Quality of life in a real-world study of patients with metastatic colorectal cancer treated with trifluridine/tipiracil

a positive ABSTRACT Background Quality of life ( qol ) is important for oncology patients, especially for those with late-stage disease. The present study was initiated to address the lack of published prospective data about the qol benefits of trifluridine/ tipiracil ( ftd / tpi ) compared with best supportive care ( bsc ) in patients with refractory metastatic colorectal cancer ( m crc ). Methods This prospective, cross-sectional, non-interventional study used multidimensional validated scales to evaluate patient-reported qol in two study cohorts of patients and also to measure differences in m crc -related symptoms and pain in a real-world clinical setting. Results Our findings demonstrate that patients with refractory m crc report better overall qol when treated with ftd / tpi than with bsc alone. In that population, statistically significant differences in mean qol measures favoured ftd / tpi over bsc for physical symptom distress, psychological distress, activity impairment, overall valuation of life, and symptomatology. The overall better qol for patients receiving ftd / tpi implies that treatment was well tolerated and was associated with a lower symptom burden. No significant differences for pain were observed between the groups. This study suggests that ftd / tpi is a well-tolerated option for the treatment of patients with refractory m crc , showcasing the value of capturing real-world qol data in routine clinical practice.


INTRODUCTION
Trifluridine/tipiracil (ftd/tpi) is an oral medication approved by Health Canada for the treatment of adult patients with refractory metastatic colorectal cancer (mcrc) who have previously been treated with, or who are not candidates for, available therapies 1 . Approval in that setting was based on results of the phase iii recourse trial, which demonstrated a statistically significant survival benefit of 7.1 months for ftd/tpi compared with 5.3 months for best supportive care (bsc). In recourse, the hazard ratio for death was 0.68, with a 95% confidence interval of 0.58 to 0.81, p < 0.001 2 . Trifluridine/tipiracil was associated with few serious adverse events, neutropenia being the most frequently observed, occurring in 38% of patients treated with ftd/tpi compared with 0% of patients treated with placebo. Based on the recourse study results, ftd/tpi is considered a potential new treatment option, and it is included in the treatment algorithm for colon cancer from the U.S. National Comprehensive Cancer Network 3 and in the guidelines published by the European Society for Medical Oncology 4 as a preferred third-line choice for cytoreduction and disease control in patients with RAS-mutated and BRAF-mutated mcrc, and as a third-line choice in patients with RAS wild-type mcrc 4 .
Despite Health Canada approval and the inclusion of ftd/tpi in international guidelines, reimbursement for ftd/ tpi is inconsistent in Canada 5  recommendation that ftd/tpi be included on the list of medications for the treatment of adult patients with mcrc for whom previous standard therapies have failed 6 . Since August 2019, ftd/tpi has been reimbursed for patients with mcrc in Quebec under the Régie de l'assurance maladie du Québec 7 ; however, patients with mcrc living in the rest of Canada are not yet able to receive public reimbursement for the drug. In August 2019, the pan-Canadian Oncology Drug Review issued a final negative recommendation, stating that ftd/tpi had only "potentially modest [progression-free survival] and [overall survival] benefit, moderate toxicities and an uncertain quality of life" 8 . Patients with mcrc outside Quebec can currently access ftd/tpi only through private insurance or the Taiho Canada (tcan) patient support program.
Patients with refractory mcrc have an unmet need and high demand for a tolerable therapy. A previous study reported that more than 700 Canadian patients applied for the drug through the tcan patient support program over a 12-month period 9 . As of February 2020, the number of patients who have accessed ftd/tpi through the tcan patient support program has increased by 1497, including 982 patients who received ftd/tpi in 2019, most with mcrc. The only other Health Canada-approved treatment option in this setting is regorafenib, which, compared with placebo, was associated with a difference in median overall survival of 1.4 months 10 . According to a systemic review and network meta-analysis, significantly more treatment toxicities are associated with regorafenib than ftd/tpi 11 , affecting tolerability and patient quality of life (qol). Regorafenib is not reimbursed in Canada.
Quality of life is important to cancer patients and can influence their well-being and survival 12 . One study found that 55% of patients with advanced cancer valued qol and length of life equally; an additional 27% valued qol over survival 13 . Although chemotherapy can alleviate some crc-related symptoms, it can introduce others, including nausea and vomiting, fatigue, and change or loss of taste [14][15][16] . Measures of qol are now recognized as being an important element for inclusion in clinical trials and are most typically captured through patient-reported outcomes [17][18][19] .
To date, few prospective comparative data have been published about the qol benefit of ftd/tpi compared with bsc. The internationally validated qol tools used in the prospective preconnect trial suggested that patients with mcrc treated with ftd/tpi can maintain their qol; however, because the trial lacked a comparator arm, it did not demonstrate whether, relative to bsc, ftd/tpi is associated with improved qol 20 . Other ftd/tpi qol studies were retrospective and did not use validated qol tools 21,22 .
In the present real-world study, we used a crosssectional, non-interventional design to understand the difference in qol for patients with refractory mcrc treated with either ftd/tpi or with bsc only. The study used 3 validated qol instruments specifically selected for their relevance to the symptomology, side effects, and clinical problems associated with mcrc, including the Rotterdam Symptom Checklist [rscl (University Medical Center Groningen, Research Institute SHARE, Groningen, Netherlands)] 23 25 , and a visual analogue scale (vas) for pain 26,27 .

