Background
Despite recent advances in chemotherapy treatments, metastatic pancreatic adenocarcinoma (mPAC) remains incurable and survival rates are still low [
1]. In the absence of treatments achieving long-term survival, chemotherapy aims to slow tumor progression and relieve symptoms [
2,
3]. However, mPAC is an aggressive disease, and the burden of physical symptoms, together with the adverse effects of chemotherapy, lead to patients’ rapid physical decline and the subsequent deterioration of their quality of life (QoL) [
2,
3]. In this scenario, health-related quality of life (HRQoL) becomes a significant concern in the management of patients with mPAC [
4,
5].
Given the acknowledged importance of HRQoL, several clinical trials have reported HRQoL and other QoL-related outcomes in mPAC patients [
5,
6]. Typically, trials including mPAC report a remarkable burden of toxicities associated with chemotherapy regimens. Nevertheless, some of these trials provide supportive evidence that favor chemotherapy in terms of HRQoL, particularly in patients who responded to it [
3,
6]. In line with these observations, global HRQoL and HRQoL-related outcomes, such as pain and fatigue, were associated with survival in the clinical trial setting [
7‐
10]. However, the methodologies used to evaluate HRQoL—usually introduced as secondary endpoints—were heterogeneous, thus they precluded comparisons between studies, often leading to inconsistent results [
5,
6]. Furthermore, thorough selection criteria from randomized controlled trials (e.g., patients with relatively good physical status) are unlikely to reflect a real-life practice scenario. To date, real-world data is limited to a few studies proving the association between HRQoL and survival [
11‐
15], and the positive impact of nab-paclitaxel plus gemcitabine treatment in HRQoL of patients with partial response or stable disease [
16].
The American Society of Clinical Oncology (ASCO) defines clinically meaningful outcomes in the clinical trial setting as a balance between toxicity and efficacy, stressing the importance of HRQoL measures as indicators of toxicity [
17]. Similarly, real-life studies recommend routine assessments of HRQoL to guide treatment decisions [
18]. In this observational prospective study, we describe the evolution of HRQoL in patients with mPAC treated with first-line chemotherapy in routine clinical practice, and analyze patients’ clinical characteristics that may influence their HRQoL.
Methods
Study design and population
This was an observational, prospective, multicenter study including patients with histologically confirmed mPAC (either recurrent or de novo) from 12 Spanish centers, who had started first-line chemotherapy. Adult patients (i.e., aged ≥18 years) with a Karnofsky Index (KI) ≥70 and a life expectancy ≥6 months, who attended routine follow-up visits between October 2014 and October 2015, were consecutively included in the study. No other recruitment sources were considered. Patients that were pregnant or breastfeeding, participating in a clinical trial, or unable to understand or answer questions related to their health were excluded from the study. Baseline demographic and clinical data were collected from the medical records at the initiation visit (i.e., at study entry, before starting the first-line treatment). Assessments, which were scheduled to match routine visits, were performed at baseline, at Days 15 and 30, and every 4 weeks up to 6 months after starting chemotherapy (Treatment Period). After this time, patients were followed up every 6 months for up to 18 months to assess survival (Follow-Up Period).
Endpoints, variables and assessments
The primary endpoint was the evolution of HRQoL in patients with mPAC treated with first-line chemotherapy in a routine clinical practice setting. HRQoL was assessed using two self-administered questionnaires: the European Organization for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire (QLQ-C30) (Version 3.0), and the European Quality of Life-5 Dimensions (EQ-5D). The QLQ-C30 is a 30-item, cancer-specific multidimensional questionnaire designed for prospective clinical studies that includes five functional scales (role, physical, emotional, cognitive and social functioning), three symptom scales (fatigue, pain, and nausea and vomiting), a global health and quality-of-life scale, additional specific symptoms commonly reported by cancer patients (dyspnea, loss of appetite, sleep disturbance, constipation, and diarrhea), and the perceived financial impact of the disease and treatment [
19]. The EQ-5D is a 5-item generic measure of health status, including five dimensions (mobility, self-care, usual activities, pain/discomfort and anxiety/depression) and a Visual Analog Scale (VAS) [
20]. Both questionnaires, which have been translated to Spanish and validated in the Spanish population [
21,
22], were administered during routine follow-up visits at baseline and at each visit during the Treatment Period.
