From July 2, 2018, to July 20, 2021, 39 patients were enrolled and received ≥1 dose of OBI-3424. Patient demographics and baseline characteristics are displayed in Table 1. The median age was 67 years (range, 39–78 years). Patients were predominantly male (22/39, 56%), Caucasian/White (29/39, 74%), non-Hispanic/Latino (36/39, 92%), and had Eastern Cooperative Oncology Group performance status 1 (32/39, 82%). The most common tumor types were prostate (8/39, 21%) and colorectal (5/39, 13%) carcinoma.

Patients were heavily pretreated. The median number of prior therapies was 5.7 (range, 2–15). Twenty-two patients (22/39, 56.4%) had prior surgery and 7 patients (7/39, 18.0%) had received prior radiation therapy.

Prior therapy included investigational drugs (n = 13 patients, 33.3%), either as monotherapy or in a combination regimen. Overall, 94.9% (37/39) of patients had received combination therapy (>50 combination regimens) and only 2 patients had received only monotherapy. Antimetabolites were the most widely administered class of drug with 21 patients receiving any of the following four drugs: 5-FU, capecitabine, gemcitabine, and pemetrexed. Six patients had received FOLFIRI, 8 had received FOLFOX, and 5 had received FOLFIRINOX.

Prior therapy included platinum agents in 69% (27/39) of patients or taxanes (paclitaxel, docetaxel, cabazitaxel, or nab-paclitaxel) in 54% (21/39) of patients. Of the 8 patients with prostate cancer, 6 had prior surgery, 8 had chemotherapy, and 7 patients were receiving androgen deprivation/hormonal therapy.

In Schedule A, OBI-3424 was well tolerated at doses of up to 8 mg/m2. Hematologic DLTs (thrombocytopenia and anemia) were observed in all 6 patients who received a dose of 12 mg/m2 on Schedule A. These DLTs led to dose modifications and/or delays in subsequent cycles, and therefore, 8 mg/m2 was considered the MTD of OBI-3424 when administered as per Schedule A. Given that the platelet count nadirs were occurring on day 15 and day 21, the schedule of administration was modified to Schedule B. OBI-3424 administered every 3 weeks was tolerated at doses of up to 14 mg/m2 and the MTD was not reached. Overall, 19 patients required a red blood cell transfusion for grade ≥3 anemia, including 11 patients who required 2 or more separate red blood cell transfusions. Three patients receiving 14 mg/m2 OBI-3424 required a transfusion; 1 of these patients required 2 transfusions and 1 patient required 3 transfusions. Considering the significant burden associated with these transfusions, the RP2D and regimen of OBI-3424 were determined to be 12 mg/m2 every 3 weeks (Schedule B).

The median number of doses administered was 4 (range, 1–38). TEAEs occurred in 32 of the 39 patients (82%; Table 2); the most common TEAEs were anemia (25/39, 64%), thrombocytopenia (19/39, 49%), nausea (10/39, 26%), and fatigue (8/39, 21%). Gastrointestinal TEAEs were minimal. Nausea occurred in 10 patients (1 patient in each of the 1, 4, and 6 mg/m2 [Schedule A] cohorts, 1 patient in the 8 and 2 patients in each of the 10, 12, and 14 mg/m2 [Schedule B] cohorts), and only 5 patients reported vomiting (1 in the 4 mg/m2 [Schedule A] cohort, 3 in the 12 mg/m2 [Schedule A] cohort, and 1 in the 14 mg/m2 [Schedule B] cohort). Diarrhea occurred only at the higher doses of 12 mg/m2 (2 patients in each of the Schedule A and Schedule B cohorts) and 14 mg/m2 (Schedule B; 1 patient). Patients received prophylaxis with antinausea medications 1 h prior to the initial dose of OBI-3424; these medications included ondansetron at 8 mg intravenously and then orally every 8 h as needed or another 5-HT3 antagonist. Antidiarrheal therapies including dicyclomine (if predominant issue was cramping or abdominal pain), diphenoxylate/atropine and/or loperamide, and, if necessary, budesonide, were implemented on an as-needed basis.

