Clinical and quality of life outcomes with trifluridine/tipiracil: PRECONNECT and TALLISUR studies

Abstract

Background: QLQ-C30 Global Health Status (GHS) and Eastern Cooperative Oncology Group performance status (ECOG PS) data from PRECONNECT and TALLISUR studies were pooled. Materials & methods: Association between changes in ECOG PS and QLQ-C30 GHS in patients with metastatic colorectal cancer (mCRC) receiving trifluridine/tipiracil (FTD/TPI) was evaluated using Cox regression analysis. Results: 1100 patients were included. There was no clinically relevant change from baseline in QLQ-C30 GHS score through cycle 7. 63.0% of patients maintained/improved ECOG PS. The presence of liver metastasis increased the risk of QLQ-C30 GHS score/ECOG PS deterioration. The association between time to ECOG PS deterioration and change in QLQ-C30 GHS score over time was significant (HR 1.71 [95% CI: 1.4, 2.2]). Conclusion: ECOG PS and QLQ-C30 GHS scores were maintained during FTD/TPI treatment, and these measures were associated. Factors that increased the risk of deterioration included the presence of liver/lung metastasis, younger age and shorter time since first metastasis.

Clinical Trial Registration: PRECONNECT (EudraCT Number: 2016-002311-18) and TALLISUR (EudraCT-Number: 2017-000292-83).

Tweetable abstract

Association between time to Eastern Cooperative Oncology Group performance status deterioration and change in QLQ-C30 Global Health Status score over time found in patients receiving FTD/TPI.

Keywords::

• metastatic colorectal cancer
• performance status
• quality of life
• trifluridine/tipiracil

Recommended first and second-line treatment for patients with unresectable metastatic colorectal cancer (mCRC) is fluoropyrimidine-based chemotherapy (e.g., 5-FU or capecitabine), often in combination with oxaliplatin or irinotecan [1]. After failure of two lines of therapy, approximately 42% of patients go on to receive further lines of therapy [2]. Treatment options for patients in the third-line setting include regorafenib, trifluridine/tipiracil (FTD/TPI), EGFR inhibitors (if not previously used), and anti-programmed cell death protein 1 inhibitors (for microsatellite instability-high mCRC) [1]. While prognosis is poor for patients initiating third-line treatment, evidence from randomized controlled trials for improved outcomes has emerged in recent years for regorafenib and FTD/TPI. In the CORRECT trial in patients with mCRC who had progressed on standard treatment, regorafenib improved median overall survival (mOS) compared with placebo (6.4 and 5.0 months, respectively) and median progression-free survival (mPFS) was 1.9 and 1.7 months, respectively [3]. In the pivotal RECOURSE study, FTD/TPI (approved for use in patients with mCRC who have progressed on, or who cannot tolerate, standard therapies) improved mOS (7.1 and 5.3 months, respectively) and mPFS (2.0 and 1.7 months, respectively) compared with placebo [4,5]. FTD/TPI also delayed the time to performance status deterioration (5.7 months with FTD/TPI vs 4.0 months with placebo) (as measured by the Eastern Cooperative Oncology Group performance status [ECOG PS]) [4,5]. In the observational PRECONNECT study, mPFS was 2.8 months and median time to ECOG PS deterioration was 8.7 months, largely confirming clinical trial findings in a real-world setting [6]. The recent phase 3 SUNLIGHT trial showed an improved mOS in with FTD/TPI + bevacizumab compared with FTD/TPI (10.8 and 7.5 months, respectively) in patients with mCRC who had received two previous chemotherapy regimens, and this combination may become the new standard of care in this setting [7].

