Purpose:

NUC-1031 is a first-in-class ProTide modification of gemcitabine. In PRO-002, NUC-1031 was combined with carboplatin in recurrent ovarian cancer.

Patients and Methods:

NUC-1031 was administered on days 1 and 8 with carboplatin on day 1 every 3 weeks for up to six cycles. Four dose cohorts of NUC-1031 (500, 625, and 750 mg/m2) with carboplatin (AUC4 or 5) were investigated. Primary endpoint was recommended phase II combination dose (RP2CD). Secondary endpoints included safety, investigator-assessed objective response rate (ORR), clinical benefit rate (CBR), progression-free survival (PFS), and pharmacokinetics.

Results:

A total of 25 women with recurrent ovarian cancer, a mean of 3.8 prior lines of chemotherapy, and a median platinum-free interval of 5 months (range: 7–451 days) were enrolled; 15 of 25 (60%) were platinum resistant, 9 (36%) were partially platinum sensitive, and 1 (4%) was platinum sensitive. Of the 23 who were response evaluable, there was 1 confirmed complete response (4%), 5 partial responses (17%), and 8 (35%) stable disease. The ORR was 26% and CBR was 74% across all doses and 100% in the RP2CD cohort. Median PFS was 27.1 weeks. NUC-1031 was stable in the plasma and rapidly generated high intracellular dFdCTP levels that were unaffected by carboplatin.

Conclusions:

NUC-1031 combined with carboplatin is well tolerated in recurrent ovarian cancer. Highest efficacy was observed at the RP2CD of 500 mg/m2 NUC-1031 on days 1 and 8 with AUC5 carboplatin day 1, every 3 weeks for six cycles. The ability to deliver carboplatin at AUC5 and the efficacy of this schedule even in patients with platinum-resistant disease makes this an attractive therapeutic combination.

Translational Relevance

The antimetabolite gemcitabine has shown some evidence of synergy with platinum; however, it is rarely used in the management of platinum-resistant ovarian cancer. In this PRO-002 study, the ProTide NUC-1031, which is a phosphoramidate modification of gemcitabine, was successfully administered with carboplatin in 25 heavily pretreated patients with recurrent ovarian cancer, of whom 15 (60%) were platinum resistant and 9 (36%) had prior exposure to gemcitabine. The recommended phase II combination dose was defined as 500 mg/m2 NUC-1031 (days 1 and 8) with carboplatin AUC5 (day 1) given every 3 weeks for up to six cycles. At this dose, strong efficacy was observed, even in platinum-resistant and gemcitabine pretreated patients, and myelosuppression was mild. The safety, pharmacokinetic, and efficacy profile of NUC-1031 plus carboplatin indicates this combination could be an effective treatment strategy for recurrent ovarian cancer.

The antimetabolite chemotherapy gemcitabine (2′-2′-difluorodeoxycytidine; dFdC) has been used for the treatment of breast, ovarian, non–small cell lung, pancreatic, bladder, and other cancers since the 1990s (1–3). Like other nucleoside analogues, gemcitabine is a prodrug and is dependent on active transport into the cancer cell and subsequent stepwise enzymatic metabolism to gemcitabine monophosphate (dFdCMP), diphosphate (dFdCDP), and triphosphate (dFdCTP; refs. 4, 5). The cytotoxic activity of gemcitabine is attributable to dFdCTP which is misincorporated into replicating DNA (in the place of deoxycytidine) resulting in masked chain termination and cell death (4, 5). In addition, the intermediate metabolite dFdCDP inhibits ribonucleotide reductase and prevents the formation of deoxycytidine, a pyrimidine nucleoside essential for DNA synthesis (5). The first step in this activation process, the conversion of gemcitabine to dFdCMP, is rate limited by the availability of the enzyme deoxycytidine kinase (dCK) which is typically expressed at low levels in resistant ovarian cancer (5). In addition, the enzyme cytidine deaminase that is often highly expressed in resistant ovarian cancer degrades the majority of gemcitabine to an inactive metabolite, 2′,2′-difluorodeoxyuridine (dFdU; ref. 6).

NUC-1031 is a ProTide form of gemcitabine that is chemically synthesised to overcome these limitations (7). It is comprised of preactivated dFdCMP protected by a biolabile phosphoramidate motif (7). NUC-1031 enters the cell independent of the Human Equilibrative Nucleoside Transporter (hENT1) channel whereupon the motif is cleaved by intracellular esterases releasing dFdCMP (7). Having bypassed the first rate-limiting phosphorylation step by the enzyme dCK, dFdCMP is then rapidly converted to the active metabolites dFdCDP and dFdCTP (7). In a 68-patient phase I dose-escalation and dose-expansion first-in-human (FIH) study, NUC-1031 displayed good efficacy and tolerability when given intravenously on days 1, 8, and 15 in 28-day cycles in heavily pretreated patients with advanced cancers, including gynecologic malignancies (8). High levels of dFdCTP were detected in patients' peripheral blood mononuclear cells (PBMC; ref. 8). Importantly, grade 3/4 myelosuppression was minimal at doses of 825 mg/m2 and below (8). A recommended phase II monotherapy dose range for NUC-1031 was defined as 825 mg/m2. In light of the favorable impact on bone marrow function observed with NUC-1031, we conducted PRO-002 to explore giving NUC-1031 alongside AUC4 or 5 carboplatin in patients with recurrent ovarian cancer. By testing four different dosing schedules, an efficacious dose of NUC-1031 was identified that could be given alongside AUC5 carboplatin with minimal myelotoxicity. This was defined as 500 mg/m2 NUC-1031 given on days 1 and 8 every 3 weeks with AUC5 carboplatin given on day 1 every 3 weeks, both for up to six cycles.