Baseline demographic characteristics considered in the study included age, sex, and weight loss in the last 3 months. Clinical and treatment characteristics included relevant concomitant diseases, date of diagnosis of the primary tumor, location of the metastatic disease, previous adjuvant treatments for non-metastatic disease (if any)—regimen, number of cycles, start and completion dates, and response—, and first-line treatment prescribed (schedule and initiation date). Together with the administration of HRQoL questionnaires, routine assessments were performed at baseline and during the 6-month Treatment Period, including blood tests, biochemistry, tumor evaluation, performance status (PS) (ECOG and KI), treatment modification, and concomitant medication. Other variables were reasons for discontinuation of chemotherapy, recorded during the Treatment and Follow-Up Periods, dose modifications and adverse events, recorded during the Treatment Period. Adverse events (AE) were classified according to the Common Terminology Criteria for Adverse Events (CTC-AE, version 4.02) of the National Cancer Institute (NCI) (
http://ctep.cancer.gov/reporting/ctc.html).
Statistical analysis
Categorical variables and outcomes were presented as frequencies and percentages, whereas continuous variables and outcomes were presented as mean and standard deviation (SD), or as median and interquartile range (Q1,Q3). Categorical outcomes were compared using the Chi-Square test or the Fisher’s exact test. Correspondingly, continuous outcomes were compared using the T-Test, the ANOVA, or their non-parametric counterparts Wilcoxon or Kruskal-Wallis tests. OS and PFS curves were plotted using the Kaplan-Meier estimator, and compared using the log-rank test. Survival analyses are described by median and 95% confidence interval (CI). The improvement or deterioration in HRQoL was defined as an increase or decrease of ≥5 points in the EORTC QLQ-C30 global health status score with respect to baseline [
23]. Overall changes in the score throughout the study visits were assessed using the adjusted linear mixed model. The significance level for all analyses was set at a two-sided α = 0.05.
To estimate the minimum sample size needed, we assumed a rate of QoL improvement of 5%. The rate of QoL improvement was defined as a minimum increase of 5–10 points in EORTC QLQ-C30 global health status scores with respect to baseline. Based on this assumption, a sample size of 110 patients was deemed necessary to estimate the proportion of patients with an improvement in QoL 6 months after starting first-line chemotherapy with a ± 4.1% precision and a 95% CI. Presuming that 10% of patients may not complete the EORTC QLQ-C30, the approximate number of patients included in the study was 120. All analyses were performed using the statistical SAS software for Windows (version 9.4).
Discussion
In this observational prospective study, we found that patients receiving any first-line chemotherapy treatment in routine clinical practice experienced positive changes in various health domains, resulting in an improvement of the HRQoL; this improvement was more notorious in patients with poorer HRQoL at treatment start ―therefore, with most room for improvement. Mean scores of the QLQ-C30 global scale and pain, appetite, sleep disturbance, nausea, and constipation symptom scales improved throughout the chemotherapy treatment, whereas scores from the EQ-5D questionnaire (dimensions and VAS) remained unchanged. The evolution of HRQoL throughout the first-line treatment, measured using the QLQ-C30 global health status score, was significantly influenced by patients’ PS and treatment response. Baseline QLQ-C30 global health status scores influenced OS and PFS, suggesting a prognostic value for this factor.
In this study, two instruments were used to measure HRQoL: the EORTC QLQ-C30 scale and the EQ-5D. While QLQ-C30 showed good responsiveness and was sensitive to changes in HRQoL throughout the Treatment Period, EQ-5D (questionnaire and VAS) did not reflect any changes in HRQoL. This finding is in line with a previous observational study in patients with mPAC, in which QLQ-C30, but not EQ-5D, enabled to identify changes in HRQoL [
16]. In contrast, another clinical trial reported significant changes in the scores of pain symptoms using EQ-5D in patients with advanced pancreatic cancer [
24]. To our knowledge, very few studies on mPAC patients used EQ-5D to assess HRQoL, likely because other cancer-specific tools, such as QLQ-C30 and cancer site-specific instruments, are available. In fact, the National Cancer Institute of Canada Clinical Trial Group chose QLQ-C30 as a standard questionnaire in clinical trials [
25]. Even though EQ-5D lacked sensitivity, its response rates were slightly higher than those of the QLQ-C30 questionnaire: mean response rate throughout study visits was 79.0 and 76.1% for EQ-5D dimensions and VAS, respectively, and 70.3% for the complete QLQ-C30 questionnaire. The EORTC QLQ-C30 global health status score, used to investigate correlations with baseline PS, treatment response and prognosis, reached a mean response rate of 78.6% throughout the study. While a consensus on acceptable response rates for patient-reported outcomes remains unestablished, a 70–80% response rate can be considered reasonable to good [
26]. Although assessing the validity of these questionnaires was out of the scope of our study, the sensitivity observed with the QLQ-C30 questionnaire suggests it is suitable for measuring changes in HRQoL of mPAC patients undergoing first-line chemotherapy.