Given the importance of maintaining quality of life (QoL) in advanced stage cancer patients who have likely received multiple treatment modalities, it is essential that late-stage cancer therapies demonstrate positive effects on patient-reported QoL outcomes alongside traditional clinical outcomes [8]. Therefore, specific QoL measures have been included in open-label and real-world studies of FTD/TPI, and as well as confirming the safety and efficacy data of randomised controlled trials, these studies showed that patient-reported QoL was maintained or improved in most patients [6,9–11]. PRECONNECT was an international, multicentre, open-label, Phase IIIb trial, designed to further characterize the safety, efficacy and QoL of FTD/TPI in adults with mCRC (NCT03306394) [6,10]. QoL was measured by the European Organization for Research and Treatment of Cancer Quality of Life (EORTC QLQ-C30) health questionnaire [12], and preliminary data indicated that Global Health Status (GHS) scores showed no clinically relevant changes from baseline among patients initiating treatment with FTD/TPI [6,10]. GHS score was improved or maintained in 55.9% of patients [10]. Country specific post-hoc analysis of the PRECONNECT study have been published from France, Italy, Australia, Turkey and Brazil and confirmed the outcomes with FTD/TPI in a third line setting in these countries [13–17]. TALLISUR was a prospective, multicentre, open-label Phase IV trial of FTD/TPI to evaluate the QoL (EudraCT: 2017-000292-83) [9]. This study also used EORTC QLQ-C30 to assess QoL changes and showed no deterioration in mean QoL scores after initiation of FTD/TPI treatment [9].

Multiple factors can have a negative effect on patient QoL, including disease progression, toxicity of treatment, co-morbidities, and surgical recovery [18]. In this analysis, we present pooled QoL data from PRECONNECT and TALLISUR to explore if patient-reported outcomes measures assessed with these questionnaires can be used as prognosis factors, and if there is an association between patient-reported QoL and changes in clinician-reported outcomes measures (ECOG PS).

Materials & methods

Study design, patients & treatment

This was a pooled analysis of QoL data from PRECONNECT (EudraCT Number: 2016-002311-18) and TALLISUR (EudraCT-Number 2017-000292-83); a detailed description of the methods from both trials have been reported previously [6,9,10]. For both studies, all relevant Institutional Review Board/Independent Ethics Committee approvals were gained before commencement. The studies were conducted in accordance with the ethical principles stated in the Declaration of Helsinki. All patients gave written informed consent before participation.

In both studies, patients with mCRC (refractory or unsuitable for standard therapies) received FTD/TPI (35 mg/m² twice-daily) on days 1–5 and 8–12 of each 28-day cycle between 2016 and 2020. Patients completed the EORTC QLQ-C30 (Version 3.0) questionnaire at each treatment cycle in both studies. The QLQ-C30 questionnaire consists of five functional scales (physical, role, cognitive, emotional, and social), three symptom scales (fatigue, pain, and nausea and vomiting), six single items (dyspnoea, loss of appetite, insomnia, constipation, and diarrhoea) and perceived financial impact of the disease, and one GHS scale [12]. A high score for a functional scale represents a healthy level of functioning, a high score for the GHS represents a high QoL, but a high score for a symptom scale represents a high level of problems. ECOG PS was evaluated at baseline and during each cycle.

Analyses

QoL data were pooled for all patients who received ≥1 dose of FTD/TPI and answered the QLQ-C30 questionnaire at baseline and at least once post baseline. Only treatment cycles with ≥10% of the pooled patient population were included in the analysis. A clinically relevant change in QoL was defined as ≥10-point change in QLQ-C30 score from baseline, or death. ECOG PS at baseline versus final value ECOG PS on treatment was assessed, and time to ECOG PS deterioration (to ≥2) was analysed using Kaplan Meier methods.

For QLQ-C30 scores, analyses included description of mean changes from baseline, a Cox proportional hazard model with the ECOG PS at baseline as covariate on time until definitive QLQ-C30 GHS deterioration, as well as a multivariable Cox proportional hazard model on time until definitive QLQ-C30 GHS deterioration. Multivariable Cox proportional hazard model analyses to define potential prognostic factors for deterioration in QLQ-C30 GHS and ECOG PS were performed using a stepwise selection process (including/retaining factors significant at the 10% level; Wald chi-square test). The following variables were included in the model: time from first metastasis to first study drug intake (<18 vs ≥18 months); number of metastatic sites (1–2 vs ≥3); ECOG PS (0 vs 1/2); ECOG PS (0/1 vs 2); liver metastasis (presence vs absence); lung metastasis (presence vs absence); lymph node metastasis (presence vs absence); KRAS status (mutant vs wild-type/unknown); KRAS status (unknown vs wild-type/mutant); BMI (<18.5 kg/m² vs ≥18.5 kg/m²); BMI (<25 kg/m² vs ≥25 kg/m²); BMI (<30 kg/m² vs ≥30 kg/m²); primary tumor site (right vs left); gender (male vs female); age (<70 vs ≥70 years); and previous drug treatment with anti-VEGF (presence vs absence). The association between the changes in QoL over time and the time to deterioration of ECOG PS to ≥2 was investigated using a Cox proportional hazard model with time-dependent covariate.