Four patients experienced dyspnea at higher doses of OBI-3424: 1 patient each in the 10 mg/m2 (Schedule B) and 12 mg/m2 (Schedule A) cohorts and 2 patients in the 14 mg/m2 (Schedule B) cohort. Dyspnea was determined not to be related to treatment with OBI-3424. A 78-year-old male patient with carcinoma of the prostate required 2 separate red blood cell transfusions for worsening anemia and low hemoglobin, which were likely the cause of the dyspnea in this patient. A 66-year-old female patient had anemia, pleural effusion, and pneumothorax (one-third of right lower lobe with diminished sounds and faint crackling), all of which may have contributed to dyspnea in this patient. The dyspnea in a 66-year-old male patient and a 63-year-old female patient, both with colorectal cancer, was likely due to worsening anemia, and in the case of the female patient, a mitral valve prolapse and hypoalbuminemia.

Nineteen patients (49%) had grade ≥3 TEAEs, including anemia (16/39, 41%), thrombocytopenia (10/39, 26%), neutropenia (3/39, 8%), and lymphocytopenia (2/39, 5%). Three patients, all of whom received OBI-3424 at 12 mg/m2 on Schedule A, experienced grade ≥3 anemia as a serious TEAE, and 2 of these patients also experienced grade ≥3 thrombocytopenia as a serious TEAE. Nineteen patients required transfusions; 10 of those 19 patients were receiving doses ≥12 mg/m2 (5/6 patients receiving 12 mg/m2 on Schedule A, 2/6 patients receiving 12 mg/m2 on Schedule B, and 3/6 patients receiving 14 mg/m2 on Schedule B).

Eleven patients (28%) required dose interruption and/or reduction due to TEAEs. Ten patients (26%) experienced dose reductions due to thrombocytopenia, anemia, neutropenia, and fatigue, and 4 patients (10%) experienced dose interruptions due to anemia, thrombocytopenia, dehydration, and orthostatic hypotension. Two patients (5%) discontinued treatment. One patient, a 54-year-old male patient with castrate-resistant prostate cancer metastatic to the bone without visceral metastases, who received 3 doses of OBI-3424 at 2 mg/m2 on Schedule A, experienced a grade 3 right cerebral subdural hematoma. The patient had received treatment for 27 days when he presented to the emergency department following several falls that resulted in confusion. On admission, a computed tomography (CT) scan revealed a right cerebral subdural hematoma, and the patient underwent an emergency craniotomy. The next day, the patient was doing relatively well with no neurologic deficits and no headache. Four days later, a CT scan revealed an increase in the hematoma, and another craniotomy was performed. The day after surgery, the patient developed a right frontal hematoma with subsequent inability to move the upper left extremity, facial droop, dysarthria, and seizures. Two days later, a CT scan showed increased hemorrhaging and edema, which the hematologist considered to be related to steroid therapy for cytopenia. The patient was discharged to hospice care in stable condition 1 day later. Ten days after discharge, the patient suffered another subdural hematoma and died. During hospitalization, the patient received 2 separate transfusions 14 days apart due to worsening anemia and thrombocytopenia. The subdural hematomas were determined to be due to head trauma resulting from several falls and were not related to treatment with OBI-3424.

The second patient discontinued treatment with OBI-3424 due to leukemia secondary to oncologic chemotherapy (the patient had received extensive previous treatments, including DNA-alkylating agents, and had received only 1 dose of OBI-3424 at the 4.0 mg/m2 dose level; therefore this adverse event was also not attributed to the treatment drug). There were no grade 5 drug-related TEAEs.

Treatment with OBI-3424 at doses of ≥8 mg/m2 using Schedule A and ≥12 mg/m2 using Schedule B was associated with grade ≥3 anemia and thrombocytopenia. Changes in hemoglobin and platelet counts during treatment are illustrated in Fig. 3a, b.