Despite the importance of assessing HRQoL to evaluate the balance between toxicity and effectivity of chemotherapy, studies focusing on changes in HRQoL of mPAC patients treated with chemotherapy are scarce, heterogeneous, and have often reported conflicting results. Kristensen et al. systematically reviewed 23 clinical trials in advanced pancreatic cancer, which included the assessment of HRQoL as a secondary endpoint. Of the 14 studies reporting changes in HRQoL compared to baseline, five observed an improvement in at least one treatment arm, three observed worsening in at least one treatment arm, and the remaining seven reported no change [
6]. Our results, showing a statistically significant 13-point increase in mean QLQ-C30 global health status scores from baseline to Month 6, are in line with the five studies reporting an improvement in HRQoL in the trial setting [
7,
9,
27‐
29]. Likewise, the decrease in pain scores during the Treatment Period in our cohort, which indicates an improvement of this symptom, are in line with the seven studies (out of eight clinical trials reporting on the evolution of pain scores) showing an improvement in pain [
9,
28,
30‐
34]. Even though the mean QLQ-C30 global health status score improved in our study population throughout the Treatment Period, 17% of patients showed only deterioration. Interestingly, patients with worse HRQoL at baseline were more likely to show an improvement throughout the Treatment Period than those with better HRQoL at baseline.
Asides from assessing the evolution of HRQoL in real-life patients receiving first-line chemotherapy, we investigated the prognostic value of baseline HRQoL in our study population. Previous clinical trials and real-life studies have demonstrated that baseline HRQoL and subsequent changes during treatment (global and subscales) are associated with survival of patients with pancreatic cancer [
8,
9,
11‐
14,
35]. Our results confirmed this trend in a real-life setting, with patients scoring ≥50 in the QLQ-C30 global health status scale and showing higher OS and PFS.
Remarkably, besides low QLQ-C30 global health status scores at treatment start, low baseline PS was significantly associated with a greater improvement in HRQoL throughout the treatment, presumably because of the greater room for improvement in these patients. Thus, although QLQ-C30 global health status scores at baseline were lower in patients with poorer PS (18- and 21-point differences compared to patients with better PS for KI and ECOG, respectively), PS scores of both groups of patients were consistent after 2 months of chemotherapy. This finding encourages priority assessment of HRQoL in patients with poorer PS. Regarding treatment response, patients with stable disease, or partial or complete response had persistently higher QLQ-C30 global health status scores than those who progressed during the first 3 months of treatment. Of note, HRQoL assessment was restricted to patients receiving first-line treatment, thus gradually reducing the number of patients in the progression subgroup. The observed trend was consistent with data from clinical trials, indicating a relationship between HRQoL and disease progression [
3,
5,
6].
In addition to the general limitations of observational designs, such as the uneven sample size across variables due to missing data, the results of this study must be interpreted in the context of the risk of bias associated with a decrease in the study sample over time. This limitation, also observed in previous studies assessing HRQoL [
6], implies that patients who discontinue treatment because of disease progression or death—and, therefore, are likely to have poorer HRQoL—are not followed up any longer. Consequently, the study population is gradually biased towards a better HRQoL as the follow-up progresses. Nevertheless, our overall purpose was to describe the changes in HRQoL during first-line chemotherapy, making it necessary to interrupt follow-up in patients initiating second-line chemotherapy. Future studies investigating the evolution of HRQoL throughout further treatment lines shall follow up patients during larger periods, irrespective of the treatment outcome. Furthermore, since a specific questionnaire for assessing HRQoL in patients with pancreatic adenocarcinoma has been validated [
36], its use in future studies may be more adequate.
Conclusions
In summary, our study shows that most patients starting first-line chemotherapy improve their HRQoL throughout the treatment, although this trend might not be applicable to patients who interrupt treatment early in the first few months due to progression or toxicity. Unlike clinical trials, which usually exclude patients with low PS, our study revealed that these patients—which are likely to have poorer HRQoL—may benefit more from chemotherapy in terms of HRQoL. In line with the psychometric properties reported in validation studies of the EORTC QLQ-C30, in our experience, the administration of the questionnaire during routine follow-up visits of patients with mPAC was feasible. Taken together, our results suggest that, in addition to PS, the EORTC QLQ-C30 global health status score may help to identify mPAC patients that are more likely to benefit from chemotherapy., and that HRQoL may be a useful factor to stratify patients in clinical trials. Finally, our study shows that the EORTC QLQ-C30 scale is a responsive tool for identifying changes in HRQoL of mPAC patients who have started first-line chemotherapy.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.