Nintedanib plus letrozole in early breast cancer: a phase 0/I pharmacodynamic, pharmacokinetic, and safety clinical trial of combined FGFR1 and aromatase inhibition

Abstract

Background: The combined use of a FGFR1 blocker and aromatase inhibitors is appealing for treating breast cancer patients with FGFR1 amplification. However, no pharmacodynamic studies have addressed the effects of this combined target modulation. We conducted a phase 0/I clinical trial in an adjuvant setting, with the goal of obtaining pharmacodynamic proof of the effects of combined aromatase and FGFR1 inhibition and to establish the RP2D for nintedanib combined with letrozole.

Patients and methods: Women with early-stage luminal breast cancer were eligible for enrollment in the study. Dose level 1 was nintedanib (150 mg/bid) plus letrozole (2.5 mg/day) administered for a single 28-day cycle (DLT assessment period), followed by a classic 3 + 3 schedule. FGF23 and 17-B-estradiol levels were determined on days 0 and 15; pharmacokinetic parameters were assessed on days 1 and 28. Patients were allowed to continue treatment for 6 cycles. The primary study endpoint was a demonstration of FGFR1 modulation (defined as a 25% increase in the plasma FGF23 level).

Results: A total of 19 patients were enrolled in the study (10 in the expansion cohort following dose escalation). At the RP2D (nintedanib 200 mg/bid plus letrozole 2.5 mg/day), we observed a 55% mean increase in the plasma FGF23 level, and 81.2% of the patients had no detectable level of 17-B-estradiol in their plasma (87.5% of the patients treated with letrozole alone). Nintedanib and letrozole displayed a pharmacokinetic interaction that led to three- and twofold increases in their respective plasma concentrations. Most G3 toxic events (5 out of 6: 2 diarrhea and 3 hypertransaminasemia) occurred subsequent to the DLT assessment period.

Conclusion: Combined treatment with nintedanib (200 mg/bid) plus letrozole (2.5 mg/day) effectively suppressed FGFR1 and aromatase activity, and these respective doses can be used as starting doses in any subsequent trials. However, drug-drug interactions may produce tolerability issues when these drugs are co-administered for an extended time period (e.g., 6 months). Patients enrolled in future trials with these drugs should be carefully monitored for their FGF23 levels and signs of toxicity, and those findings should guide individualized treatment decisions.

Trial registration: This trial was registered at www.clinicaltrials.gov under reg. # NCT02619162, on December 2, 2015.

Keywords: FGF23; FGFR1; Letrozole; Nintedanib; Pharmacodynamics; Phase 0 clinical trial; Phase I clinical trial.

Conflict of interest statement

MQF received research funds from the following companies: Novartis, Roche, Bayer, Boehringer-Ingelheim, Astra Zeneca, and MEI Pharma.

Figures

Fig. 1

Trial design. Patients completing the treatment for early hormone receptor-positive breast cancer that were eligible for letrozole treatment and were receiving it for a minimum of 4 weeks were enrolled in the trial. This design allowed testing whether letrozole adjuvant treatment was actually achieving its therapeutic goal (17-B-estradiol suppression) and whether the concurrent administration of nintedanib exerted any negative influence on it even in the absence of pharmacokinetic interactions between both drugs. Cycle 1 was mandatory and included pharmacokinetic profiling and pharmacodynamic endpoints (FGF23 and 17-B-estradiol suppression). Cycles 2 to 6 were included in order to provide the option for any potential benefit that a long-term administration of a multi-tyrosine-kinase inhibitor could imply in the adjuvant setting, as long as patients willingly decided to continue treatment and no toxicity was observed. After completing 1 to 6 cycles, patients continued on standard adjuvant letrozole treatment. And end-of-treatment (EOT) visit was scheduled 28 days after the last nintedanib dose

Fig. 2

Pharmacodynamic parameters. a Change of FGF23 plasma concentration from baseline to day + 15 in level 2 patients. Horizontal error bars: standard error. The average concentration (columns) shifted from 24.9 pg/mL (baseline) to 38.7 pg/mL (d + 15). Each dot represents the value of a single patient. *P < 0.05 (Wilcoxon). b Dot plot showing the relationship between BIBF1120-BS AUC and FGF23 on day 15 at the individual patient level. c Change of 17-B-estradiol levels from baseline to day + 15, the first day of the last cycle, and at the EOT visit. Each gray dot represents one patient with 17-B-estradiol levels below the LLD. Red dots represent the individual 17-B-estradiol levels of those patients that showed a concentration above the LLD. Regarding the patients below the LLD, although the exact levels might be lower, all of them are plotted at the level of Y = 5 pg/mL which was our LLD

Fig. 3

Pharmacokinetic parameters. a Plasma letrozole levels at both dose levels, comparing in each the mean plasma levels in nanograms per milliliter on day 29 versus 1 along the 24-h course. b, c The same as in a for both active nintedanib metabolites BIBF1120-BS and BIBF1202-ZW, respectively. N = 3 in level 1; N = 16 in level 2. Error bars: standard error