Subgroup analyses from patients with pre-treated metastatic colorectal cancer receiving trifluridine/tipiracil: results of the TALLISUR trial

Background In the pivotal phase III RECOURSE trial, trifluridine/tipiracil (FTD/TPI) improved progression-free and overall survival (PFS, OS) of patients with pre-treated metastatic colorectal cancer (mCRC). Subsequently, the TALLISUR trial provided post-authorisation efficacy and safety data and patient-reported outcomes on quality of life (QoL) in a German patient cohort. The present analysis reports the final data on efficacy, safety and QoL and investigates the impact of baseline characteristics and associated prognostic subgroups on outcome. Methods In this prospective, multi-centre, Germany-wide, phase IV study, patients with pre-treated mCRC were given the choice to receive either FTD/TPI or best supportive care (BSC). To assess the primary endpoint, QoL, EORTC QLQ-C30 questionnaires were employed. Secondary endpoints included QoL assessed through EQ-5D-5L questionnaires, OS, PFS and safety. Additionally, 3 subgroups were defined according to a post-hoc analysis of the RECOURSE trial: best, good and poor prognostic characteristics (BPC, GPC, PPC). Patients with < 3 metastatic sites at inclusion and/or ≥ 18 months from diagnosis to inclusion were considered to have GPC. GPC patients without liver metastasis at inclusion were considered to have BPC. All remaining patients were considered to have PPC. Results Of 195 patients, 186 decided to receive FTD/TPI and 9 to receive BSC. The low number of patients in the BSC-arm did not allow statistically meaningful analyses. Treatment with FTD/TPI was associated with maintained QoL. For all patients, median OS was 6.9 months (95% CI 6.1 – 8.3) and for the defined subgroups (BPC n = 20 vs GPC n = 65 vs PPC n = 121) 12.2, 7.9 and 6.8 months (95% CI 6.0 – 18.2, 6.2 – 13.3, 5.4 – 8.1). The most frequent TEAEs were neutropenia (29.6%), anaemia (24.7%) and nausea (23.7%). Febrile neutropenia occurred in 1.1%. Conclusions Treatment of patients suffering from pre-treated mCRC with FTD/TPI was associated not only with prolonged survival and delayed progression, but also with maintained QoL. Independent of other baseline characteristics such as ECOG performance status and age, low metastatic burden and indolent disease were factors associated with favourable outcome. Clinical trial registration EudraCT-Number 2017–000292-83, first registration 19/06/2017. Supplementary Information The online version contains supplementary material available at 10.1186/s12885-024-12599-7.


Background
Trifluridine/tipiracil (FTD/TPI) is approved by the European Medicines Agency as monotherapy for the treatment of adult patients with mCRC who have been previously treated with or are not considered candidates for available therapies including fluoropyrimidine-, oxaliplatin-and irinotecan-based chemotherapies, anti-VEGF and -EGFR agents [1].
This approval was granted based on results of the pivotal phase III RECOURSE trial [2,3].Patients with pre-treated mCRC (limited to Eastern Cooperative Oncology Group performance status (ECOG PS) 0 or 1) received either FTD / TPI (n = 534) or placebo (n = 266) plus best supportive care (BSC).By adding FTD / TPI to BSC, median overall survival (mOS; 7.2 vs 5.2 months; hazard ratio (HR) 0.69, p < 0.0001) and median progression-free survival (mPFS; 2.0 vs 1.7 months; HR 0.48, p < 0.001) were significantly improved.The toxicity profile was characterised by haematological side effects with neutropenia (38%) and leukopenia (21%) being most frequent.Febrile neutropenia was observed in 4% of patients receiving FTD / TPI.Effect of FTD / TPI on health-related quality of life (QoL) was not assessed based on patient-reported outcomes.However, time to deterioration of the ECOG PS from 0 / 1 to ≥ 2 was significantly longer in patients treated with FTD / TPI (5.7 vs 4.0 months; HR 0.66, p < 0.001).
In a post-hoc exploratory analysis, the effect of prognostic factors on RECOURSE efficacy outcomes was examined [4].To this end, 3 subgroups were defined: best, good and poor prognostic characteristics (N = 800; BPC, n = 153; GPC, n = 386; PPC, n = 414).Patients with < 3 metastatic sites at inclusion and/ or ≥ 18 months from diagnosis to inclusion were considered to have GPC.GPC patients without liver metastasis at inclusion were considered to have BPC.All remaining patients were considered to have PPC.Here, mOS was longer in patients with BPC and GPC compared to patients with PPC (16.4 vs 9.3 vs 5.3 months).
The phase IV TALLISUR (Trifluridine / tipirAcil quaLity of LIfe StUdy in mCRC patients) was a prospective, interventional, multi-centre, Germany-wide, open-label and non-randomised study (EudraCT-Number 2017-000292-83, first registration 19/06/2017).It was designed to generate post-authorisation efficacy and safety data and to assess QoL under FTD / TPI treatment.Patients with pre-treated mCRC were briefed by the treating physician to make an informed decision on either receiving FTD / TPI or solely BSC.
Here, we present the final efficacy data as well as patient-reported QoL results of the TALLISUR study stratified according to prognostic factors or subgroups like age and sidedness complementing previously published results of an interim analysis [5].