Efficacy data were pooled for all patients who received ≥1 dose FTD/TPI. Objective response rate (ORR; defined as a best overall response rate of complete or partial response), disease control rate (DCR; defined as a best overall response rate of complete or partial response, or stable disease), and PFS (defined as the time from first administration of FTD/TPI to first radiological or clinical tumor progression, or death) were assessed. The survival function of the time-dependent parameters was estimated via Kaplan-Meier method and 95% confidence intervals (CI) for median durations.

Results

Patient population

In total, 1100 patients (914 patients from PRECONNECT; 186 from TALLISUR) received ≥1 dose of FTD/TPI and 975 received ≥1 dose of FTD/TPI and completed a baseline and ≥1 post-baseline QLQ-C30 questionnaire. Most patients had a baseline ECOG PS score of 0 (n = 506) or 1 (n = 545); 20 patients had a score of 2 and for 29 patients the score was not reported (Table 1). Compared with the subgroup of patients with an ECOG PS score of 0, the subgroups of patients with ECOG PS scores of 1 or 2 were more likely to be female, were older, and had lower/worse QLQ-C30 GHS scores (Table 1).

Table 1. Baseline characteristics of the pooled population from PRECONNECT and TALLISUR according to ECOG status at baseline.

Characteristic ^†	All (n = 1100)	ECOG 0 (n = 506)	ECOG 1 (n = 545)	ECOG 2 (n = 20)
Median (min–max) age, years	63 (24–89)	62 (24–83)	64 (28–89)	69 (35–85)
Gender
Female	442 (40.2)	185 (36.6)	234 (42.9)	10 (50.0)
Male	658 (59.8)	321 (63.4)	311 (57.1)	10 (50.0)
Primary tumor site
Right colon	286 (26.0)	128 (25.3)	149 (27.4)	3 (15.0)
Left colon	730 (66.4)	346 (68.4)	349 (64.2)	15 (75.0)
Not specified	83 (7.6)	32 (6.3)	46 (8.5)	2 (10.0)
Time from first metastasis to first FTD/TPI intake
<18 months	206 (19.2)	84 (17.1)	113 (21.2)	6 (37.5)
≥18 months	865 (80.8)	408 (82.9)	421 (78.8)	10 (62.5)
Number of metastatic sites
<3	832 (75.6)	385 (76.1)	414 (76.0)	12 (60.0)
≥3	268 (24.4)	121 (23.9)	131 (24.0)	8 (40.0)
RAS mutation-positive	577 (52.5)	274 (54.2)	275 (50.5)	10 (50.0)
BRAF mutation-positive	23 (2.1)	12 (2.4)	10 (1.8)	0 (0.0)
Median (Q1–Q3) QLQ-C30 GHS score	67 (50–83)	67 (50–83)	58 (50–75)	50 (42–58)

† n (%) unless otherwise stated.

ECOG PS: Eastern Cooperative Oncology Group performance status; FTD/TRI: Trifluridine/tipiracil; Min: Minimum; Max: Maximum; QLQ-C30 GHS: Cancer Quality of Life Questionnaire Global Health Status.

Patient-reported outcomes: QLQ-C30

In the total population, there was no clinically relevant change from baseline in the mean QLQ-C30 GHS score at any assessment-point up to cycle 7 of FTD/TPI treatment (Figure 1). In total 450/975 (46.2%) patients had a clinically meaningful definitive deterioration in QLQ-C30 GHS (369 had a definitive deterioration and 81 patients died). Median (95% CI) time to deterioration was 3.9 months (3.6, 4.5).

Figure 1. Evolution of global health status – mean change from baseline in QLQ-C30 GHS score at each time point^†.

