cancer, germline mutations in genes encoding activin and inhibin subunits (INHBA and INHA) have been described in women with young onset ovarian cancers, raising the question that in some settings, alteration of this pathway may play a role in tumorigenesis and inherited mutations may confer cancer susceptibility risk (15). These mutations are believed to disrupt activin and inhibin production, altering the ratio of activin and inhibin in favor of epithelial ovarian cell growth.

Additional data suggest that mutations more commonly associated with distinct ovarian cancer histologies including BRCA1 (serous), ARID1A (endometriosis associated clear cell and endometrioid), and FOXL2 (granulosa cell) may promote dysregulation in activin signaling and may further enhance dependence on this pathway (16, 17). For example, downregulation of BRCA1 indirectly enhances activin signaling through downregulation of the activin antagonist follistatin alongside upregulation of ActR2B, with the net effect of establishing an autocrine loop feeding cell proliferation (16). Moreover, the tumor suppressor activity of ARID1A is mediated through SMAD3, a key transducer of activin signaling (17, 18). Similarly, adult granulosa cell tumors (GCT) harbor pathognomonic FOXL2 C134W mutations that drive tumorigenesis, in part, through upregulation of ActR2B pathway signaling (19–23). In addition, FOXL2 C134W inhibits activin A transcriptional activation of follistatin, the β-subunit of follicle-stimulating hormone (FSH). Reduced follistatin in granulosa cells expressing FOXL2 C134W may contribute to tumor formation as a direct result of the loss of activin antagonism (24).

The importance of activin to a wide variety of human physiologic mechanisms makes it an important drug target. Given the potential multifaceted role of activin A signaling in tumor proliferation, cancer stem cell survival, and cancer cachexia, STM 434, a soluble receptor-Fc fusion protein targeting human activin A, was developed. STM 434 is a ligand trap that binds activin A with high potency, thereby inhibiting activin A signaling. STM 434 also binds and neutralizes other ligands of the endogenous ActR2B receptor, including myostatin, which are implicated in cancer cachexia and skeletal muscle wasting. Here, we describe the first-in-human phase I experience of STM 434 in patients with granulosa cell ovarian cancer and other advanced solid tumors.

Study oversight

This was an open-label, multicenter, single-arm, first-in-human phase I study. The study was conducted in accordance with the Declaration of Helsinki and the principles of Good Clinical Practice. The protocol was approved by institutional review boards within each institution, and all patients provided written informed consent before undergoing any study procedures. The study was designed by the lead investigators and the sponsor, Atara Biotherapeutics. The sponsor collected and analyzed the data in conjunction with the authors.

Study design and endpoints

The primary endpoint of the study was to define the maximum tolerated dose (MTD) of STM 434 injection administered as monotherapy in subjects with ovarian cancer or other advanced solid tumors. Secondary endpoints included evaluating the incidence of adverse events (AE) as assessed according to the Common Terminology Criteria for Adverse Events (CTCAE) v4.03 (NCI CTCAE v4.03, 2010), pharmacokinetics, overall response rate (ORR), and progression-free survival (PFS). Exploratory objectives included measuring changes in multiple metabolic parameters including change from baseline in lean body mass and 6-minute walk test distance (6MWD).

Eligible patients were age ≥18 years with Eastern Cooperative Oncology Group (ECOG) performance status score 0 to 1 and measurable disease using RECIST v1.1 (25). Patients were required to have recurrent metastatic or locally advanced disease considered refractory or intolerant to standard treatment, tumors for which no standard treatment was available, or serous ovarian, fallopian tube, primary peritoneal cancers, GCTs, or clear cell tumors considered platinum refractory/resistant or platinum intolerant.

Treatment plan

The study incorporated a 3 + 3 dose-escalation design to define an MTD with intravenous STM 434 and evaluate the safety, tolerability, and pharmacokinetic profiles of intravenous STM 434 at 5 planned dose levels (0.25, 0.5, 1, 2, and 4 mg/kg i.v.). STM 434 was initially administered intravenously every 4 weeks. After pharmacokinetic data from the first dosing cohort (0.25 mg/kg) showed a shorter than anticipated half-life, the dosing schedule was subsequently modified to a once every 2-week interval. In addition to the planned doses, the protocol was amended to study a dose of 8 mg/kg every 2 weeks, due to lower than anticipated exposures predicted to yield anticatabolic effects and to further investigate the potential for antitumor activity. Patients were treated with STM 434 until disease progression based on radiographic measurements according to RECIST v1.1, unacceptable toxicity, or withdrawal of consent.