Methods
Methods have been described before [5].Briefly, FTD / TPI was administered orally twice daily on days 1 to 5 and days 8 to 12 of each 28-day cycle as long as a benefit was observed or until unacceptable toxicity occurred.Based on body surface area, the dose was 35 mg/m 2 given twice daily and was not allowed to exceed 80 mg per dose.Depending on individual safety and tolerability, a maximum of three dose reduction levels (30, 25 and 20 mg/m 2 ) were permitted.In the event of toxicities, the dose interruption, resumption and reduction criteria as stated in the SmPC were followed.
The primary endpoint was pre-defined as the rate of responders with stabilised (> -10 and < 10 scores) or improved (≥ 10 scores) QoL response (QoL-RR).Response was calculated as the mean score of the EORTC QLQ-C30 global health status / QoL scale from the second cycle until the end of treatment / observation compared to the baseline score.
Questionnaires were scheduled to be filled in within 2 days before or on the first day of any treatment / observation cycle.In order to obtain additional information, an extended time period was analysed including questionnaires being filled in between the day after the last administration (FTD / TPI-group) / day 12 (BSC-group) of the previous cycle and the first day of the respective cycle.
Based on baseline characteristics, 3 subgroups were defined according to Supplementary Table 1 as previously described by Tabernero and colleagues [4].Patients with good prognostic characteristics (GPC) were defined as patients with low metastatic burden (1 or 2 metastatic sites at randomization) and less indolent disease (≥ 18 months from diagnosis of first metastasis).Patients with GPC without liver metastases were defined as patients with best prognostic characteristics (BPC) Patients with poor prognostic characteristics (PPC) were defined as patients with high metastatic burden (≥ 3 metastatic sites at randomization) and/or aggressive disease (< 18 months from diagnosis of first metastasis).
Tumour response was assessed by imaging procedures (magnetic resonance imaging or computed tomography scan) and evaluated according to Response Evaluation Criteria in Solid Tumours (RECIST) version 1.1 [10].