Change in (a) hemoglobin value and (b) platelet count from baseline to the minimum observed postbaseline value. Solid diagonal line represents no change; vertical and horizontal lines are borders between NCI CTCAE grades. Vertical distance from the diagonal represents magnitude of change; above the diagonal is an increase, below a decrease. For analytes in which the reference range varied by age/sex, the lowest value was used to create the borders. Color gradient indicates dose intensity (lighter = lower dose intensity, darker = higher dose intensity). C1 D1 cycle 1, day 1, NCI CTCAE National Cancer Institute Common Terminology Criteria for Adverse Events, Q3W every 3 weeks.

Mean plasma concentration versus time profiles of single doses of OBI-3424 and its active metabolite (OBI-2660) at doses ranging from 1 to 14 mg/m2 are illustrated in Fig. 4a, b. Pharmacokinetic profiles revealed that OBI-3424 clearance varied widely between patients and across dose levels, whereas the concentration of OBI-2660 increased with increasing dose and exhibited less variation across dose levels.

a OBI-3424 and b OBI-2660 during cycle 1 (days 1 and 8 combined).

OBI-3424 and OBI-2660 pharmacokinetic parameters by dose level are summarized in Supplementary Table S1 (Supplementary Material). Maximum serum concentrations (Cmax) of OBI-3424 generally occurred shortly after the 30-min drug infusion (between 33 and 50 min). The time to maximum concentration of the active metabolite OBI-2660 was observed to be slightly delayed compared with that for OBI-3424, with the Cmax achieved between 1.33 h and 1.75 h after the start of drug infusion. Exposures (Cmax and area under the concentration-time curve [AUC]) to OBI-3424 increased with increasing doses; however, an analysis of dose proportionality for OBI-3424 using the power model was inconclusive. Exposures to OBI-2660 increased across the dose range of 1 mg/m2 to 14 mg/m2 but in a less than proportional manner.

Exposure (AUC) of OBI-2660 was much lower than that of OBI-3424, with the AUC approximately 1.4% to 3.4% of OBI-3424. The half-life of OBI-3424 was short (0.21–0.74 h), whereas OBI-2660 had a longer half-life (1.87–3.08 h). Mean clearance of OBI-3424 ranged from 4.8 L/h/m2 to 8.9 L/h/m2, and volume of distribution at steady state ranged from 2.5 L/m2 to 4.3 L/m2 for OBI-3424. No accumulation of exposure (Cmax and AUC) between 2 doses (cycle 1 day 1 and cycle 1 day 8) was observed for either OBI-3424 or OBI-2660.

Thirty-two of the 39 patients provided archival (28/32) or fresh biopsy (4/32) tumor tissue specimens for AKR1C3 expression testing using the validated, automated IHC assay (NeoGenomics). Among the 32 specimens, 6 tumor tissue specimens did not have sufficient viable tumor cells and were non-evaluable. AKR1C3 H-scores by patient are listed in Fig. 5.

Antitumor activity in patients treated with OBI-3424 at postbaseline scans.

Thirty-three of 39 patients were evaluable for tumor response by RECIST v1.1. Six patients were not evaluable because of early discontinuation due to disease progression (n = 2), consent withdrawal (n = 2), clinically significant deterioration in the patient’s condition (n = 1), or death (n = 1).

A patient with cholangiocarcinoma, treated with OBI-3424 at 10 mg/m2 (Schedule B), had a partial response (33% reduction in tumor measurements). Stable disease lasting for ≥6 weeks was noted in 21 patients (64%), with 15/21 patients (71%) experiencing stable disease for ≥4 cycles (12 weeks), including a patient with esthesioneuroblastoma who remained on study for >28 cycles (12 mg/m2 OBI-3424, Schedule A, dose reduced to 9 mg/m2 every 3 weeks after 4 cycles). The remaining 11 patients (33%) had progressive disease. Best response by RECIST v1.1 is shown in Fig. 5.

Reasons for discontinuation of the study drug for all patients were as follows: disease progression (n = 27), consent withdrawn (n = 5), clinically significant deterioration in the patient’s condition (n = 4), death (subdural hematoma unrelated to study drug; n = 1), logistical reasons (unable to travel to receive treatment due to COVID-19 protocols; n = 1), and leukemia secondary to chemotherapy (unrelated to study drug; n = 1).