Safety profile of trifluridine/tipiracil monotherapy in clinical practice: results of the German compassionate-use program for patients with metastatic colorectal cancer

Colorectal cancer (CRC) is the third most common cancer in Germany [1]. During the last two decades, new combination chemotherapies (e.g with oxaliplatin, irinotecan, and flouropyrimidines) and the development of therapeutic monoclonal antibodies (e.g bevacizumab, cetuximab, panitumumab and ramucirumab) and other targeted agents (e.g. aflicercept, regorafenib) have notably prolonged the median survival time of pts. with metastatic CRC (mCRC). In recent clinical trials, the median overall survival (OS) from first-line therapy in metastatic disease has reached approximately 30 months [2‐8]. Nevertheless, the prognosis of patients, which are refractory or intolerant to all approved drugs is poor and there is still an unmet medical need for these patients [9‐13], especially for those who are in a good performance status and eligible for further therapies.

FTD/TPI is an oral antimetabolite and has been shown to be effective in the treatment of patients refractory or intolerant to approved drugs for mCRC [14]. The drug combines trifluridine, a thymidine-based nucleoside analogue, and tipiracil, which improves the bioavailability of trifluridine by inhibiting the enzyme thymidine phosphorylase, which is involved in its catabolism [15]. In the randomized global phase 3 trial RECOURSE FTD/TPI prolonged OS and progression free survival (PFS) compared to placebo with a favorable safety profile [14]. In the RECOURSE trial, 800 pts. with mCRC refractory or intolerant to fluoropyrimidine, irinotecan, oxaliplation, bevacizumab and to epidermal growth factor (EGFR) antibodies (in pts. with KRAS wild-type tumors) were randomly assigned in a 2:1 ratio to either FTD/TPI (35 mg/m² per dose twice daily on days 1 to 5 and 8 to 12, every four weeks) or placebo. In addition in both study arms patients received best supportive care (BSC). FTD/TPI, significantly prolonged median OS compared to placebo (7.1 vs. 5.3 months; hazard ratio (HR) 0.68; p < 0.0001). In addition FTD/TPI significantly prolonged median PFS; 2.0 vs. 1.7 Months; HR 0.48; p < 0.0001), improved disease control rate ((DCR; 44.0% vs. 16.3%; p < 0.0001) and prolonged median time to deterioration of ECOG performance status compared to placebo. FTD/TPI was well tolerated in the RECOURSE trial. The most overall common adverse events (AE) were leukopenia, neutropenia, anemia and thrombocytopenia and the most common AE ≥ grade 3 was neutropenia (38% of patients treated with FTD/TPI).

However, data from clinical trials does not always reflect clinical experience in a real-life setting. In RECOURSE, for example, patients with ECOG performance status ≥2 were excluded and the rate of patient who had received regorafenib prior to FTD/TPI was only 18%. This is explained by the lack of approval of regorafenib at the time point of initiation of the RECOURSE trial. Thus, little information is available regarding the safety profile of FTD/TPI for patients with poor performance status or patients that had been pretreated with regorafenib.

To collect more information about the usage and tolerability of FTD/TPI in clinical practice, we prospectively investigated the characteristics of patients with mCRC refractory or intolerant to standard chemotherapies treated with FTD/TPI monotherapy within a compassionate use program (CUP), which was conducted from January 12 to August 14, 2016 in designated CRC centers in Germany and compared the rate and severity of AE to the data from the phase III trial RECOURSE. CUPs provide treatment options for patients with unmet medical needs and offer the opportunity to obtain early data on efficacy, safety and use of a new drug in a real-world setting. In our analysis, we differentiated between patients who had received regorafenib prior to treatment with FTD/TPI and those who had not.

