Background
The burden of liver cancer is high, with approximately 840,000 people worldwide receiving a diagnosis of liver cancer each year and 780,000 dying as a result of the disease [
1]. Hepatocellular carcinoma (HCC) is the most common type of liver cancer, accounting for about 75–85% of cases [
2,
3]. The prognosis for people with HCC is poor, with an estimated 5-year survival of only 12% following diagnosis [
1,
3]. The most common risk factors for HCC are viral hepatitis B, viral hepatitis C, heavy alcohol consumption, and non-alcoholic fatty liver disease, the latter being driven by the increasing prevalence of obesity and type 2 diabetes mellitus [
2‐
4]. The prevalence of underlying risk factors affects disease incidence, which is highest in East Asia, South-East Asia, and sub-Saharan Africa, primarily driven by geographic distribution of viral hepatitis prevalence and is higher in men than in women [
2‐
4].
Novel approaches for HCC management, using targeted agents and immunotherapies administered as monotherapy [
5,
6] or combination therapies [
7‐
9], have experienced several advancements in recent years. There is a paucity of published data on the lived experience of patients with HCC. Only a few qualitative interview studies have elicited the signs, symptoms, and impacts of the disease directly from patients with HCC [
10,
11]. A US study that included 10 patients with HCC (performance status ≤ 2) who had received systemic therapy identified diarrhea, fatigue, skin toxicities, and loss of appetite as patients’ most important concerns [
11]. Pain, in particular abdominal pain, emerged as another important experience upon prompted questioning [
11]. In a study from Taiwan, which included 33 patients who had received surgery, transarterial embolization/transarterial chemoembolization (TAE/TACE), or drug therapy for their HCC, both physical symptoms (e.g., digestive problems, sleep problems) and psychosocial stress were identified as being a significant part of the patient experience [
10]. Interviews and pre-testing were also conducted with individuals with HCC as part of developing the HCC-specific module of the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30) [
12], QLQ-HCC18 [
13]. Most of the participants in that study had very early stage (Child–Pugh A) disease [
13]. Fatigue, body image, and pain were among the issues considered important. The interview studies are informative, but there is still a limited understanding of the patient experience of HCC at different disease stages.
Evaluating patients’ experiences of their HCC in clinical trials is vital so that patients and clinicians can have clearer expectations regarding the effects of new treatments on symptoms and functional impacts that influence health-related quality of life (HRQoL). An in-depth understanding of these experiences is needed in order to select or to develop tailored measures of symptoms and impacts that can be used in HCC trials. This information is useful for regulators and payers, to guide decision making for product approval and reimbursement [
14‐
18]. It is also an essential part of the development and testing of patient-reported outcome (PRO) instruments, to ensure that they are valid and fit for purpose in the intended target population [
19‐
22]. PRO assessment should be included in clinical studies to provide direct evidence of treatment benefit from patients. PROs have become an important endpoint when evaluating the effectiveness of study treatments in clinical studies, and they aid the understanding of benefit-risk evaluation [
23].
The aim of the current study was to obtain a detailed understanding of the patient experience of HCC-related signs or symptoms (signs/symptoms) and impacts on daily functioning and HRQoL.
Methods
Qualitative concept elicitation interviews were conducted with individuals with HCC to identify the signs/symptoms and impacts of HCC that are most relevant to patients and to develop a model describing the concepts (symptoms and impacts) that people with HCC may experience [
18,
20]. A qualitative content analysis method was used. Salient concepts from the interviews were mapped to PRO instruments frequently used in HCC studies.
Patients and procedures
The qualitative patient interviews were conducted in accordance with recommendations provided by International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Good Research Practices Task Force [
20]. The study was approved by the New England Institutional Review Board.
Participants (aged ≥ 18 years) with confirmed diagnosis of early, intermediate, or advanced HCC (Barcelona clinic liver cancer (BCLC) stage A, B, or C) were recruited in the USA with the assistance of a patient advocacy group (Blue Faery), a clinical research organization (Renovatio Clinical), and a patient recruitment agency (Global Perspectives). Processes for recruitment included discussion boards and Facebook support groups (Blue Faery), email and social media outreach, and physician contact. Interview participants’ performance status, stage, and etiology of liver cirrhosis were not captured. All participants provided their written informed consent online before being invited for interviews and received a gift voucher (US$150) for their participation in the study. The qualitative interviews were held by telephone by three trained interviewers. All interviewers were educated to at least Bachelor’s level in life sciences subjects and were experienced in conducting individual patient concept elicitation interviews across a wide range of therapeutic areas. Participants were not known to the interviewers prior to their participation in the study.