Dashed lines represent the threshold for clinically meaningful changes.

^†Only treatment cycles with ≥10% of the pooled patient population were included in the analysis.

C: Cycle; QLQ-C30 GHS: Cancer Quality of Life Questionnaire Global Health Status; SD: Standard deviation.

Clinical-reported outcomes – ECOG PS

The last recorded ECOG PS value within the FTD/TPI treatment period was the same or better than baseline ECOG PS in 621/985 (63.0%) of patients with available baseline ECOG PS (Table 2). The median time to ECOG PS deterioration was 7.1 months (95% CI: 6.2, 8.3); 399 patients had a deterioration of ECOG PS to ≥2 or had died without previous ECOG ≥2 deterioration. In the pooled data, median PFS was 2.8 months (95% CI: 2.7, 2.9). Patients with ECOG PS 0 at baseline had longer median PFS than patients with ECOG PS 1 at baseline (3.0 months [95% CI: 2.9, 3.3] and 2.4 months [95% CI: 2.2, 2.7], respectively; Figure 2).

Table 2. Last ECOG PS value under treatment with trifluridine/tipiracil according to ECOG PS value at baseline.

Baseline ECOG PS	Last ECOG PS value under treatment, n (%)
	0	1	2	3	4	5	All
0	258 (25.5)	150 (14.9)	47 (4.7)	12 (1.2)	–	1 (0.1)	468 (46.3)
1	44 (4.4)	306 (30.3)	103 (10.2)	45 (4.5)	3 (0.3)	–	501 (49.6)
2	–	2 (0.2)	11 (1.1)	3 (0.3)	–	–	16 (1.6)
3	–	–	–	–	–	–	–
4	–	–	–	–	–	–	–
Missing	10 (1.0)	9 (0.9)	4 (0.4)	1 (0.1)	1 (0.1)	–	25 (2.5)
All	312 (30.9)	467 (46.2)	165 (16.3)	61 (6.0)	4 (0.4)	1 (0.1)	1010 ^† (100)

† Patients with at least one post-baseline ECOG PS measurement under treatment.

ECOG PS: Eastern Cooperative Oncology Group performance status.

Figure 2. Kaplan–Meier curve of progression-free survival data according to baseline ECOG PS from the PRECONNECT and TALLISUR studies.

CI: Confidence interval; ECOG PS: Eastern Cooperative Oncology Group performance status; PFS: Progression-free survival.

In the pooled population, the ORR was 2.1% (95% CI: 1.3, 3.1) and DCR was 31.9% (95% CI: 29.2, 34.8).

Prognostic factors for QLQ-C30 GHS or ECOG PS deterioration

The multivariable Cox proportional hazard model including 924 patients showed that the risk of QLQ-C30 GHS definitive deterioration increased in the presence of lung or liver metastasis and in patients <70 years of age (Table 3). A time from first metastasis to first FTD/TPI intake of <18 months, BMI <25 kg/m² and baseline ECOG PS 0-1 were associated with a decreased risk of QLQ-C30 GHS definitive deterioration (Table 3). Because of the low number of patients with ECOG PS 2, the association of ECOG PS 0-1 with a decreased risk of QLQ-C30 GHS definitive deterioration may not be relevant. All other variables described in the Methods section did not influence QLQ-C30 GHS definitive deterioration.

Table 3. Significant ^† prognostic factors associated with a clinically relevant ^‡ deterioration in QLQ-C30 GHS score.

Factor	HR (95% CI)	p-value
ECOG PS at baseline 0–1 (vs ≥2)	0.822 (0.68, 1.00)	0.0491
Time from first metastasis to first FTD/TPI intake <18 months (vs ≥18 months)	0.668 (0.53, 0.84)	0.0006
BMI <25 kg/m^2 (vs ≥25 kg/m^2)	0.835 (0.69, 1.01)	0.0646
Presence of lung metastasis Yes (vs no)	1.189 (0.98, 1.44)	0.0775
Age <70 years (vs ≥70 years)	1.280 (1.02, 1.61)	0.0345
Presence of liver metastasis Yes (vs no)	1.343 (1.08, 1.67)	0.0086

† At a 10% level.