Dose-limiting toxicities (DLT) were defined as any grade ≥3 (according to CTCAE v4.03) nonhematologic toxicity (excluding toxicities which could be definitively attributed to other underlying disease), any grade ≥4 hematologic toxicity lasting 7 days, febrile neutropenia, or grade 3 thrombocytopenia with active bleeding. The DLTs monitoring period included the first 28 days of treatment, regardless of dosing schedule. At least 3 patients were required to be evaluable for DLTs at each dose level prior to further escalation. “Backfill” slots were also permitted at doses that had been declared safe, whereas newer doses levels were under evaluation. Dose escalation continued until the MTD was identified on the basis of all available safety data. Intrasubject dose escalation, to the next highest dosing level previously declared as safe, was permitted provided the patient had not had a DLT or evidence of radiographic disease progression.

Assessments

Response assessments were conducted according to RECIST v1.1 with tumor measurements obtained by CT and/or MRI. Assessments were performed at baseline and then approximately every 8 weeks. Duration of therapy was recorded from the day of the first dose to the last dose of STM 434 administered. Type and duration of prior oncologic therapies were captured in the study database. Tumor markers, as applicable, were measured at baseline and then every other cycle. Body composition (lean body mass) was measured by dual energy absorptiometry (DXA) scan at baseline and then every other cycle. SIx-minute walk tests (6MWT), a measure of muscle function, were performed at baseline and then with each subsequent cycle (26). Serum tumor biomarkers, activin A, and FSH, were measured every cycle, and laboratory parameters, including complete blood count and comprehensive metabolic profile, were measured at baseline and with each cycle. ECOG performance status was assessed at baseline, with each cycle, and at end of study. Safety was monitored continuously throughout the study period.

Efficacy and safety analyses

PFS was estimated using the Kaplan–Meier methods. Ninety-five percent confidence intervals were calculated using the binomial exact test. Safety analyses were descriptive. The safety analysis population included all subjects who received at least 1 dose of STM 434. The efficacy analysis population included all subjects who received at least 1 cycle of therapy (corresponding to either 1 or 2 doses of STM 434, depending on whether they were enrolled to an every 4 weeks or every 2 weeks dosing cohort, respectively).

Pharmacokinetic analyses

Pharmacokinetic parameters, terminal half-life, clearance, mean residence time, volume of distribution, and volume of steady state were estimated for patients according to a predefined pharmacokinetic schedule. Pharmacokinetic analyses were summarized using descriptive statistics.

Patient characteristics

Between October 17, 2014 and January 13, 2017, a total of 32 subjects were enrolled to one of 8 dosing cohorts (Fig. 1). Baseline demographics and disease characteristics are presented in Table 1. The 32 subjects enrolled in the study were a median of 61 years of age (range: 41–79 years). Primary cancer types included 20 patients with ovarian cancer (13 granulosa, 7 epithelial), 3 with colorectal, 2 with pancreas, and 7 with other solid tumors. Patients had received a median of 3 prior lines of therapy (range: 1–11).

Figure 1.

STM 434: Dose-escalation cohorts (3 + 3 design). Schema depicting the dose-escalation design of the first-in-human, phase I clinical trial of STM 434.

Figure 1.

STM 434: Dose-escalation cohorts (3 + 3 design). Schema depicting the dose-escalation design of the first-in-human, phase I clinical trial of STM 434.

Close modal
Table 1.

Patient baseline characteristics

Baseline characteristicAll patients (n = 32)
Age at screening 
 Median (range), years 61 (41–79) 
  <18 0 (0) 
  ≥18–65 22 (69) 
  ≥65–85 10 (31) 
  ≥85 0 (0) 
ECOG performance status 0–1 [n (%)] 32 (100) 
Prior therapies [median (range)] 3 (1–11) 
Sex and child-bearing potentiala [n (%)] 
 Male 6 (19) 
 Female 26 (81) 
 Child-bearing potential 0 (0) 
 Postmenopausal 10 (38) 
 Surgically sterile 18 (69) 
Ethnicity, n (%) 
 Hispanic or Latino 7 (22) 
 Non-Hispanic or Latino 24 (75) 
 Not reported 1 (3) 
Racea, n (%) 
 White 28 (88) 
 Black/African American 2 (6) 
 Asian 1 (3) 
 Not reported 1 (3) 
Tumor type, n (%) 
 Ovarian granulosa cell 13 (41) 
 Epithelial ovarian 7 (22) 
 Colon 3 (9) 
 Pancreas 2 (6) 
 Other solid tumorb 7 (22) 
Baseline characteristicAll patients (n = 32)
Age at screening 
 Median (range), years 61 (41–79) 
  <18 0 (0) 
  ≥18–65 22 (69) 
  ≥65–85 10 (31) 
  ≥85 0 (0) 
ECOG performance status 0–1 [n (%)] 32 (100) 
Prior therapies [median (range)] 3 (1–11) 
Sex and child-bearing potentiala [n (%)] 
 Male 6 (19) 
 Female 26 (81) 
 Child-bearing potential 0 (0) 
 Postmenopausal 10 (38) 
 Surgically sterile 18 (69) 
Ethnicity, n (%) 
 Hispanic or Latino 7 (22) 
 Non-Hispanic or Latino 24 (75) 
 Not reported 1 (3) 
Racea, n (%) 
 White 28 (88) 
 Black/African American 2 (6) 
 Asian 1 (3) 
 Not reported 1 (3) 
Tumor type, n (%) 
 Ovarian granulosa cell 13 (41) 
 Epithelial ovarian 7 (22) 
 Colon 3 (9) 
 Pancreas 2 (6) 
 Other solid tumorb 7 (22) 