Study Design, Objectives, and Endpoints
This cross-sectional, non-interventional study compared patient-reported qol, symptoms, and pain between two cohorts of patients. The primary objective was to quantify the difference in qol between patients with mcrc who were treated with ftd/tpi or with bsc in a real-world setting. Secondary objectives were to quantify the differences in mcrc-related symptoms and pain between patients with mcrc who were treated with ftd/tpi or with bsc.

Study Oversight and Confidentiality
The study protocol was submitted to independent ethics committees or institutional review boards (or both) for review and written approval. The study was conducted in accordance with the principles of the Helsinki Declaration of 1975, as revised in 2008. In addition, the study was consistent with the International Conference on Harmonization Good Clinical Practice, the Good Epidemiology Practices, and applicable regulatory requirements. Patients undertook no risks in participating in the study.
Consent was obtained from all patients before they completed any questionnaires. All patients who entered the study were assigned a unique identification number before they completed the questionnaires; that number was used to ensure that all patient-reported outcome scales and all study documentation were related to an individual patient. All data collected were encrypted, and only authorized members of the study team had access to the patient identification list.
The study was conducted between 27 November 2018 and 29 January 2020. Patients were enrolled from across Canada, with participating sites in Quebec, Ontario, and Alberta.

Patient Eligibility and Recruitment
Patients were eligible for the study if they had received at least 2 prior lines of treatment for mcrc. Because the study was non-interventional, there was no requirement to assign patients to any specific treatment, and they were neither randomized nor stratified with respect to sex, Eastern Cooperative Oncology Group performance status (ecog ps), or number of prior therapies. Patients who, as an outpatient at their oncology centre, had received at least 1 cycle of ftd/ tpi per label 28 were assigned to one study group. Patients who had not received ftd/tpi because of patient refusal, clinical comorbidities, ecog ps greater than 1, or a rapidly progressing tumour were assigned to the other study group. Patients were excluded if they had previously participated in a clinical trial.

Data Collection and QOL Tools
Patient demographics were captured by the treating oncologists on the subject eligibility form, including information about patient age, sex, number of prior lines of therapy for mcrc, KRAS status, and ecog ps at the start of current treatment, medications concomitant with the current intervention, and current treatment details. The physician also confirmed the inclusion and exclusion criteria for participating patients. Data collection was performed using a combination of validated qol scales, including the rscl, the fcsi, and the vas for pain, described briefly below and included in the supplementary appendixes.

Rotterdam Symptom Checklist
The rscl 23,24 measures qol in cancer patients specifically, covering the domains of physical symptom distress, psychological distress, activity level, and overall global life quality.