This open-label, phase Ib combination dose-escalation/expansion study was conducted at two clinical centers in the United Kingdom: the Churchill Hospital, Oxford and the Hammersmith Hospital, London. The study was performed in accordance with the Declaration of Helsinki and the principles of Good Clinical Practice (9). The protocol was approved by the West London Research Ethics Committee and all patients provided written informed consent before undergoing any study procedures. The primary objective of the study was to determine the recommended phase II combination dose (RP2CD) of NUC-1031 (on days 1 and 8) and carboplatin (on day 1) given in a 21-day schedule for up to six cycles. Secondary objectives were to evaluate the safety and tolerability of the combination and its efficacy by objective response rate (ORR), clinical benefit rate (CBR), progression-free survival (PFS), and best overall response (BOR), utilizing the evaluation criteria determined by the Gynecologic Cancer Intergroup (GCIG) and RECIST version 1.1 (10–12). Research objectives included assays to explore the relationship between NUC-1031 pharmacokinetics, pharmacodynamics, and clinical activity.

Eligible patients were ages ≥18 years with Eastern Cooperative Oncology Group (ECOG) performance score 0–2; adequate organ function; histologically confirmed epithelial ovarian, fallopian tube or primary peritoneal cancer (here collectively termed “ovarian cancer”); and evaluable recurrent disease on radiological imaging (in accordance with RECIST v1.1). Patients must have had a platinum-free interval (PFI) of ≤24 months. PFI was defined as the time since last (adjuvant or nonadjuvant) platinum (carboplatin or cisplatin) or platinum-containing chemotherapy until start of subsequent nonplatinum therapy or consent for PRO-002 (whichever occurred first; ref. 13). Exclusion criteria included prior allergy to gemcitabine or carboplatin and symptomatic central nervous system metastases.

NUC-1031 was administered intravenously on days 1 and 8 every 21 days via a Groshong or peripherally inserted central catheter. Initially, consistent with the phase I study (8), NUC-1031 was administered in a 10- to 15-minute bolus injection but, in March 2016, this was substituted by a 250 mL saline solution formulation of NUC-1031 that was administered over 30 minutes. Carboplatin was administered via a 1-hour intravenous infusion on day 1 of each 21-day cycle, immediately prior to the administration of NUC-1031. On the basis of the findings of the PRO-001 (phase I) study (8), a dose below the MTD of 750 mg/m2 was chosen as the starting dose of NUC-1031 to combine with carboplatin AUC4. The dose was escalated sequentially in cohorts of 3 to 6 patients using an accelerated titration method (Table 1) with alternating escalation of NUC-1031 and carboplatin with each planned cohort. Dose escalation stopped when the MTD/RP2CD had been determined. The MTD was defined as the highest dose level for which fewer than 2 of 6 (or <33%) patients experienced dose-limiting toxicities (DLT). The expansion cohort was then dosed at the MTD to confirm the RP2CD.

Table 1.

Number of patients and doses of NUC-1031 and carboplatin by cohort.

CohortNUC-1031 (days 1 and 8, every 3 weeks) × 6 cyclesCarboplatin (day 1 only, every 3 weeks) × 6 cyclesNumber of patients (n = 25)
750 mg/m2 AUC4 
750 mg/m2 AUC5 
2B 500 mg/m2 AUC5 12 
625 mg/m2 AUC4 
CohortNUC-1031 (days 1 and 8, every 3 weeks) × 6 cyclesCarboplatin (day 1 only, every 3 weeks) × 6 cyclesNumber of patients (n = 25)
750 mg/m2 AUC4 
750 mg/m2 AUC5 
2B 500 mg/m2 AUC5 12 
625 mg/m2 AUC4 

CT-based tumor assessments were conducted according to RECIST v1.1 at screening and weeks 9 and 18. Serum CA125 was measured at baseline and at day 1 of each treatment cycle. Safety parameters were continually assessed and based on adverse events (AE) graded according to the NCI Common Terminology Criteria for Adverse Events version 4.03, clinical laboratory data, and physical examinations. In view of the association between gemcitabine and pulmonary toxicity, lung function assessments, comprised of spirometry, lung volumes, and gas transfer, were performed at baseline and at the end of study participation (14). Blood samples were collected for pharmacokinetic analysis at set timepoints after the end of NUC-1031 infusion (pre-dose, end of infusion, 30 minutes, 2 hours, 24 hours) on cycle 1 day 1. A DLT was defined as any of the following occurring during the first treatment cycle: grade 3 or 4 nonhematologic toxicity (excluding nausea/vomiting/diarrhea or rash that responded to standard medical treatment), grade 3 or 4 nausea/vomiting/diarrhea that occurred despite standard medical treatment, grade 4 neutropenia lasting >7 days, febrile neutropenia, grade 4 thrombocytopenia, any toxicity related to NUC-1031 that was unresolved after a treatment delay of >14 days, or isolated/recurrent toxicity that was judged by the investigator to be a DLT. The use of granulocyte colony-stimulating factor was permitted from cycle 2 only.