NOTE: The clinical and pathologic features of the 32 patients with ovarian granulosa cell tumors and other advanced solid tumors who received STM 434 are summarized. Data are presented as N (%) unless indicated otherwise.

aMore than one response can be provided for race and child-bearing potential and, as such, the percentage may total more than 100%.

bSquamous cell carcinoma (2), chondrosarcoma (1), urachal (1), leiomyosarcoma (1), renal cell carcinoma (1), and thymic carcinoma (1).

Safety and tolerability

A total of 32 patients were evaluable for safety. The most common treatment-emergent AEs (TE-AE), regardless of attribution, was fatigue, occurring in 13 (41%) of subjects at grade 1 and 2 severity. Epistaxis occurred in 11 (34%) of subjects and in 1 subject (8 mg/kg every two weeks cohort) was grade 3 in severity. Gingival bleeding, peripheral edema, and headache each occurred in 7 subjects each (22%) but only at grade 1 and 2. Grade 1 and 2 TE-AEs occurring at >10% frequency and all grade 3 TE-AEs, regardless of frequency, are summarized in Table 2. No treatment-related grade 4 AEs were observed, and no deaths occurred on study or within 30 days after the last dose of STM 434.

Table 2.

Grade 1/2 TE-AEs occurring in >10% of subjects and grade 3 TE-AEs regardless of frequency

AE, n (%)Grade 1/2Grade 3All grades (n = 32)
Fatigue 13 (41) 13 (41) 
Epistaxis 10 (31) 1 (3)a 11 (34) 
Gingival bleeding 7 (22) 7 (22) 
Headache 7 (22) 7 (22) 
Peripheral edema 7 (22) 7 (22) 
Abdominal pain 4 (13) 2 (6)b 6 (19) 
Diarrhea 6 (19) 6 (19) 
Pyrexia 6 (19) 6 (19) 
Rash, maculo-papular 6 (19) 6 (19) 
Anemia 3 (9) 2 (6)c 5 (16) 
Decreased appetite 4 (13) 4 (13) 
Dyspnea 4 (13) 4 (13) 
Nausea 4 (13) 4 (13) 
AE, n (%)Grade 1/2Grade 3All grades (n = 32)
Fatigue 13 (41) 13 (41) 
Epistaxis 10 (31) 1 (3)a 11 (34) 
Gingival bleeding 7 (22) 7 (22) 
Headache 7 (22) 7 (22) 
Peripheral edema 7 (22) 7 (22) 
Abdominal pain 4 (13) 2 (6)b 6 (19) 
Diarrhea 6 (19) 6 (19) 
Pyrexia 6 (19) 6 (19) 
Rash, maculo-papular 6 (19) 6 (19) 
Anemia 3 (9) 2 (6)c 5 (16) 
Decreased appetite 4 (13) 4 (13) 
Dyspnea 4 (13) 4 (13) 
Nausea 4 (13) 4 (13) 

NOTE: Listed above are AEs reported in at least 10% of the patients (n = 32) with advanced solid tumors who received STM 434 and that were deemed by the investigators to be related to study drug. No treatment-emergent grade 4 or 5 events were reported.

a8 mg/kg every 2 weeks cohort.

b0.25 mg/kg every 4 weeks and 0.5 mg/kg every 2 weeks cohorts.

c2.0 mg/kg every 2 weeks cohort.