In our prospective CUP conducted in heavily pretreated patients with mCRC in Germany, we found that patient characteristics as well as the safety profile of FTD/TPI were comparable with those reported in the pivotal RECOURSE trial [14]. Mean age of patients, gender distribution, KRAS mutational status of patients as well as time from diagnosis of colorectal cancer was consistent between the German CUP and the phase 3 trial, though the mean body surface area in the CUP was slightly higher (1.87 vs. 1.781, respectively) leading to a higher daily dose of FTD/TPI (130 mg) compared with the RECOURSE trial (120 mg) (suppl. Table 1). However, in comparison with the RECOURSE trial, the patient population within the German CUP had two major differences indicating that phase 3 trials do not always reflect clinical experience in a real-life setting. First, in our study 9.8% of patients with an ECOG performance status of ≥2 were included whereas patients with poor performance status were excluded in the RECOURSE trial (0%). Thus, our cohort represents a more “real” patient population in daily clinical practice. We do not observe any remarkable difference in the safety profile of FTD/TPI or in the duration of treatment in patients with an ECOG PS ≥2. Among the 22 patients with ECOG PS ≥2 only 3 discontinued treatment with FTD/TPI due to AEs. Thus FTD/TPI seems to be tolerated and effective even in patients with poorer performance status. We did not detect any unexpected safety signals with comparable characteristics and incidence of AEs and SAEs. However, the incidence of hematological AEs and other laboratory abnormalities was lower compared to the RECOURSE trial, presumably due to a underreporting of these AEs in CUPs. This is in line with the findings of a postmarketing surveillance study in 3.420 patients treated with FTD/TPI from May to November 2014 in Japan [17]. In this large observational study the safety profile was also similar to those from the RECOURSE trial. Furthermore, an expanded-access program (EAP) in the USA was carried out to further assess the safety profile of FTD/TPI in a real-world setting in 549 US patients with refractory mCRC. In this EAP, patients had a comparable exposure duration to that reported in 64 US patients who had participated in the RECOURSE trial, with no unexpected safety concerns [18]. In the EAP, only 4% of patients discontinued the treatment with FTD/TPI due to AEs [18]. Finally, a recent Spanish CUP conducted with refractory 538 mCRC patients supported the findings of the German CUP presented here. In the Spanish CUP, FTD/TPI was generally well tolerated and the real-world data analysis was consistent with that reported in phase 3 trials of FTD/TPI in patients with mCRC [19]. In contrast to other European countries regorafenib is not longer available for patients in Germany since 2016. For this reason the number of patients in the German CUP, which were pretreated with regorafenib, is significantly lower than in patients of other European CUPs published recently. For example 70% of patients in the Italian CUP were pretreated with regorafenib compared to only 33.6% in the here reported patient population [20]. In addition, in Germany patients with advanced CRC are not only treated in large academic centers but also in smaller hospitals or private practices. For this reason 118 centers participated in the German CUP and enrolled 226 patients in total. So, there was a high variability in the experience of treating advanced CRC among the physicians. In contrast in other European countries like the Netherland the treatment of patients with advanced cancers is more centralized. In line, in the Netherlands’ CUP 148 patients were enrolled in only 17 centers [21]. Taken together, the here reported patient population of the German CUP differs from those previously reported.

The data from this CUP is limited because long-term follow-up to analyze survival outcomes of the patients was not allowed due to regulatory affairs. Nevertheless, our study confirms findings of previous data that the safety profile of FTD/TPI was comparable to the pivotal trial RECOURSE and that the drug is well tolerated in a real life setting [18, 19, 22]. Moreover our data show that patients with an ECOG performance status of ≥2 and/or previously treated with regorafenib can recieve FTD/TPI safely and effectively. Further clinical trials should investigate these patient groups with respect to efficacy as well as safety with a focus on the sequence of FTD/TPI and regorafenib.

In a prospective CUP in Germany conducted with pretreated mCRC patients the safety profile of FTD/TPI was investigated in a real-world setting. FTP/TPI was well tolerated in clinical practice. The patient characteristics and the toxicities of FTD/TPI documented in the German CUP were comparable with those reported in the pivotal RECOURSE trial without any unexpected safety signals. However, in the German CUP a higher number of patients had an ECOG performance status of ≥2 (9.7% vs. 0% in RECOURSE) and the previous use of regorafenib was more common in the real-life setting (33.6% vs.18% in RECOURSE).