Efficacy and safety analyses

ORR with exact binomial 95% confidence intervals (CI) and Kaplan–Meier estimates for PFS (based on clinical symptoms and/or RECIST v1.1 progression) were calculated; safety analyses were descriptive. With respect to primary objectives and endpoints, no specific hypotheses were tested statistically. The primary focus was on determining the RP2CD, the safety profile, and the identification of a range of biologically active doses and the pharmacokinetics of NUC-1031 in patients with ovarian cancer. Baseline characteristics of the patients, together with safety, pharmacokinetic, biomarker, and antitumor activity summaries were provided by dose level of NUC-1031/carboplatin and overall. No formal interim analysis was performed in this study. Safety, pharmacokinetic, and pharmacodynamic biomarker data were reviewed on an ongoing basis in line with the cohort progression criteria. Progression was defined as the date of objective progression by RECIST, by CA125 (GCIG) criteria, or the date of symptomatic progression, whichever occurred first.

Pharmacokinetic analysis

Review of preliminary safety and available pharmacokinetic data from the dose escalation was performed after completion of each dosing cohort. Blood samples (0.05, 0.55, 2.05, and 24 hours) were collected for pharmacokinetic analyses at set timepoints after the end of NUC-1031 infusion on cycle 1 day 1. Plasma was assayed for NUC-1031, dFdC, and dFdU and PBMCs for dFdCTP using ultraperformance liquid chromatography tandem mass spectrometry (UPLC MS-MS; ref. 15). Given the relatively sparse pharmacokinetic sampling employed during the study, a Bayesian post hoc approach utilizing a base population model was taken to estimate pharmacokinetic parameters (Cmax, AUC0–24, AUC0-∞, t1/2, Vss, and CL) for patients who provided concentration–time observations.

Pharmacokinetic modeling

The base population pharmacokinetic model was developed using non-linear mixed effects (NONMEM) using phase IB pharmacokinetic data (PRO-001) as part of the ongoing development of NUC-1031 (16). The model was built in a stepwise manner with NUC-1031 being characterized first followed by dFdC and then dFdU. Random variability was tested for clearance (CL) and volume (V). The evaluation of the model was based on the values given by the objective function (−2log likelihood), the Akaike information criteria (AIC), Bayesian information criteria, and the coefficient of determination of linear regression of the observed versus predicted values. A visual predictive check was performed to assess whether the model was predictive.

Statistical analysis

Sample size calculations were based on a Fleming design (17). No formal statistical analyses were planned or performed on safety or efficacy data.

Between November 27, 2014 and November 10, 2016, 25 eligible patients were enrolled in the study. At the time of data cutoff on May 30, 2017, no patients remained on study; 18 successfully completed treatment and 7 discontinued (3 due to disease progression, 2 due to adverse events, 1 due to unmanageable toxicity, and 1 due to physician decision). All 25 patients were considered toxicity evaluable. Two patients discontinued the study before their first post-enrollment (9 weeks) CT scan for toxicities that had not resolved to baseline within 2 weeks and the remaining 23 participants were response evaluable. Twenty (80%) of the 25 patients were included in the CA125 evaluable-for-response set; the remaining 5 patients (20%) did not have a baseline CA125 level of ≥2×upper limit of normal (ULN). See Table 2 for patient characteristics. Patients had a mean age of 64 years (range: 37–77 years) and the majority had a good performance status (92% ECOG PS 0 or 1). All had primary ovarian cancer and 23 of 25 (92%) had serous histology. The majority of study patients (15/25, 60%) had platinum-resistant disease with a PFI of <6 months; the median PFI among them was 5 months (range: 7–184 days). Of the remaining 10 patients, 9 (36%) had partially platinum-sensitive disease with a median PFI of 8 months (range: 195–311 days) and 1 (4%) was platinum sensitive with a PFI of 14.8 months. Patients had received a mean of 3.8 prior lines of treatment (median 3, range: 2–8), including first-line treatment. Nine (35%) had received gemcitabine as part of a preceding regimen, 3 of whom entered PRO-002 having progressed immediately after gemcitabine and carboplatin treatment. BRCA testing was not universally conducted at the time of the study so BRCA status was unknown for 14 (56%) patients; 5 of the remaining 11 (20%) were BRCA wild-type and 6 (24%) were confirmed BRCA mutation carriers (5 with BRCA 1 and 1 with BRCA 2 mutations).

Table 2.

Demographic and clinical characteristics at baseline.

Demographic and clinical characteristics at baseline.
Demographic and clinical characteristics at baseline.

Patients were sequentially recruited into the study: 6 were recruited into Cohort 1, 1 into Cohort 2, 12 into Cohort 2B, and 6 into Cohort 3 (Table 1). Eighteen (72%) patients completed all six cycles of study treatment, while 7 discontinued prematurely for progressive disease or other reasons. Among the 25 treated patients, a mean of 5.1 cycles of NUC-1031 plus carboplatin were administered. Across all dose levels, 68% and 81% of planned doses of NUC-1031 and carboplatin were administered, respectively. The majority of patients had dose reductions and dose modifications of both drugs and 12 (48%) patients required ≥1 NUC-1031 dose reduction. Summarized in Table 3.

Table 3.

Treatment completion and intensity by cohort.

Treatment completion and intensity by cohort.
Treatment completion and intensity by cohort.