Nine patients (28%) experienced serious AEs (SAE), with the following SAEs occurring in 1 subject each (a single subject may have had more than 1 event): abdominal abscess, abdominal pain, anemia, ascites, acute cholecystitis, dehydration, epistaxis, gastritis, hemorrhagic ascites, infection, pain, and pyrexia. Of these 9 SAEs, 4 were deemed related to STM 434: anemia and epistaxis (8 mg/kg every 2 weeks cohort), hemorrhagic ascites (0.5 mg/kg every 4 weeks cohort), and gastritis (8 mg/kg every 2 weeks cohort).

Three patients withdrew from STM 434 treatment as a result of at least possibly STM 434–related AEs including: grade 2 dyspnea on exertion in the 0.5 mg/kg every 4 weeks cohort, grade 3 hemorrhagic ascites in the 0.5 mg/kg every 4 weeks cohort, and grade 1 abdominal wall hemorrhage in the 4 mg/kg every 2 weeks cohort. In some cases, the observed bleeding diathesis persisted for weeks after discontinuation of STM 434.

DLT and MTD

A total of 5 DLTs (all grade 3) were observed in 3 patients: (i) hemorrhagic ascites lasting 2 months (0.5 mg/kg every 4 weeks), (ii) stomatitis and gastritis (8 mg/kg every 2 weeks), and (iii) epistaxis and increased alkaline phosphatase (8 mg/kg every 2 weeks). As 2 subjects experienced DLTs in the 8 mg/kg every 2 weeks cohort, this was considered to have exceeded the MTD and 4 mg/kg every 2 weeks was formally determined to be the MTD. However, given the overall safety profile of the agent with bleeding complications observed at dose levels as low as 0.5 mg/kg every 2 weeks, the safety review committee recommended against proceeding with dose expansion.

Pharmacokinetic analysis

The pharmacokinetic evaluable population consisted of 31 subjects. Data from cohort 1 (0.25 mg/kg every 4 weeks) showed a half-life and clearance that was 2-fold higher in humans compared with observations in rodent and monkey preclinical studies. Therefore, based on the half-life of 5.0 days, the safety committee determined that every 2 weeks dosing should be administered. Further pharmacokinetic data from cohort 2 (0.5 mg/kg every 4 weeks) confirmed the half-life observations from cohort 1. Therefore, subsequent cohorts were administered at an every 2 weeks schedule to optimize drug exposure. STM 434 exhibited approximately linear pharmacokinetics over a dose range of 0.25 mg/kg to 8 mg/kg with little intersubject variability (Fig. 2). Across dose cohorts, the half-life of STM 434 in cycle 1 ranged from 5.0 days in the 0.25 mg/kg every 4 weeks, 0.5 mg/kg every 2 weeks, and 2 mg/kg every 2 weeks cohorts) to 7.6 days in the 8 mg/kg every 2 weeks cohort. Mean clearance in cycle 1 was 0.3 mg/kg in all cohorts except for the 0.25 mg/kg every 4 weeks, 0.5 mg/kg every 4 weeks, and 0.5 mg/kg every 2 weeks cohorts, in which it was 0.4 mg/hour/kg and the 2 mg/kg every 2 weeks cohort, which was 0.5 mg/hour/kg.

Figure 2.

Pharmacokinetics of STM 434. Exposure to STM 434, as measured by the maximum concentration (Cmax) and AUC 0–336 hours and AUC 320–1,448 hours, appears to increase in a linear manner with doses from 0.25 to 8 mg/kg.

Figure 2.

Pharmacokinetics of STM 434. Exposure to STM 434, as measured by the maximum concentration (Cmax) and AUC 0–336 hours and AUC 320–1,448 hours, appears to increase in a linear manner with doses from 0.25 to 8 mg/kg.

Close modal

Exposure to STM 434, as measured by the maximum concentration (Cmax) and area under the concentration–time curve from 0 to 336 hours (AUC 0–336 hours), appeared to increase in a linear manner with doses from 0.25 to 8 mg/kg (Fig. 2). In the 0.25 mg/kg every 4 weeks cohort, the mean Cmax in cycle 1 was 7.3 μg/mL and increased with dose and frequency to 149.7 mg/mL in the 8 mg/kg every 2 weeks cohort. The mean AUC 0–336 hours in cycle 1 increased approximately linearly from 577.5 (179.04) hour · μg/mL in the 0.25 mg/kg every 4 weeks cohort to 18,566.7 (1,960.44) hour · μg/mL in the 8 mg/kg every 2 weeks cohort. The subjects who experienced DLTs were not outliers for pharmacokinetic exposure. For the study population as a whole, subjects had a mean (SD) of 8.0 (8.72) weeks of therapy with STM 434, entered a mean (SD) of 2.8 (2.03) cycles, and had a mean (SD) cumulative drug exposure of 7.8 (8.59) mL (Supplementary Table S1).