‡ Deterioration of at least 10 points.

CI: Confidence interval; ECOG PS: Eastern Cooperative Oncology Group performance status; FTD/TRI: Trifluridine/tipiracil; QLQ-C30 GHS: Cancer Quality of Life Questionnaire Global Health Status.

The multivariable Cox proportional hazard model analyses including 1038 patients showed that the risk of ECOG PS deterioration to ≥2 increased with the presence of liver metastasis, a time from first metastasis to first FTD/TPI intake of <18 months, and a lower BMI (Table 4). As would be expected, the risk of ECOG PS deterioration was lower when baseline ECOG PS was 0 versus ≥1 (Table 4).

Table 4. Significant ^† prognostic factors associated with time to ECOG PS deterioration (to ≥2).

Factor	HR (95% CI)	p-value
ECOG PS at baseline 0 (vs ≥1)	0.463 (0.37, 0.58)	<0.0001
Time from first metastasis to first FTD/TPI intake <18 months (vs ≥18 months)	1.421 (1.11, 1.82)	0.0052
BMI <18.5 kg/m^2 (vs ≥18.5 kg/m^2)	1.614 (1.09, 2.39)	0.0175
BMI <25 kg/m^2 (vs ≥25 kg/m^2)	1.659 (1.33, 2.07)	<0.0001
Presence of liver metastasis Yes (vs no)	1.308 (1.02, 1.67)	0.0318

† At a 10% level.

CI: Confidence interval; ECOG PS: Eastern Cooperative Oncology Group performance status; FTD/TRI: Trifluridine/tipiracil.

Association between QLQ-C30 & ECOG PS outcomes

Better ECOG PS at baseline was associated with a longer time to deterioration of QLQ-C30 GHS (median [95% CI] time to deterioration 4.2 months [3.7, 5.3] and 3.5 months [2.9, 4.3] with ECOG PS 0 and 1, respectively; HR 0.84 [95% CI: 0.70, 1.02]; Figure 3). Patients with an ECOG PS 0 at baseline were less likely to have deterioration in most QLQ-C30 subdomains during treatment with FTD/TPI than patients with a baseline ECOG PS 1 (Figure 4).

Figure 3. Time to Global Health Status (QLQ-C30) deterioration^† by ECOG PS status at baseline.

^†Deterioration of at least 10 points.

CI: Confidence interval; ECOG PS: Eastern Cooperative Oncology Group performance status; QLQ-C30 GHS: Cancer Quality of Life Questionnaire Global Health Status.

Figure 4. Impact of baseline ECOG PS (0 vs 1) on time to definitive deterioration of at least 10 points QLQ-C30 scores.

ECOG PS: Eastern Cooperative Oncology Group performance status; QLQ-C30: Quality of Life health Questionnaire.

At last post-baseline evaluation, patients whose ECOG PS improved (n = 45) or stabilised (n = 544) from baseline had no deterioration of QLQ-C30 GHS (mean score change: +0.74 and -2.84, respectively). For those patients whose ECOG PS deteriorated (n = 340), QLQ-C30 GHS also deteriorated (mean score change: –8.88). In a Cox regression analysis, the association between time to ECOG PS deterioration and clinically relevant change in QLQ-C30 GHS score over time (change of ≥10 vs <10) was significant (HR [95% CI]: 1.71 [1.4, 2.2]; p < 0.0001). Physical functioning and role functioning, and symptoms of appetite and fatigue were the QLQ-C30 dimensions most impacted in patients with a worsening ECOG PS (Table 5).

Table 5. Changes from baseline at end of study in individual QLQ-C30 dimensions by change at end of study in ECOG PS.