Safety

All 16 patients in the dose-escalation part of the study were DLT evaluable. Among them, 4 experienced at least one DLT during cycle 1 of their treatment. The first was a DLT of grade 4 thrombocytopenia in a patient in Cohort 1 (750 mg/m2 NUC-1031 and carboplatin AUC4) prompting the enrollment of a further 3 patients to this dose level. The same patient had four additional DLTs: grade 3 fatigue, grade 3 febrile neutropenia, grade 4 leucopenia, and grade 4 neutropenia. As no further DLTs were observed in Cohort 1, the carboplatin dose was escalated to AUC5 and given with 750 mg/m2 NUC-1031 for Cohort 2. The first patient to be enrolled in this cohort experienced grade 3 neutropenia before C1D8 which was not a DLT but required a dose interruption and reduction. In view of the rapid emergence of this AE, the investigators agreed to discontinue enrollment in this cohort and opened Cohort 2B in which the carboplatin was given at AUC5 alongside 500 mg/m2 NUC-1031. Three patients were recruited to this cohort and no DLTs were observed. Cohort 3 was then opened exploring 625 mg/m2 NUC-1031 and carboplatin AUC4. Six patients were enrolled in this cohort, and 4 DLTs were observed in 3 patients: grade 4 thrombocytopenia in one, grade 3 fatigue in one, and grade 3 fatigue with grade 3 hyponatremia in the third. The MTD was therefore defined as 500 mg/m2 NUC-1031 and carboplatin AUC5 (dose used in Cohort 2B). Cohort 2B was then expanded to include an additional 9 patients (to 12 in total) which confirmed this as the RCP2D.

Across the dose-escalation and dose-expansion parts of the study, the majority of patients (88%) experienced an AE of grade 3 or higher and underwent at least one dose modification of NUC-1031 (84%) and carboplatin (68%; Table 4). A total of 24 (96%) patients experienced at least one grade 1 or 2 treatment-related AE. The most common grade 1 or 2 AEs observed across all doses that were causally attributable to carboplatin or NUC-1031 were: neutropenia (68%), hypomagnesemia (64%), nausea (52%), fatigue (48%), anemia (48%), leukopenia (48%), and transaminitis (48%; Table 5). Neutropenia was the principal reason for dose modification (76%). Although the numbers of patients per cohort were too small for formal statistical comparison, across all cohorts, grades 1–4 myelosuppression (anemia, neutropenia, and thrombocytopenia) were NUC-1031 dose dependent, affecting 67% of patients in Cohort 1 (NUC-1031 750 mg/m2) and 58% in Cohort 2B (NUC-1031 500 mg/m2). No allergic reactions to carboplatin or NUC-1031 were observed in any study participants.

Table 4.

Grade 3/4 adverse events by system.

Grade 3/4 adverse events by system.
Grade 3/4 adverse events by system.
Table 5.

Grade 1 and 2 adverse events causally related to study drug affecting >5% study participants by system.

Grade 1 and 2 adverse events causally related to study drug affecting >5% study participants by system.
Grade 1 and 2 adverse events causally related to study drug affecting >5% study participants by system.

Pharmacokinetic model

The base population pharmacokinetic model that best described NUC-1031 and its metabolites dFdC and dFdU is shown in Supplementary Fig. S1. On the basis of AIC and goodness-of-fit (GOF) plots, a three-compartment model was found as the most appropriate model to describe the data for NUC-1031. The rate of conversion of NUC-1031 to dFdC was also described in a three-compartment model, whereas a single-compartment model was used for the conversion of dFdC to dFdU. Population pharmacokinetic parameters of NUC-1031 and its metabolites dFdC and dFdU are shown in Supplementary Table S2. GOF plots for NUC-1031, dFdC, and dFdU are shown in Supplementary Fig. S2. The R squared values of the linear regression for NUC-1031, dFdC, and dFdU were found to be 0.77, 0.53, and 0.88, respectively. The predictive nature of the model to determine the pharmacokinetics of NUC-1031 and its metabolites dFdC and dFdU is shown in Supplementary Fig. S3.

This model was used to estimate pharmacokinetic parameters for each of the 21 participants who provided evaluable concentration–time observations. The individual post hoc pharmacokinetic parameter estimates obtained from the final model were then used to calculate pharmacokinetic exposure parameters for each analyte. The base population pharmacokinetic model provided an adequate fit to the individual observed plasma concentration–time data for NUC-1031, dFdC, and dFdU (Supplementary Fig. S4).

Pharmacokinetics

Twenty-one participants had evaluable pharmacokinetic samples: 8 participants from Cohort 2B (NUC-1031 500 mg/m2), 6 participants from Cohort 3 (NUC-1031 625 mg/m2), and 7 participants from Cohort 1 and 2 (NUC-1031 750 mg/m2). Three pharmacokinetic datasets were not obtained.

The mean infusion times for NUC-1031 500, 625, and 750 mg/m2 dose levels were 19.3 (range: 9–30), 14.7 (range: 5–26), and 26.6 (range: 14–30) minutes, respectively. Similar to PRO-001, AUC0-t and Cmax were used to compare pharmacokinetics between doses and in relation to potential interactions (8). Following administration, NUC-1031 achieved substantially higher concentrations than its metabolites dFdU and dFdC on day 1 as shown in Fig. 1.

Figure 1.