Response, duration, and discontinuation of therapy

The efficacy population, defined as having received a minimum of 1 cycle of STM 434, consisted of 30 patients. Sixteen (53.3%) patients experienced a best response of stable disease and 12 (40%) had progressive disease according to RECIST. Two (6.7%) patients were nonevaluable by RECIST and discontinued prior to their first response assessment due to clinical deterioration and were counted as nonresponders. The best percent change in target lesions observed in each patient is shown in Fig. 3A. Duration of therapy is shown for all 32 patients enrolled on the study (Fig. 3B). Among the 30 patients in the efficacy population, the median duration of therapy was 6 weeks (range: 2–48). One patient, with thymic cancer enrolled in the 0.5 mg/kg every 4 weeks cohort, continued on study with a best response of stable disease for 12 cycles (48 weeks). Specifically, among the patients with GCTs, 10 of 12 achieved a best response of stable disease. However, the median duration of therapy for these 10 patients was only 7.1 weeks (range: 2.3–22.1). The median ratio of duration of STM 434 therapy to duration of prior therapy was 1.08. Seven patients discontinued therapy due to either disease progression of lack of clinical benefit as judged by the principal investigator, 2 discontinued due to an AE, and 1 patient withdrew consent.

Figure 3.

Greatest change in target lesions and duration of therapy. A, Waterfall plot showing the greatest change in target lesions by RECIST v1.1 criteria. Only patients with baseline and at least one postbaseline assessment of target lesion are included. Two patients in the efficacy population are not included in this figure. Colors (shown in legend) represent the different dose cohorts. Asterisks indicate that the tumor type is granulosa ovarian. B, Swimmer plot showing duration of therapy for the 32 patients who received STM 434.

Figure 3.

Greatest change in target lesions and duration of therapy. A, Waterfall plot showing the greatest change in target lesions by RECIST v1.1 criteria. Only patients with baseline and at least one postbaseline assessment of target lesion are included. Two patients in the efficacy population are not included in this figure. Colors (shown in legend) represent the different dose cohorts. Asterisks indicate that the tumor type is granulosa ovarian. B, Swimmer plot showing duration of therapy for the 32 patients who received STM 434.

Close modal

Pharmacodynamics and metabolic response to therapy

Activins function biologically to enhance biosynthesis and secretion of FSH. Therefore, to confirm on-target and pathway inhibition by STM 434, FSH levels were measured at baseline (pretreatment) and at defined intervals with each cycle. As expected, STM 434 induced a reciprocal decrease in FSH in most subjects. FSH decreased by a mean (SD) of 8.1 (11.45) mIU/mL for all cohorts (n = 27) from baseline to the last available postbaseline observation. The change in FSH for each cohort is shown in Fig. 4A.

Figure 4.

Pharmacodynamic and metabolic response to therapy. A, Mean absolute change in FSH compared with baseline. B, Mean percentage change in total lean body mass (LBM) compared with baseline. C, Mean absolute change in 6MWD in meters.

Figure 4.

Pharmacodynamic and metabolic response to therapy. A, Mean absolute change in FSH compared with baseline. B, Mean percentage change in total lean body mass (LBM) compared with baseline. C, Mean absolute change in 6MWD in meters.

Close modal

Other metabolic parameters of activin A inhibition were also measured. As activin A overexpression is implicated in muscle wasting and cancer cachexia, lean body mass and six-minute walk distance were also assessed. Lean body mass increased by a mean percent (SD) of 3.00 (10.4; n = 21; Fig. 4B). The 6-minute walk distance increased by a mean (SD) of 20.0 (60.2) meters (n = 22; Fig. 4C).

To explore whether the changes observed in lean body mass and 6-minute walk distance were reflected in other biologic and functional parameters of performance, we also examined changes in laboratory parameters (Supplementary Table S2) and ECOG performance status (Supplementary Table S3). The mean (SD) changes from baseline to end of treatment for albumin, hemoglobin, and absolute lymphocyte count were −0.2 g/dL (0.2), −0.8 g/dL (1.1), and 188.3/μL (979.5), respectively. In terms of ECOG performance status, at baseline, 23 of 32 patients had an ECOG score of 0, with the remaining 9 patients with an ECOG score of 1. At the end of study, ECOG score was only available on 13 patients, of which 8 patients had ECOG score of 0 and 5 patients had an ECOG score of 1. Among these 13 patients who had both baseline and end of study ECOG scores, 10 patients had stable ECOG scores at baseline to end of treatment, 1 patient's ECOG score improved from 1 to 0, and 2 patients' ECOG score worsened from 0 to 1.