Patient enrolment and demographics
Between 22 September 2017 (first patient included) and 08 January 2019 (last patient included), 195 patients entered the trial across 44 sites in Germany (Fig. 1).
At the time of the interim analysis, 1 patient was wrongly documented to not have received therapy [5].This patient was now included in this final analysis.Except for one study centre, each centre included at least 1 patient who received at least one FTD / TPI treatment.One centre included only one patient for BSC treatment who did not start close observation in the end.
Due to freedom of choice of each patient to receive active therapy or BSC, the two study arms were imbalanced with respect to the number of patients (186 in the FTD / TPI-group vs 9 in the BSC-group).Therefore, it was not possible to conduct any statistically meaningful analyses using data obtained from patients receiving BSC.
The median duration of FTD / TPI treatment was 68 days (range 1 -719 days) and the median number of cycles was 3 (range 1 -23 cycles). 2 patients (1.1%) received a maximum of 23 cycles.The median cumulative FTD / TPI dose administered was 657.4 mg/m 2 (range 32.5 -743.9 mg/m 2 ) and the median relative dose intensity was 98.1% (range: 4.7 -110.0%).Throughout the study, 72 patients (38.7%) had at least one cycle with a dose reduction to < 90% of the target dose.During the treatment period, 12 patients (6.5%) received at least one dose of granulocyte colony-stimulating factor.
Baseline characteristics and demographics are shown in Table 1.
Patients in the FTD / TPI and BSC-group were median 67 (range 40 -88) or 78 (range 54 -82) years old and 62.9% or 55.6% male, respectively.89.2% of patients in the FTD / TPI-group and 44.4% of patients in the BSCgroup presented an ECOG PS 0 -1.Primary tumours were located more often on the left than on the right side of the colon (74.2% vs 22.6% in the FTD / TPI-and 66.7% vs 22.2% in the BSC-group).While most patients were characterised by multiple metastatic sites, 23 patients presented a single metastatic site (15 patients with liver only metastasis, 6 patients with lung only metastasis).52.6% and 68.4% of patients with ≤ 2 metastatic sites also had a lung or liver metastasis, respectively, compared to 92.3% and 93.4% of patients with > 2 metastatic sites.Most patients received FTD / TPI as a third (38.2%) or fourth (26.3%) line therapy option.Nevertheless, 31 (16.7%)patients received FTD /TPI as first or as second line therapy.The baseline characteristics of these patients are described in more detail in Supplementary Table 2.After end of treatment with FTD / TPI, 36% of patients received subsequent therapy.

Quality of life
Return rates for EORTC QLC-C30 and EQ-5D-5L are shown in Supplementary Table 3.
Questionnaires from 106 patients receiving FTD / TPI were evaluable for the primary endpoint, QoL-RR.The primary endpoint was reached for FTD / TPI-treated patients with 58.5% QoL-RR (95% CI 48.5 -68.0).When allowing for the above-mentioned extended period to fill in the questionnaires, QoL-RR was reported even higher with 64.2% (95% CI 54.3 -73.2).
Mean score changes from baseline of selected EORTC QLQ-C30 items up to cycle 6 are shown in Fig. 2 illustrating a maintained QoL during FTD / TPI treatment.
The estimated median time to deterioration of the EORTC QLQ-C30 global health status / QoL score by at least 10 scores was 121 days (95% CI 84.0 -172.0).Similarly, the estimated median time to deterioration of the EQ VAS was 113 days (95% CI 85.0 -140.0).
Efficacy data obtained from patients receiving BSC is summarised in Supplementary Table 5.
Response rates of patients treated with FTD / TPI are shown in Supplementary Table 6.Objective response rate (ORR) was 2.2% and disease control rate (DCR) was 27.4%.

Subgroup analyses
Regarding the primary endpoint, QoL-RR, subgroup analyses are shown in Table 2.
Results were comparable between patients with BPC, GPC and PPC.Also, patients with right-and left-sided primary tumour had a comparable QoL-RR.Similarly, QoL-RR of patients with synchronous and metachronous metastases did not differ greatly.Also, patients who had been diagnosed < 18 compared to ≥ 18 months ago had similar QoL-RR.Furthermore, neither the number of metastatic sites nor the number of previous lines of therapies affected QoL-RR.Patients with ECOG PS ≥ 2 presented a comparable QoL-RR compared to patients with ECOG PS < 2. Interestingly, however, male patients had a slightly lower QoL-RR compared to female patients.
Regarding OS, a univariate Cox regression analysis was performed based on patients' baseline characteristics (Supplementary Table 7).Two characteristics showed statistical significance: ECOG 0 -1 compared to ECOG 2 -3 and diagnosis of metastatic disease < 18 compared to ≥ 18 months ago.However, according to a multivariate Cox regression analysis, no statistical significance was observed (Supplementary Table 8).
Survival was comparable between patients with synchronous and metachronous metastases (Fig. 3b), patients with right and left sided primary tumour (Fig. 3c) and between younger and elderly patients (Fig. 3d).Also, similar OS was observed between patients who received FTD / TPI as first line therapy (mOS 6.0 months, n = 8) or after 1 or 2 previous lines of therapy (mOS 6.5 months, n = 94).The median duration of treatment for each subgroup is shown in Supplementary Table 9.