Pharmacokinetic profile of metabolites NUC-1031, dFdU, and dFdC. Mean dose-normalized NUC-1031, dFdC, and dFdU plasma concentrations from Cohort 2B (NUC-1031 500 mg/m2 and carboplatin AUC5) and Cohort 3 (NUC-1031 625 mg/m2 and carboplatin AUC4) using observed and predicted model concentrations, plotted on a semi-log scale. dFdC, difluorodeoxycytidine; dFdU, 2′,2′-difluorodeoxyuridine; dFdCTP, gemcitabine triphosphate; AUC, area under the plasma concentration–time curve.

Figure 1.

Pharmacokinetic profile of metabolites NUC-1031, dFdU, and dFdC. Mean dose-normalized NUC-1031, dFdC, and dFdU plasma concentrations from Cohort 2B (NUC-1031 500 mg/m2 and carboplatin AUC5) and Cohort 3 (NUC-1031 625 mg/m2 and carboplatin AUC4) using observed and predicted model concentrations, plotted on a semi-log scale. dFdC, difluorodeoxycytidine; dFdU, 2′,2′-difluorodeoxyuridine; dFdCTP, gemcitabine triphosphate; AUC, area under the plasma concentration–time curve.

Close modal

The mean AUC0–24 of NUC-1031 increased with increasing dose whereas the mean AUC0–24 of dFdC decreased with increasing dose. The mean AUC0–24 of dFdU was found to be similar between doses. The mean Cmax of NUC-1031 and dFdU increased between Cohort 2B (NUC-1031 500 mg/m2) and Cohort 3 (NUC-1031 625 mg/m2) but was lower in Cohort 1 and 2 (NUC-1031 750 mg/m2). The mean Cmax of dFdC decreased with increasing dose (Supplementary Table S2). The reduction in NUC-1031 and dFdU Cmax values in Cohort 1 and 2 can be explained by significantly longer infusion times for NUC-1031 in Cohort 1 and 2 (NUC-1031 750 mg/m2) with mean infusion time of 27 minutes versus 19 and 14 minutes for Cohorts 2B and 3, respectively, resulting in lower amounts of NUC-1031 infused per minute and subsequent reduced concentrations from which to calculate the Cmax (plasma and PBMC) values. Intracellular concentrations of the active anticancer metabolite dFdCTP remained high throughout the 24-hour pharmacokinetic sampling window with median Cmax values of 2.66, 4.16, and 3.22 μg/mL at 500, 626, and 750 mg/m2, respectively (Supplementary Appendix S1). Cmax values generated during the PRO-001 study at the 500 and 625 mg/m2 doses were very similar to those seen for PRO-002, with median values of 3.31 and 4.09 μg/mL, respectively, and higher for the 750 mg/m2 dose with a median Cmax of 8.44 μg/mL (8).

A comparison of the PRO-002 pharmacokinetic parameters with those generated during the PRO-001 study indicates that combination with carboplatin did not alter the pharmacokinetic profile of NUC-1031 (8). In Cohort 2B (NUC-1031 500 mg/m2), Cohort 3 (NUC-1031 625 mg/m2), and Cohort 1 and 2 (NUC-1031 750 mg/m2), the median concentrations of NUC-1031 AUC0–24 and Cmax were 118 μg/hour/mL and 412 μg/mL, 176 μg/hour/mL and 573 μg/mL, and 259 μg/hour/mL and 499 μg/mL, respectively. These values were similar to the median AUC0–24 and Cmax values generated following administration of 500, 625, and 750 mg/m2 NUC-1031 alone (PRO-001 study): 122 μg/hour/mL and 654 μg/mL, 161 μg/hour/mL and 419 μg/mL, and 272 μg/hour/mL and 718 μg/mL, respectively. These results demonstrate that there was no clinically relevant pharmacokinetic drug–drug interaction between NUC-1031 and carboplatin.

Efficacy

Radiological response to treatment was determined using CT scans conducted at weeks 9 and 18 and compared with baseline scans according to RECIST v1.1 (Table 6). Among the 23 response-evaluable patients, 1 (4%) had a confirmed response of a CR, 5 had confirmed PR (22%) resulting in an ORR of 26%, 8 (35%) had stable disease of >6 weeks duration, and 8 (35%) progressed on study. Four of the six responses occurred in Cohort 2B and two in Cohort 1. One patient in Cohort 3 achieved a PR by their 18-week CT scan but discontinued because of fatigue. CBR, defined as the proportion of patients with a best overall response according to RECIST v1.1 of CR, PR, and SD, was highest in Cohort 2B with 11/11 (100%) evaluable patients obtaining clinical benefit from treatment (Fig. 2). In terms of CA125 response, 5 patients were not evaluable as they did not have baseline CA125 levels ≥2 × ULN (35 U/mL). Of the remaining 20 who were CA125 response assessable, 11 (55%) patients showed a CA125 response: 10 (50%) with confirmed PR and 1 (5%) with confirmed CR (Fig. 3). The best change in CA125 occurred in Cohorts 2 and 2B with mean changes of 72% and 61%, respectively.

Table 6.

RECIST responses by cohort.

RECIST responses by cohort.
RECIST responses by cohort.
Figure 2.