Consistent with the hypothesis that STM 434 may have the potential to alter fatty metabolism, one patient with ovarian granulosa cell cancer enrolled in the 0.75 mg/kg every 2 weeks cohort was noted to have complete radiographic resolution of hepatic steatosis previously observed on her baseline CT scan after 2 months of treatment (Supplementary Fig. S1). No other changes in her clinical status, dietary behavior, or concomitant medications could otherwise explain this change.

We report results from a first-in-human, first-in-class study evaluating the safety, pharmacokinetics, pharmacodynamics, and preliminary efficacy of STM 434 in 32 patients with advanced solid tumors, including 12 patients with ovarian GCTs. Multiple episodes of mucocutaneous bleeding syndrome mimicking hereditary hemorrhagic telangiectasia occurred at several dose levels among patients treated with STM 434. This safety signal was ultimately hypothesized to be related to off-target inhibition of BMP9, a related ligand of ActR2B, the target of STM 434. Preclinical toxicity studies had identified bleeding diathesis as one potential risk of STM 434. Specifically, in a study of cynomolgus monkeys of the STM 434 surrogate STM 217, 3 of 7 monkeys treated with 3 mg/kg and 2 of 6 monkeys treated with 10 mg/kg developed red muzzle discharge consistent with epistaxis, compared with 0 of 7 control monkeys. One monkey in the 10 mg/kg group also developed significant bleeding from an ischial pad lesion 7 weeks after starting treatment. Platelet counts were normal in the monkeys, and it was hypothesized that the bleeding was the result of inhibition of bone morphogenetic protein 9 (BMP9), another member of the TGFβ superfamily that serves as a ligand for the activin receptor-like kinase-1 (ACVRL1 or ALK1). ACVRL1 bears significant homology to ActR2B, the target of STM 434. Mutations in activin receptor-like kinase-1 (ACVRL1) have been reported in hereditary hemorrhagic telangiectasia (HHT), which is associated with a similar spectrum of mucocutaneous bleeding disorders as well as capillary telangiectasias (27). Interestingly, 1 patient with ovarian GCT treated at the 0.75 mg/kg every 2 weeks dose developed diffuse telangiectasias on the upper chest, back, and face (Supplementary Fig. S2). The telangiectasias resolved 3 months following cessation of STM 434.

No definitive evidence of antitumor activity was observed with STM 434, including in patients with GCT in which activin A biology was expected to play a significant biological role. Because of the overall safety profile of STM 434 as well as the absence of clear efficacy, the decision was made to discontinue further development of STM 434.

Nevertheless, STM 434 demonstrated the expected on-target pathway inhibition, with a reciprocal decrease in FSH for most patients. In addition, treatment with STM 434 led to increased lean body mass and in most cohorts, an increase in 6MWT distance. Most patients experienced stable ECOG performance status at baseline and at the end of study. Other metabolic effects that implicate activins' potential role in regulating fat metabolism were observed. This included the complete radiographic resolution of fatty liver after 2 months of STM 434 therapy in 1 patient with ovarian granulosa cell cancer. These metabolic effects of activin inhibition, specifically the potential to increase lean body and physical stamina, may have an important role in supportive care, including in preventing and diminishing cancer cachexia. Although the toxicity observed with this particular agent render it unacceptable for use in cancer cachexia reversal, these data suggest that activin pathway inhibitors may play a potentially important supportive care role in oncology and provide a better-informed opportunity to develop a second-generation agent. However, it will be critical that future drugs targeting this pathway avoid off-target BMP9 inhibition and the bleeding diathesis it appears to induce.

In summary, we report data from the first in human, phase I dose-escalation study evaluating STM 434, an activin inhibitor, as treatment for patients with advanced cancers. Although STM 434 did not lead to tumor responses, metabolic effects that may have a future role in treating cancer cachexia were observed.

V. Makker reports receiving other commercial research support from Eli Lilly, AstraZeneca, Takeda, Karyopharm, Eisai, Merck, and IBM, and is a consultant/advisory board member for Tesaro, Eisai, Merck, Karyopharm, Takeda, and Arqule. J.F. Liu is a consultant/advisory board member for Tesaro, AstraZeneca, Clovis, and Mersana Therapeutics. D.W. Rasco reports receiving other commercial research support from Atara Biotherapeutics. W.H. Navarro and C.M. Haqq hold ownership interest (including patents) in Atara Biotherapeutics. D.M. Hyman reports receiving commercial research grants from Loxo Oncology, PUMA Biotechnology, AstraZeneca, and Bayer, and is a consultant/advisory board member for Chugai Pharma, CytomX Therapeutics, Boehringer Ingelheim, AstraZeneca, Pfizer, Bayer, and Genentech/Roche. No potential conflicts of interest were disclosed by the other authors.