Discussion
The aim of the TALLISUR study was to complement the pivotal phase III RECOURSE trial with post-authorisation efficacy and safety data and patient-reported outcomes on QoL from German patients [2,3].
Here, we reported the final efficacy, safety and QoL results which were in line with results from an interim analysis published before [5].Additionally, we presented the survival and QoL results of comprehensive subgroup analyses.
In a post-hoc exploratory analysis of the RECOURSE trial, the effect of prognostic factors were investigated and 3 subgroups were defined: BPC, GPC, PPC [4].Low metastatic burden (≤ 2 metastatic sites at inclusion) and indolent disease (≥ 18 months from diagnosis of metastatic disease to inclusion) were factors of good Interestingly, QoL-RR was comparable between all three subgroups.This observation is particularly important, since QoL-RR is the major objective when applying last line treatment.The reason for this observation remains unclear, however.Although the RECOURSE trial reported a longer mPFS and mOS for patients treated in fourth and fifth line compared to third line, no difference was observed in regards of the number of metastatic sites (1 -2 vs ≥ 3).However, QoL was not assessed in RECOURSE at all.Whereas patients with BPC profited the most, patients from all prognostic subgroups evidently benefitted from treatment with FTD / TPI.Time of onset of metastasis (synchronous vs metachronous), localisation of primary tumour (right vs left) and age at inclusion (≥ 65) did not affect efficacy of FTD / TPI nor its effect on QoL.Male patients had a slightly lower QoL-RR compared to female patients.This could also be explained by male patients generally being older than female patients (62.5% of patients ≥ 65 years old were male) when enrolling into the study.
When interpreting the QoL results, one must bear in mind that the return rates of questionnaires were higher when considering the extended period (Supplementary Table 3).This also illustrates that especially with late-stage cancer patients, formal assessment of patientreported outcomes is typically an arduous task.
TEAEs observed in TALLISUR were in line with the safety profile of FTD / TPI [2,3].This is particularly relevant for TEAEs believed to likely affect QoL, such as nausea, vomiting, diarrhoea, and fatigue [12].
As indicated by these subgroup analyses, FTD / TPI is a viable option for virtually all patients who are not considered candidates for other available therapies.This is also stressed by 16.7% of patients receiving FTD / TPI as first or as second line therapy.
The TALLISUR results are in line with previous research investigating the effect of FTD / TPI on QoL [13][14][15].While stabilising disease, FTD / TPI is also associated with maintaining QoL, two crucial objectives in the late stage setting of mCRC [16,17].
Recently, FTD / TPI has been approved by EMA in combination with bevacizumab for the treatment of mCRC patients after two prior lines of therapy based on the pivotal phase III SUNLIGHT trial [18].In line with the TALLISUR results, QoL was maintained for patients in the monotherapy arm as well as for patients in the combination arm [19].
Based on the FRESCO-2 trial, fruquintinib has been approved by EMA for treatment of mCRC patients who have been previously treated with available standard therapies and who have progressed on or are intolerant to treatment with either FTD / TPI or regorafenib [20].Here, patients who had been treated with FTD / TPI and/or regorafenib were randomised to receive either fruquintinib or placebo.This situation does not reflect the situation of the TALLISUR study, however.Beyond therapy with FTD / TPI monotherapy or regorafenib monotherapy, QoL was not negatively impacted by fruquintinib therapy [21].

Conclusions
In summary, the TALLISUR trial provided patientreported QoL and post-authorisation efficacy and safety data from German patients with pre-treated mCRC receiving FTD / TPI.Overall, treatment with FTD / TPI prolonged survival and delayed progression with a manageable toxicity profile while concomitantly being associated with maintained QoL.

Table 1
Patient baseline characteristics and demographics

Table 2
Quality of life by subgroups