Waterfall chart comparing the best change (%) from baseline of target lesions based on RECIST 1.1. The dashed lines represent the threshold for progressive disease at +20% and partial response at −30% from baseline. Twenty-two of the 23 evaluable patients are included in this graph as 1 patient in Cohort 2B did not have any target lesions. Cohort 1 (NUC-1031 750 mg/m2 and carboplatin AUC4), Cohort 2 (NUC-1031 750 mg/m2 and carboplatin AUC5), Cohort 2B (NUC-1031 500 mg/m2 and carboplatin AUC5) and Cohort 3 (NUC-1031 625 mg/m2 and carboplatin AUC4). AUC, area under the plasma concentration–time curve; RECIST, Response Evaluation Criteria In Solid Tumors version 1.1.

Figure 2.

Waterfall chart comparing the best change (%) from baseline of target lesions based on RECIST 1.1. The dashed lines represent the threshold for progressive disease at +20% and partial response at −30% from baseline. Twenty-two of the 23 evaluable patients are included in this graph as 1 patient in Cohort 2B did not have any target lesions. Cohort 1 (NUC-1031 750 mg/m2 and carboplatin AUC4), Cohort 2 (NUC-1031 750 mg/m2 and carboplatin AUC5), Cohort 2B (NUC-1031 500 mg/m2 and carboplatin AUC5) and Cohort 3 (NUC-1031 625 mg/m2 and carboplatin AUC4). AUC, area under the plasma concentration–time curve; RECIST, Response Evaluation Criteria In Solid Tumors version 1.1.

Close modal
Figure 3.

Waterfall chart showing best change (%) from baseline of CA125 levels in patients with CA125 > 2× ULN at baseline. The dashed line represents the threshold of 50% fall in CA125 compared to baseline (GCIG partial response). Cohort 1 (NUC-1031 750 mg/m2 and carboplatin AUC4), Cohort 2 (NUC-1031 750 mg/m2 and carboplatin AUC5), Cohort 2B (NUC-1031 500 mg/m2 and carboplatin AUC5), and Cohort 3 (NUC-1031 625 mg/m2 and carboplatin AUC4). AUC, area under the plasma concentration–time curve.

Figure 3.

Waterfall chart showing best change (%) from baseline of CA125 levels in patients with CA125 > 2× ULN at baseline. The dashed line represents the threshold of 50% fall in CA125 compared to baseline (GCIG partial response). Cohort 1 (NUC-1031 750 mg/m2 and carboplatin AUC4), Cohort 2 (NUC-1031 750 mg/m2 and carboplatin AUC5), Cohort 2B (NUC-1031 500 mg/m2 and carboplatin AUC5), and Cohort 3 (NUC-1031 625 mg/m2 and carboplatin AUC4). AUC, area under the plasma concentration–time curve.

Close modal

Of the 6 patients (22%) who had a confirmed RECIST v1.1 response in the full analysis set, 1 was platinum sensitive and 5 were partially platinum sensitive at study entry. Of the 8 patients with lasting SD, 7 had platinum-resistant and 1 had partially platinum-sensitive disease. Of the 8 patients who progressed between their 9 and 18 week scans, 7 were platinum resistant and 1 was partially platinum sensitive. Overall, all 23 evaluable patients (100%) included in the exploratory analysis had progression at the time of PFS analysis. The median PFS duration was 27.1 weeks (6.2 months; range: 5–46 weeks; 95% CI: 19.1–29.7).

Platinum-based chemotherapies like carboplatin and cisplatin remain the mainstay of treatment for ovarian cancer and, even in the current era of targeted therapies, the emergence of platinum resistance has adverse prognostic implications (12, 13). Strategies to extend or recover platinum sensitivity have been explored by combining platinum-based chemotherapy with other synergistic agents. However, combinations that display synergism in vitro can cause additive toxicities in the clinical setting. Gemcitabine, while synergizing with both carboplatin and cisplatin in cell lines and xenografts, causes severe dose-limiting myelosuppression when given with cisplatin in the clinic (18). Even in less heavily pretreated platinum-sensitive patients, carboplatin and gemcitabine are administered at reduced doses of AUC4 (days 1) and 1,000 mg/m2 (days 1 and 8), respectively, to lessen myelotoxicity (12, 13). More recently, this regimen has been superseded by carboplatin and pegylated liposomal doxorubicin (PLD), wherein carboplatin can be administered at AUC5 but at a longer cycle length of 4 weeks (19). When evaluated in patients with partially platinum-sensitive ovarian cancer in the CALYPSO study, the carboplatin and PLD combination yielded an ORR of 39% (20).

As preclinical studies showed NUC-1031 bypasses chemoresistance mechanisms, and a FIH phase I study showed it was less myelotoxic than gemcitabine, we questioned whether NUC-1031 could be repositioned in combination with AUC5 carboplatin for the treatment of recurrent partially platinum-sensitive and -resistant ovarian cancer. The majority of patients in PRO-002 had PROC and were heavily pretreated (8). At the RP2CD of 500 mg/m2 NUC-1031 and AUC5 carboplatin, treatment was well tolerated and all 11 evaluable patients in this cohort derived clinical benefit, 6 with best responses of PR and 5 with SD of >6 weeks. Interestingly, 5 of these 11 patients had previously received gemcitabine and 1 had progressed on a gemcitabine-containing regimen prior to entering PRO-002. These findings support preclinical studies showing that NUC-1031 overcomes resistance mechanisms associated with gemcitabine (7, 21). At this dose of NUC-1031, myelotoxicity was low grade and manageable, enabling the concomitant administration of carboplatin at AUC5. Other toxicities were also minimal; of note no lung toxicity was observed in any dosing cohorts. Although patients were precluded from the study if they had a history of allergy to platinum agents or gemcitabine, we did not observe any de novo allergic reactions to carboplatin or NUC-1031 in any of the study participants.