Conception and design: J.J. Tao, V. Makker, C.M. Haqq, D.M. Hyman

Development of methodology: J.J. Tao

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J.J. Tao, N.A. Cangemi, K.A. Cadoo, J.F. Liu, D.W. Rasco, D.M. Hyman

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): J.J. Tao, N.A. Cangemi, V. Makker, W.H. Navarro, D.M. Hyman

Writing, review, and/or revision of the manuscript: J.J. Tao, V. Makker, K.A. Cadoo, J.F. Liu, D.W. Rasco, W.H. Navarro, C.M. Haqq, D.M. Hyman

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J.J. Tao, N.A. Cangemi, W.H. Navarro, C.M. Haqq, D.M. Hyman

Study supervision: W.H. Navarro, C.M. Haqq, D.M. Hyman

The clinical trial was supported by Atara Biotherapeutics and an institutional Core Grant.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Loomans
HA
,
Andl
CD
. 
Intertwining of Activin A and TGFβ signaling: dual roles in cancer progression and cancer cell invasion
.
Cancers
2014
;
7
:
70
91
.
2.
Loomans
HA
,
Andl
CD
. 
Activin receptor-like kinases: a diverse family playing an important role in cancer
.
Am J Cancer Res
2016
;
6
:
2431
47
.
3.
Kim
H
,
Watkinson
J
,
Varadan
V
,
Anastassiou
D
. 
Multi-cancer computational analysis reveals invasion-associated variant of desmoplastic reaction involving INHBA, THBS2 and COL11A1
.
BMC Med Genomics
2010
;
3
:
51
.
4.
Steller
MD
,
Shaw
TJ
,
Vanderhyden
BC
,
Ethier
J-F
. 
Inhibin resistance is associated with aggressive tumorigenicity of ovarian cancer cells
.
Mol Cancer Res
2005
;
3
:
50
61
.
5.
Lonardo
E
,
Hermann
PC
,
Mueller
M-T
,
Huber
S
,
Balic
A
,
Miranda-Lorenzo
I
, et al
Nodal/Activin signaling drives self-renewal and tumorigenicity of pancreatic cancer stem cells and provides a target for combined drug therapy
.
Cell Stem Cell
2011
;
9
:
433
46
.
6.
Lonardo
E
,
Frias-Aldeguer
J
,
Hermann
PC
,
Heeschen
C
. 
Pancreatic stellate cells form a niche for cancer stem cells and promote their self-renewal and invasiveness
.
Cell Cycle
2012
;
11
:
1282
90
.
7.
Argilés
JM
,
Orpí
M
,
Busquets
S
,
López-Soriano
FJ
. 
Myostatin: more than just a regulator of muscle mass
.
Drug Discov Today
2012
;
17
:
702
9
.
8.
Elkina
Y
,
von Haehling
S
,
Anker
SD
,
Springer
J
. 
The role of myostatin in muscle wasting: an overview
.
J Cachexia Sarcopenia Muscle
2011
;
2
:
143
51
.
9.
Smith
RC
,
Lin
BK
. 
Myostatin inhibitors as therapies for muscle wasting associated with cancer and other disorders
.
Curr Opin Support Palliat Care
2013
;
7
:
352
60
.
10.
Matzuk
MM
,
Finegold
MJ
,
Mather
JP
,
Krummen
L
,
Lu
H
,
Bradley
A
. 
Development of cancer cachexia-like syndrome and adrenal tumors in inhibin-deficient mice
.
Proc Natl Acad Sci U S A
1994
;
91
:
8817
21
.
11.
Zhou
X
,
Wang
JL
,
Lu
J
,
Song
Y
,
Kwak
KS
,
Jiao
Q
, et al
Reversal of cancer cachexia and muscle wasting by ActRIIB antagonism leads to prolonged survival
.
Cell
2010
;
142
:
531
43
.
12.
Cobellis
L
,
Reis
FM
,
Luisi
S
,
Danero
S
,
Pirtoli
L
,
Scambia
G
, et al
High concentrations of activin A in the peritoneal fluid of women with epithelial ovarian cancer
.
J Soc Gynecol Investig
2004
;
11
:
203