QLQ-C30 dimension	Change in QLQ-C30 scores, mean (SD), by ECOG PS change
	All (n = 929)	Improvement in ECOG PS (n = 45)	Stabilization in ECOG PS (n = 544)	Worsening in ECOG PS (n = 340)
Global health status ^†	-4.86 (20.8)	0.74 (19.1)	-2.84 (20.3)	-8.88 (21.2)
Functional scales
Cognitive	-2.99 (18.5)	-1.11 (17.2)	-1.62 (17.6)	-5.46 (19.9)
Physical	-5.95 (18.5)	-2.07 (14.8)	-3.46 (16.8)	-10.44 (20.6)
Emotional	-1.93 (20.2)	1.85 (23.2)	-0.21 (19.2)	-5.21 (21.0)
Role	-8.24 (28.1)	0.00 (26.4)	-4.27 (26.8)	-15.73 (28.8)
Social	-2.73 (26.3)	7.41 (22.6)	-1.17 (26.0)	-6.60 (26.7)
Symptom scales
Appetite	8.62 (30.3)	8.89 (32.9)	6.36 (29.4)	12.19 (31.1)
Constipation	1.84 (29.4)	2.22 (27.0)	0.92 (29.1)	3.27 (30.1)
Diarrhea	2.49 (28.6)	-4.44 (31.5)	2.70 (28.6)	3.09 (28.1)
Insomnia	2.16 (30.7)	-8.15 (27.7)	0.43 (30.1)	6.31 (31.4)
Dyspnea	5.83 (26.4)	0.00 (21.6)	4.09 (25.9)	9.37 (27.3)
Fatigue	8.08 (23.0)	3.46 (23.4)	6.15 (22.6)	11.77 (23.3)
Financial difficulties	-1.38 (22.9)	-7.41 (23.5)	-1.67 (22.3)	-0.10 (23.6)
Nausea and vomiting	6.25 (21.1)	2.96 (24.2)	4.49 (19.2)	9.52 (23.0)
Pain	3.84 (27.3)	-4.07 (23.1)	0.89 (25.8)	9.64 (29.0)

† N = 922 (improvement in ECOG PS: n = 45; stabilization in ECOG PS: n = 542; worsening in ECOG PS: n = 335) for patients who had Global Health Status baseline and postbaseline.

Numbers in bold represent a clinically relevant change in quality of life; defined as ≥10-point change in QLQ-C30 score from baseline, or death.

ECOG PS: Eastern Cooperative Oncology Group performance status; QLQ-C30: Quality of Life health Questionnaire; SD: Standard deviation.

Discussion

In this pooled analysis of QoL data from over 1000 patients, we showed that patient-reported QoL (QLQ-C30 GHS) was maintained during treatment in most patients with mCRC who received FTD/TPI. More than 50% of patients did not have any deterioration of ECOG PS status during treatment. We also observed a significant association between time to ECOG PS deterioration and change in QLQ-C30 GHS score over time, and baseline ECOG PS and QLQ-C30 GHS score; a better ECOG PS score (0 vs 1) at baseline was associated with a longer time to deterioration of QLQ-C30 GHS score and a decreased likelihood of deterioration in most of the subdomains of the QLQ-C30 questionnaire.

The potential impact of a treatment on patient QoL is now recognized as an important component in treatment selection [19], and for some patients with advanced cancer and receiving later lines of therapy, equal value is placed on QoL and length of life [20]. However, in clinical trials of third- or fourth-line therapies for CRC, QoL outcomes have not been the focus, and this pooled analysis provides important information on the impact of FTD/TPI. Furthermore, in routine clinical practice, questionnaires such as the QLQ-C30 may be too time-consuming and hence patient QoL is often not captured in an objective way. In contrast, ECOG PS is assessed routinely. Therefore, the association between ECOG PS and QLQ-C30 GHS at baseline, and the subsequent association between ECOG PS deterioration and change in QLQ-C30 GHS score (and sub-domain scores) during treatment with FTD/TPI is important. These findings demonstrate that patient-reported outcome measures can be used to reinforce clinician-reported outcome measures, highlighting the importance of a shared dialogue between patients and physicians in treatment decision making. If patient-reported changes in QLQ-C30 GHS are used in clinical practice, this may provide an earlier surrogate of investigator-assessed ECOG-PS due to a potentially earlier read-out.