This study had limitations; it was small and the population, although mostly platinum resistant, was heterogenous as it contained patients with platinum-sensitive, partially platinum-sensitive, and resistant ovarian cancer. The regimen examined would need further evaluation in a larger study to more clearly compare and characterize disease response by platinum sensitivity and allow comparison with current standard-of-care regimens. In addition, to reflect standard of care, in PRO-002 radiological tumor assessments were scheduled after every 9 weeks (three cycles of treatment), longer than in comparable chemotherapy studies in which imaging is conducted every 4 weeks (10, 11). As a RECIST response is only confirmed if it is maintained for two consecutive scans, our reported rates of confirmed ORR and radiological PFS are probably conservative (10). Finally, at the time of this study, BRCA1/2 gene testing was only approved for patients with ovarian cancer who had an indicative familial or personal cancer history and was known for 13 (52%) of the study patients. Of the 23 response-evaluable patients, 6 were BRCA 1 or 2 germline carriers (26%), 6 were BRCA negative (26%), and 11 were BRCA unknown (48%). Among the patients with best responses, the two CRs were observed in 1 BRCA-negative patient and 1 with BRCA unknown, and the 7 PRs were seen in 4 BRCA-negative, 2 BRCA-unknown, and 1 BRCA-positive patients. Although numbers are small, this suggests there was no evidence of superior response to chemotherapy in the BRCA carriers, who were predominantly platinum resistant at the time of study entry. Instead, response was more aligned with PFI across the study participants. A comparison of the PRO-002 pharmacokinetic parameters with those generated following monotherapy (PRO-001 study) indicates that combining with carboplatin does not alter the profile of NUC-1031, suggesting there is no clinically relevant pharmacokinetic interaction between NUC-1031 and carboplatin. Overall, this study provides encouraging evidence to support the use of NUC-1031 in combination with carboplatin AUC5 in patients with recurrent ovarian cancer.

E. Ghazaly reports grants from Nucana plc during the conduct of the study. N.W. McCracken was employed with Nucama plc at the time of the writing of the article which provided the NUC-1031 compound and some financial support to help run the PRO-002 study. As a clinical pharmacologist and pharmacokinetic expert, N.W. McCracken wrote the pharmacokinetic section of the reports based on data that had been generated prior to starting at NuCana. At present N.W. McCracken is no longer in the employment of NuCana plc and has taken up another position in another company. D.J. Harrison reports salary from NuCana plc during the conduct of the study. S.P. Blagden reports grants from Nucana plc and other from Nucana plc during the conduct of the study. S.P. Blagden also reports other from Redx Pharma, UCB, Merck & Co, and Astex Pharma, as well as personal fees from Amphista Therapeutics and Ellipses outside the submitted work. No disclosures were reported by the other authors.

F. Kazmi: Writing–review and editing. S. Nicum: Data curation, investigation, project administration, writing–review and editing. R.L. Roux: Data curation, investigation, project administration, writing–review and editing. L. Spiers: Data curation, project administration, writing–review and editing. C. Gnanaranjan: Formal analysis, methodology, writing–review and editing. A. Sukumaran: Investigation, writing–review and editing. H. Gabra: Supervision, writing–review and editing. E. Ghazaly: Formal analysis, investigation, methodology, writing–review and editing. N.W. McCracken: Formal analysis, writing–original draft, writing–review and editing. D.J. Harrison: Formal analysis, methodology, writing–review and editing. S.P. Blagden: Conceptualization, supervision, investigation, methodology, writing–original draft, project administration, writing–review and editing.

The study was designed by the lead investigators and the sponsor (Imperial Healthcare NHS Trust). The sponsor collected and analysed the data in conjunction with the authors. Phastar provided statistical analysis support. This work was also supported by the BRC, ECMC, and CRUK Clinical Centres, and the Ovarian Cancer Action Research Centre, Imperial College London (London, United Kingdom). The study was funded and the investigational drug NUC-1031 was supplied by NuCana plc. We are grateful to Danielle McLean for her support in writing this article and the late Professor Chris McGuigan for synthesising the ProTide NUC-1031.

The study was sponsored by Imperial College London and funded by NuCana plc.