FACT Colorectal Cancer Symptom Index
The facit measurement system is a collection of questionnaires measuring health-related quality of life for people with chronic illnesses. Those questionnaires often serve as validation benchmarks for newer measures 12,29 . The fcsi 25 is a crc-specific scale designed to capture the clinically relevant problems associated with crc.

Visual Analogue Scale
The vas for pain 26,27 uses a numerical rating scale that has been shown to be reliable and valid for subjective cancer pain measurement.
The patient (or a caregiver) completed the rscl, fcsi, and vas questionnaires in paper format at the time of the visit with the oncologist for a regular mcrc appointment. The set of questionnaires took approximately 15-20 minutes to complete. Validated French versions of the questionnaires were provided for francophone patients in Quebec. The questionnaire was anonymous, and no identifying information was collected. This was a one-time data collection, and patients were not contacted for follow-up.
After the questionnaires were completed by the patient or caregiver, the treating physician reviewed and indicated causality for each item making up the rscl and fcsi scores for patients treated with ftd/tpi; the treating physician also commented whether the patient experienced worsening of pain as a result of treatment.

Data Analysis and Statistics
Patient-reported outcomes data were summarized descriptively by group for each scale, and mean scores were compared. The data are presented as frequencies and proportions, and as means with standard deviation and range.
Proportions were compared using the chi-square test. Continuous variables were compared using the Student t-test. Calculations were performed in the IBM SPSS Statistics software application (version 21: IBM, Armonk, NY, U.S.A.). Two-tailed p values less than 0.05 were considered statistically significant.
Data analysis focused on the differences in mean qol scores between the ftd/tpi-treated cohort and the bsc-treated cohort. Endpoints were analyzed as follows: n Differences between the study groups in overall qol, including physical symptom distress, activity level, psychological distress, and overall valuation of life (measured using the rscl) n Difference between the study groups in crc symptoms (measured using the fcsi) n Difference between the study groups with respect to pain (measured using the vas for pain)

Patient Enrolment and Demographics
The analysis included 105 patients: 50 in the ftd/tpi cohort and 55 in the bsc cohort.  psychological distress was reported by those who were treated with ftd/tpi than by those who were treated with bsc (p = 0.0047). All other domains were also better for patients treated with ftd/tpi, but the differences were not statistically significant: physical distress, p = 0.1727; activity level, p = 0.0520; and overall valuation of life, p = 0.1389. Table iii shows the analysis of individual symptoms in the rscl for the ftd/tpi and bsc groups. The rscl contains 39 items, with most being graded on a 4-point Likert-type scale whose responses range from "not at all" to "very much." Significantly fewer patients in the ftd/tpi cohort than in the bsc cohort reported experiencing 14 of the symptoms itemized in the rscl, including lack of appetite, irritability, tiredness, worrying, depression, nervousness, despair, difficulty sleeping, headaches, dizziness, tension, anxiety, difficulty concentrating, and shortness of breath. Although fewer patients in the ftd/tpi cohort experienced lack of energy, low back pain, decreased sex interest, abdominal aches, constipation, acid indigestion, shivering, tingling hands or feet, sore mouth or pain with swallowing, hair loss, burning or sore eyes, and dry mouth, the differences were not statistically significant. A higher proportion of patients in the ftd/tpi cohort than in the bsc cohort experienced sore muscles, nausea, vomiting, and diarrhea, but those differences were not statistically significant.
In the analysis of overall qol in the rscl, 70% of patients treated with ftd/tpi rated their overall valuation of life as excellent, good, or moderately good; 27% of the patients receiving bsc gave such ratings (p < 0.0001, Table iv). No patient in either cohort rated their overall valuation of life as extremely poor.

Symptomology and Pain
Table v presents the results of the fcsi colorectal cancerspecific scale and the vas for pain. The possible fcsi scores ranged from 0 to 36, where a higher score indicates less symptomology. Patients treated with ftd/tpi reported a higher mean score of 22.9 ± 6.0 (range: 11-34) compared with the 20.3 ± 5.5 (range: 8-31) reported by patients treated with bsc (p = 0.0197). The reported vas for pain was low in both groups, with no significant difference in the level of pain for patients receiving ftd/tpi (2.4 ± 2.6; range: 0-8.0) and for those receiving bsc (3.0 ± 2.4; range: 0-8.5; p = 0.1421).