6
.
13.
Menon
U
,
Riley
SC
,
Thomas
J
,
Bose
C
,
Dawnay
A
,
Evans
LW
, et al
Serum inhibin, activin and follistatin in postmenopausal women with epithelial ovarian carcinoma
.
BJOG
2000
;
107
:
1069
74
.
14.
Welt
CK
,
Lambert-Messerlian
G
,
Zheng
W
,
Crowley
WF
,
Schneyer
AL
. 
Presence of activin, inhibin, and follistatin in epithelial ovarian carcinoma
.
J Clin Endocrinol Metab
1997
;
82
:
3720
7
.
15.
Tournier
I
,
Marlin
R
,
Walton
K
,
Charbonnier
F
,
Coutant
S
,
Théry
J-C
, et al
Germline mutations of inhibins in early-onset ovarian epithelial tumors
.
Hum Mutat
2014
;
35
:
294
7
.
16.
Karve
TM
,
Preet
A
,
Sneed
R
,
Salamanca
C
,
Li
X
,
Xu
J
, et al
BRCA1 regulates follistatin function in ovarian cancer and human ovarian surface epithelial cells
.
PLoS One
2012
;
7
:
e37697
.
17.
Wiegand
KC
,
Shah
SP
,
Al-Agha
OM
,
Zhao
Y
,
Tse
K
,
Zeng
T
, et al
ARID1A mutations in endometriosis-associated ovarian carcinomas
.
N Engl J Med
2010
;
363
:
1532
43
.
18.
Guan
B
,
Wang
T-L
,
Shih
I-M
. 
ARID1A, a factor that promotes formation of SWI/SNF-mediated chromatin remodeling, is a tumor suppressor in gynecologic cancers
.
Cancer Res
2011
;
71
:
6718
27
.
19.
Shah
SP
,
Köbel
M
,
Senz
J
,
Morin
RD
,
Clarke
BA
,
Wiegand
KC
, et al
Mutation of FOXL2 in granulosa-cell tumors of the ovary
.
N Engl J Med
2009
;
360
:
2719
29
.
20.
Ellsworth
BS
,
Burns
AT
,
Escudero
KW
,
Duval
DL
,
Nelson
SE
,
Clay
CM
. 
The gonadotropin releasing hormone (GnRH) receptor activating sequence (GRAS) is a composite regulatory element that interacts with multiple classes of transcription factors including Smads, AP-1 and a forkhead DNA binding protein
.
Mol Cell Endocrinol
2003
;
206
:
93
111
.
21.
Pisarska
MD
,
Bae
J
,
Klein
C
,
Hsueh
AJW
. 
Forkhead l2 is expressed in the ovary and represses the promoter activity of the steroidogenic acute regulatory gene
.
Endocrinology
2004
;
145
:
3424
33
.
22.
McConechy
MK
,
Färkkilä
A
,
Horlings
HM
,
Talhouk
A
,
Unkila-Kallio
L
,
van Meurs
HS
, et al
Molecularly defined adult granulosa cell tumor of the ovary: the clinical phenotype
.
J Natl Cancer Inst
2016
;
108
. pii:
djw134
.
23.
Yanagida
S
,
Anglesio
MS
,
Nazeran
TM
,
Lum
A
,
Inoue
M
,
Iida
Y
, et al
Clinical and genetic analysis of recurrent adult-type granulosa cell tumor of the ovary: persistent preservation of heterozygous c.402C>G FOXL2 mutation
.
PLoS One
2017
;
12
:
e0178989
.
24.
McTavish
KJ
,
Nonis
D
,
Hoang
YD
,
Shimasaki
S
. 
Granulosa cell tumor mutant FOXL2C134W suppresses GDF-9 and activin A-induced follistatin transcription in primary granulosa cells
.
Mol Cell Endocrinol
2013
;
372
:
57
64
.
25.
Eisenhauer
EA
,
Therasse
P
,
Bogaerts
J
,
Schwartz
LH
,
Sargent
D
,
Ford
R
, et al
New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1)
.
Eur J Cancer
2009
;
45
:
228
47
.
26.
ATS Statement
.
Am J Respir Crit Care Med
2002
;
166
:
111
7
.
27.
Seo
J
,
Chu
H
,
Lee
JS
,
Kim
DY
. 
Mucocutaneous telangiectasia as a diagnostic clue of hereditary hemorrhagic telangiectasia: an activin receptor-like kinase-1 mutation in a Korean Patient
.
Ann Dermatol
2016
;
28
:
264
6
.