The maintenance of QoL in patients receiving FTD/TPI for mCRC shown in this study reflects results from other previously published studies assessing the QoL of patients receiving FTD/TPI for mCRC. The QUALITAS sub-study of 150 patients with mCRC showed that there was no clinically relevant change in QoL (QLQ-C30 summary score) from baseline to 10 months post-baseline [21]. In addition, in a prospective, non-interventional study of 105 patients with mCRC who received either best supportive care without FTD/TPI or at least one cycle of FTD/TPI treatment, patients receiving FTD/TPI reported better overall QoL compared with those receiving best supportive care [11]. Median PFS from our analysis of pooled efficacy data from PRECONNECT and TALLISUR was 2.8 months which was also in line with results from previously published FTD/TPI trials and observational studies, including the RECOURSE trial where median PFS was 2.0 months [5,10,22], while the impact of baseline ECOG PS on PFS was expected. Median time to ECOG PS deterioration was 7.1 months, which was in line with results seen in the RECOURSE trial [5], and maintenance of ECOG PS during FTD/TPI treatment was also observed in a retrospective analysis in Italy [23].

Despite the association between patient-reported QoL and clinician-reported performance status of patients with mCRC receiving FTD/TPI, the prognostic factors for increased risk of GHS deterioration and increased risk of ECOG PS deterioration differed. Indeed, the finding that both a BMI <25 kg/m² and time of <18 months from first metastasis to first FTD/TPI intake increased the risk of ECOG PS deterioration to ≥2 but decreased the risk of QLQ-C30 GHS deterioration is difficult to explain. However, as ECOG PS is clinician-reported and QLQ-C30 GHS is patient-reported, the results of each measure should be interpreted differently and it is feasible that different factors influence their measurement. The association between QoL and lower BMI could indicate that patients with lower BMI are receiving a higher dose intensity; however, this was not measured in the current analysis and there was no correlation between BMI and neutropenia (which in turn is correlated with dose intensity) in the PRECONNECT study.

Risk of QLQ-C30 GHS definitive deterioration was increased in patients <70 years of age; a possible explanation for this is that younger patients may be more impacted by such a serious disease in their daily life and therefore report a reduced QoL than older patients who may have fewer competing demands on their daily lives. Indeed, studies in CRC and other types of cancer have reported lower scores in QoL measures and a greater impact on psychosocial and financial factors in younger compared with older patients [24–26]. Many of the prognostic factors associated with a clinically relevant improvement in QoL are also associated with prolonged OS [27]. The presence of liver metastasis was the one factor prognostic for both increased risk of ECOG PS deterioration and QLQ-C30 GHS deterioration and is likely explained by disease progression.

Pooling data from two studies enabled the analysis of QoL data from a population of more than 1000 patients treated with FTD/TPI; however, it was not possible to pool safety data as the two studies used different MedDRA classifications. Another limitation of this analysis was the low number of patients in the subgroup with a baseline ECOG PS of 2 (n = 20) and these derived exclusively from the TALLISUR study due to the exclusion criteria of PRECONNECT (only patients with an ECOG PS of 0 or 1 were included). This should therefore be considered when interpreting data from this subgroup.

Future research in this topic should involve an exploration of the correlation of patient-reported outcome measures and clinician-reported outcomes in prospective studies. While the benefits of FTD/TPI alone on QoL were confirmed in this study, from here, future studies should also look to investigate the use of FTD/TPI for mCRC earlier than third line on QoL. The maintenance of QoL in the third-line setting, makes FTD/TPI a suitable candidate for treatment alongside other regimens, and future studies should investigate the impact of combination therapies, such as FTD/TPI + bevacizumab (as used in the SUNLIGHT trial [7]), on patient QoL.

Conclusion

In summary, results from this pooled analysis showed that QoL was maintained in most patients receiving FTD/TPI for mCRC and that there was an association between QoL (QLQ-C30 GHS) and ECOG PS. The correlation of patient-reported outcome measures and clinician-reported outcomes needs to be further explored in prospective studies.

Summary points

• It is important that late-stage cancer therapies demonstrate positive effects on patient-reported quality of life (QoL) outcomes alongside traditional clinical outcomes.

• Trifluridine/tipiracil (FTD/TPI) prolongs median overall survival and progression-free survival while maintaining QoL when used as third-line therapy in patients with metastatic colorectal cancer (mCRC).