DISCUSSION
This prospective, real-world, non-interventional study evaluated patient-reported qol and measured differences in mcrc-related symptoms and pain in 105 patients. Results demonstrated that, compared with patients with refractory mcrc who received bsc, those who received ftd/ tpi reported better overall qol. Results were consistent across the 3 qol instruments used. Patients receiving ftd/ tpi reported better outcomes across all domains of the rscl: less physical symptom distress, less psychological distress, lower activity level impairment, and better overall valuation of life. Those results are consistent with the fcsi, which indicated that, compared with patients receiving bsc, those receiving ftd/tpi reported fewer crc symptoms. The parameters measured by those indices are important to patients with mcrc.
Patients in this real-world experience (rwe) study were neither stratified nor randomized-representing one reason why the cohorts were imbalanced with respect to sex (male vs. female), ecog ps, and number of prior therapies. In this non-interventional study, most patients who received ftd/tpi had an ecog ps of 0-1, reflecting the Canadian indication for that agent 28 ; patients who declined treatment or who had an ecog ps of 2 received bsc. That difference might also confound the finding of a lower activity level in the bsc cohort; ecog ps 2 is defined as "ambulatory and capable of all selfcare but unable to carry out any work activities; up and about more than 50% of waking hours" 30 . To account for the large difference in ecog ps between the patients in the two groups, patients who had an ecog ps of 0-1 were analyzed separately, and self-reported qol was directionally better, but not statistically different, in patients treated with ftd/tpi compared with those treated with bsc. Unfortunately, the numbers of patients in the groups were small, which made the analysis inaccurate.
Active treatments can have toxicities that can negatively affect qol and make the active treatment poorly tolerated. Fatigue is commonly reported with the use of ftd/tpi, but it is a symptom that can also be attributed to the disease itself 31 . The rscl results indicated that lower proportions of patients experienced tiredness and lack of energy when treated with ftd/tpi than with bsc, suggesting that any ftd/ tpi-induced fatigue might have been offset by a reduction in disease-related fatigue. The only chemotherapy-related symptoms that were proportionally higher in the ftd/tpi cohort compared with the bsc cohort were sore muscles, nausea, vomiting, and diarrhea-differences that were not statistically significant. Those adverse events were as expected with other chemotherapeutic agents (for example, fluoropyrimidines), and in recourse, they were reported more frequently in the ftd/tpi group than in the placebo group 2 . However, as in recourse, those adverse events were easily manageable in the real-world setting Better score in patients receiving FTD/TPI, but not significantly so. Pts = patients; FTD/TPI = trifluridine/tipiracil; BSC = best supportive care; ECOG PS = Eastern Cooperative Oncology Group performance status.  (26) 10 (18) Neither good nor bad 9 (18) 12 ( with appropriate supportive treatment-a reassuring observation. Our study captured patient-reported symptoms and qol only, and so levels of neutropenia experienced by patients were not documented because that side effect is benign and easy to manage with granulocyte colonystimulating factor and likely did not affect qol.
Although average pain levels were not significantly different between the groups, it is noteworthy that pain scores were low in both groups. We do not know if the pain scores were influenced by pain medications or pain management strategies because our study did not report concurrent medications. However, it is likely that patients in this late treatment setting have engaged with palliative care and that their pain is being well controlled by palliative care physicians and nurses.
Our results add to the established evidence supporting ftd/tpi as a tolerable treatment for patients with refractory mcrc, which includes the pivotal phase iii recourse trial 2 , the prospective preconnect qol study 20 , and the Canadian retrospective rwe study of ftd/tpi in refractory mcrc 9 . The preconnect and Canadian rwe studies both generated clinically relevant data, but they were single-arm noncomparative studies. Although prospective qol data were not collected as part of recourse 2 , the recourse and Canadian rwe 9 studies both evaluated the maintenance or stabilization of a patient's baseline ecog ps status, which is a clinically important parameter in routine practice. In addition, the results of the studies are consistent with real-world data from ftd/tpi compassionate use programs in Italy 32 , Spain 33 , and Germany 34 (all of which are among the 11 countries of the Patented Medicine Prices Review Board); however, none of the compassionate use program studies specifically evaluated qol.