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.
Kroep
JR
,
Peters
GJ
,
Nagourney
RA
.
Clinical activity of gemcitabine as a single agent and in combination
.
In:
Peters
GJ
,
editor
.
Deoxynucleoside analogs in cancer therapy
.
New York: Humana Press
; 
2006
. p.
253
88
.
2.
Anderson
H
,
Hopwood
P
,
Stephens
R
,
Thatcher
N
,
Cottier
B
,
Nicholson
M
, et al
Gemcitabine plus best supportive care (BSC) vs BSC in inoperable non-small cell lung cancer–a randomized trial with quality of life as the primary outcome
.
Br J Cancer
2000
;
83
:
447
53
.
3.
Nagourney
RA
,
Link
JS
,
Blitzer
JB
,
Forsthoff
C
,
Evans
SS
. 
Gemcitabine plus cisplatin repeating doublet therapy in previously treated, relapsed breast cancer patients
.
J Clin Oncol
2000
;
18
:
2245
9
.
4.
Heinemann
V
,
Xu
Y-Z
,
Chubb
S
,
Sen
A
,
Hertel
LW
,
Grindey
GB
, et al
Cellular elimination of 2′, 2′-difluorodeoxycytidine 5′-triphosphate: a mechanism of self-potentiation
.
Cancer Res
1992
;
52
:
533
9
.
5.
Van Haperen
VWR
,
Veerman
G
,
Boven
E
,
Noordhuis
P
,
Vermorken
JB
,
Peters
GJ
. 
Schedule dependence of sensitivity to 2′, 2′-difluorodeoxycytidine (gemcitabine) in relation to accumulation and retention of its triphosphate in solid tumour cell lines and solid tumours
.
Biochem Pharmacol
1994
;
48
:
1327
39
.
6.
Raynal
C
,
Ciccolini
J
,
Mercier
C
,
Boyer
J-C
,
Polge
A
,
Lallemant
B
, et al
High-resolution melting analysis of sequence variations in the cytidine deaminase gene (CDA) in patients with cancer treated with gemcitabine
.
Ther Drug Monit
2010
;
32
:
53
60
.
7.
Slusarczyk
M
,
Lopez
MH
,
Balzarini
J
,
Mason
M
,
Jiang
WG
,
Blagden
S
, et al
Application of ProTide technology to gemcitabine: a successful approach to overcome the key cancer resistance mechanisms leads to a new agent (NUC-1031) in clinical development
.
J Med Chem
2014
;
57
:
1531
42
.
8.
Blagden
SP
,
Rizzuto
I
,
Suppiah
P
,
O'Shea
D
,
Patel
M
,
Spiers
L
, et al
Anti-tumour activity of a first-in-class agent NUC-1031 in patients with advanced cancer: results of a phase I study
.
Br J Cancer
2018
;
119
:
815
22
.
9.
Vijayananthan
A
,
Nawawi
O
. 
The importance of Good Clinical Practice guidelines and its role in clinical trials
.
Biomed Imaging Interv J
2008
;
4
:
e5
.
10.
Rustin
GJS
,
Vergote
I
,
Eisenhauer
E
,
Pujade-Lauraine
E
,
Quinn
M
,
Thigpen
T
, et al
Definitions for response and progression in ovarian cancer clinical trials incorporating RECIST 1.1 and CA 125 agreed by the Gynecological Cancer Intergroup (GCIG)
.
Int J Gynecol Cancer
2011
;
21
:
419
23
.
11.
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
.
12.
Pfisterer
J
,
Plante
M
,
Vergote
I
,
du Bois
A
,
Hirte
H
,
Lacave
AJ
, et al
Gemcitabine plus carboplatin compared with carboplatin in patients with platinum-sensitive recurrent ovarian cancer: an intergroup trial of the AGO-OVAR, the NCIC CTG, and the EORTC GCG
.
J Clin Oncol
2006
;
24
:
4699
707
.
13.
Berg
T
,
Nøttrup
TJ
,
Roed
H
. 
Gemcitabine for recurrent ovarian cancer-a systematic review and meta-analysis
.
Gynecol Oncol
2019
;
155
:
530
7
.
14.
Aapro
MS
,
Martin
C
,
Hatty
S
. 
Gemcitabine–a safety review
.
Anticancer Drugs
1998
;
9
:
191
201
.
15.
Veltkamp
SA
,
Hillebrand
MJ
,
Rosing
H
,
Jansen
RS
,
Wickremsinhe
ER
,
Perkins
EJ
, et al
Quantitative analysis of gemcitabine triphosphate in human peripheral blood mononuclear cells using weak anion-exchange liquid chromatography coupled with tandem mass spectrometry
.
J Mass Spectrom
2006
;
41
:
1633
42
.
16.
Beal
S
,
Sheiner
L
,
Boeckmann
A
,
Bauer
R
.
NONMEM user's guides (1989–2009)
.
Ellicott City, MD
:
Icon Development Solutions
; 
2009
.
17.
Fleming
TR
. 
One-sample multiple testing procedure for phase II clinical trials
.
Biometrics
1982
;
38
:
143
51
.
18.
Pinato
DJ
,
Graham
J
,
Gabra
H
,
Sharma
R
. 
Evolving concepts in the management of drug resistant ovarian cancer: dose dense chemotherapy and the reversal of clinical platinum resistance
.
Cancer Treat Rev
2013
;
39
:
153
60
.
19.
Holloway
RW
,
Grendys
EC
,
Lefebvre
P
,
Vekeman
F
,
McMeekin
S
. 
Tolerability, efficacy, and safety of pegylated liposomal doxorubicin in combination with carboplatin versus gemcitabine–carboplatin for the treatment of platinum-sensitive recurrent ovarian cancer: a systematic review
.
Oncologist
2010
;
15
:
1073
82
.
20.
Gladieff
L
,
Ferrero
A
,
De Rauglaudre
G
,
Brown
C
,
Vasey
P
,
Reinthaller
A
, et al
Carboplatin and pegylated liposomal doxorubicin versus carboplatin and paclitaxel in partially platinum-sensitive ovarian cancer patients: results from a subset analysis of the CALYPSO phase III trial
.
Ann Oncol
2012
;
23
:
1185
9
.
21.
Rizzuto
I
,
Ghazaly
E
,
Peters
GJ
. 
Pharmacological factors affecting accumulation of gemcitabine's active metabolite, gemcitabine triphosphate
.
Pharmacogenomics
2017
;
18
:
911
25
.