• This study pooled QoL data from 1100 patients from two open-label studies of FTD/TPI, PRECONNECT and TALLISUR, to explore if patient-reported QoL measures can be used as prognosis factors, and if there is an association between patient-reported QoL and changes in clinician-reported outcomes measures (specifically Eastern Cooperative Oncology Group performance status [ECOG PS]).

• As expected, QoL and ECOG PS were maintained in most mCRC patients on FTD/TPI.

• The presence of lung or liver metastasis and an age <70 years increased the risk of QoL deterioration, and <18 months from first metastasis to first FTD/TPI intake, BMI <25 kg/m² and ECOG PS 0–1 were associated with a decreased risk of QoL deterioration.

• ECOG PS deterioration was more likely with the presence of liver metastasis, <18 months from first metastasis to first FTD/TPI intake, and a lower BMI. ECOG PS deterioration was less likely when ECOG PS was 0 at baseline.

• There was an association between QoL (as measured by Cancer Quality of Life Questionnaire Global Health Status [QLQ-C30 GHS]) and ECOG PS (hazard ratio [HR] 1.71 [95% CI: 1.4, 2.2]).

• In routine practice when patient QoL assessment may be too time consuming, a measure of ECOG PS may serve as an indication of patient QoL when receiving FTD/TPI.

Author contributions

L Wyrwicz: conceptualization, methodology, validation, writing – review & editing; J Taieb: validation, writing – review & editing; T Price: validation, writing – review & editing; J-B Bachet: validation, resources, writing – review & editing; M Karthaus: validation, writing – review & editing; L Vidot: conceptualization, methodology, writing – review & editing; B Chevallier: conceptualization, validation, methodology, writing – review & editing; T Reisländer: conceptualization, validation, writing – review & editing; L Weiss: validation, resources, writing – review & editing; V Heinemann: conceptualization, methodology, validation, writing – review & editing.

Writing assistance

M Gilmour, Editorial Director at Empowering Strategic Performance (ESP) Ltd, Crowthorne, UK, provided medical writing support, which was funded by Servier, Suresnes, France in accordance with Good Publication Practice (GPP3) guidelines (http://www.ismpp.org/gpp3).

Ethical conduct of research statement

All relevant Institutional Review Board/Independent Ethics Committee approvals were gained before commencement of PRECONNECT and TALLISUR. Both studies were conducted in accordance with the ethical principles stated in the Declaration of Helsinki. All patients gave written informed consent before participation. No additional ethical approval or consent was required for this pooled analysis.

Acknowledgments

The authors thank all patients, physicians and study teams participating in this study.

Financial & competing interests disclosure

The PRECONNECT and TALLISUR studies were funded by Institut de Recherches Internationales, Servier, France, and this pooled analysis was also supported by Institut de Recherches Internationales, Servier, France. L Wyrwicz has received honoraria from Servier and Roche. J Taieb has received honoraria for speaker or advisory role from Servier, Roche, Lilly, Celgene, Shire, Amgen, Sanofi, Merck, Lilly and Sirtex. T Price has received compensation for advisory board for Amgen (paid self), Merck, Roche and Takeda (paid self). JB Bachet has received personal fees from Amgen, AstraZeneca, Bayer, Merck Serono, Pierre Fabre, Roche, Sanofi, Servier, Shire and non-financial support from Amgen, Merck Serono and Roche. M Karthaus received personal fees from Roche, Servier and non-financial support from Servier. L Vidot and B Chevallier are employees of Servier. T Reisländer is an employee of Servier Deutschland GmbH. L Weiss received honoraria from Servier, Taiho and Roche. V Heinemann has received honoraria from Merck, Amgen, Roche, Sanofi, Servier, Pfizer, Pierre-Fabre, AstraZeneca, BMS, MSD, Novartis, Boehringer Ingelheim, Celgene, Sirtex, Terumo, Oncosil, Nordic and Seagen, he has received research funding from Merck, Amgen, Roche, Sanofi, Boehringer-Ingelheim, Sirtex and Servier.

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Data sharing statement

Data are available on reasonable request. All data relevant to the study are included in the article or uploaded as supplementary information. Anonymised patient-level, study-level clinical trial data will be shared in agreement with the Servier Data-Sharing Policy available at https://clinicaltrials.servier.com/data-request-portal/.