The present study is the first to use validated qol instruments to compare qol and symptomatology in patients with mcrc treated with ftd/tpi or with bsc. The qol questionnaires in our study were completed either by the patient or a caregiver directly. That approach is aligned with various studies showing that qol and symptom reporting are most accurately captured by the patients themselves, because physicians underreport both the severity and the prevalence of toxicities and symptoms 35 . Capturing qol as part of patient-reported outcome measures is generally accepted as necessary, important, and beneficial 36 .
A benefit of rwe studies is that they examine the generalizability of interventions in practice by capturing the experience of patients who are treated in a clinical environment. Such patients are often excluded from randomized clinical trials because of personal characteristics or comorbidities, and their treatment responses and symptoms are monitored less rigidly 37 . Evidence demonstrates that capturing qol and symptoms in a real-world context improves symptom control, patient satisfaction 38,39 , and health-related quality of life 40 ; is associated with increased survival duration 36,37,[41][42][43] ; and is becoming more common in real-world clinical settings 44,45 . In recognition of those benefits, a group of Canadian experts and stakeholders, including the Colorectal Cancer Resource and Action Network, developed and published a position statement supporting the greater use of patient-reported outcomes in Canadian health care and research contexts 46 .
The present study characterizes qol in the real-world context. It had a comparative design, which showed higher qol in patients with mcrc who were treated with ftd/tpi than in those treated with bsc. However, the study was based on one-time data capture and did not track change from baseline in patient-reported qol. Likewise, it did not quantitatively capture survival outcomes or safety signals, nor did it capture prior therapies in a specific and sequenced way to permit comparison of treatment patterns with those in other jurisdictions.
A benefit of the one-time data capture was that it eliminated recall and response shift bias 47 , and also ensured that the data were complete and that no patients were lost to follow-up, which are common problems with rwe studies 37 . A distinguishing feature of the study was its use of tools specific to mcrc to capture patient qol. Those tools included 3 validated qol instruments selected for their relevance to the symptomology, side effects, and clinical problems associated with mcrc: the rscl 23,24 , the fcsi 25 , and a vas for pain 26,27 . Importantly, our real-world evaluation of qol for patients treated with ftd/tpi is reflective of the heavily pretreated mcrc population, and the results are applicable to other patients with mcrc who are treated with ftd/tpi.

CONCLUSIONS
The results of this real-world study demonstrated favourable outcomes for ftd/tpi compared with bsc for overall qol in patients with refractory mcrc. They showcase the tremendous value of real-world qol data in capturing the experience of patients with mcrc treated in a routine clinical setting. Because such data are complementary to data from randomized controlled trials, the information presented in our analysis should help to inform funding, regulatory, and health policy bodies about the urgent unmet clinical need for the growing number of patients with refractory mcrc.

ACKNOWLEDGMENTS
This study was developed and conducted by Drug Intelligence with the support of tcan. To assist with the preparation of the manuscript, tcan supported a medical writer, Chrystal Palaty phd of Metaphase Health Research Consulting Inc. to help with preparation and submission of the manuscript.

CONFLICT OF INTEREST DISCLOSURES
We have read and understood Current Oncology's policy on disclosing conflicts of interest, and we declare the following interests: WYC serves as an advisor to tcan and has received research grants and honoraria from the company; PK has received educational and research grants from tcan and participated on companyorganized advisory boards; AD is Director, Medical, at tcan.
This real-world treatment study was supported by tcan, who also contributed to study design, data collection, and data interpretation. The authors had full access to all data in the study, and all authors had responsibility for the final decision to submit the work for publication.