A standardized, semi-structured interview guide was developed based on the information obtained from a targeted literature search, blogs/forums covering individual perspectives of HCC, and discussions with clinicians (Online Resource 1). The targeted literature search was conducted using PubMed in October 2018 to identify published research describing the experiences of individuals with HCC. Five patient blog entries/forums detailing symptoms and impacts of HCC (PatientsLikeMe, Blog for a Cure, American Cancer Society, Hepatitis B Foundation, and HealthBoards cancer forums [
24‐
28]) were identified as supplementary sources of information via a targeted search. Telephone discussions were conducted with oncologists from the USA (
n = 2), Europe (
n = 2), and Asia (
n = 1). The clinicians worked in university hospitals or academic centers and had at least 7 years’ relevant practice experience.
During the qualitative interviews, interviewers asked open-ended questions to give participants the opportunity to talk freely about their HCC, including its signs/symptoms and impacts on their daily lives, and asked prompted questions to explore participants’ experiences in greater depth. To examine whether a sign/symptom might be disease and/or treatment related, interviewers probed participants on whether they or their treating physician attributed it to HCC and/or to its treatment, and whether it was experienced before, during, or after treatment. For signs/symptoms and impacts reported by participants, interviewers asked participants how disturbing the sign/symptom or impact was or is to their life using a 0–10 scale, with 0 being ‘not at all disturbing’ and 10 being ‘very disturbing’. When time permitted, participants were also asked to review questions from identified PRO instruments, but this was deprioritized when the concept elicitation part of the interview was lengthy. Interviews lasted approximately 75 to 90 min each and were audio-recorded with the participants’ permission. The study team used de-identified field notes to ensure that there would be no loss of quantitative findings for the analysis. Verbatim transcription of the audio files and general quality control were performed by a professional third party.
Data analysis
A preliminary codebook for analysis of the qualitative HCC interview transcripts was designed based on the literature searches, blogs/forums review, and discussions with clinicians. The codebook was continuously reviewed and updated as new concepts emerged from the HCC interview transcripts. Two researchers assigned codes to sections of text from the interview transcripts relating to signs/symptoms and impacts of HCC, to identify common themes, using ATLAS.ti software (version 8; ATLAS.ti Scientific Software Development GmbH, Berlin, Germany). Six interview transcripts (from interview numbers 5, 10, 15, 16, 20, and 25) were coded by both researchers, to assess inter-coder agreement and as a basis on which to discuss updates to the codebook and coding rules. Good inter-coder agreement was predefined as Krippendorff’s coefficient alpha binary > 0.7 [
29].
Concept analysis was undertaken both in the interviewed group overall and in disease-stage subgroups (BCLC stages A, B, and C). A concept was deemed ‘salient’ if at least 50% of participants mentioned the concept and it had a mean disturbance rating of five or higher (on the 0–10 scale).
The adequacy of the patient sample size was estimated using the principle of concept saturation [
20]. Qualitative study data indicate that a minimum sample size of 12–15 is usually sufficient for concept elicitation interviews [
30,
31] and having a sample size of more than 25 is generally not considered beneficial [
32]. To assess concept saturation, transcripts were grouped chronologically into five waves of five interviews each, and concepts derived from each wave of interviews were compared with concepts from the prior wave(s). If no new concepts appeared in the final wave of interviews, saturation of concepts was considered achieved.
Conceptual model
Signs/symptoms or impacts related to HCC and/or its treatment were used to construct conceptual models to capture concepts overall and by disease stage. Initial models were created from the reviews of the literature and patient blogs/forums and were refined based on the qualitative interviews. The conceptual models included information on salient concepts overall and by disease stage. Signs/symptoms or impacts that had been deemed by interviewed individuals to be exclusively treatment related were not included in the HCC conceptual model. The reason for excluding exclusively treatment-related concepts was that different patients will experience different treatment regimens, and the objective of this research was to capture the common experience of HCC.
PRO instrument mapping
The PubMed, ClinicalTrials.gov, PROQOLID, and PROLABELS databases were searched in October 2018 to identify PRO instruments used in HCC studies. PubMed was used to identify development and validation studies in HCC for each identified PRO instrument, to assess their content validity and psychometric properties. Salient signs/symptoms identified during the interviews were mapped to selected PRO instruments to determine how well these instruments captured the salient concepts.
Discussion
Engaging people living with HCC is crucial to understanding the disease experience fully and identifying signs/symptoms and impacts that are relevant to the patient experience. This qualitative study included in-depth interviews with individuals with HCC to obtain a rich understanding of the patient experience of HCC and to guide fit-for-purpose patient-centered HCC outcome measurement in clinical development. A conceptual model of HCC was developed to describe the signs/symptoms and impacts of HCC experienced by patients, overall and by disease stage. To our knowledge, this is the first conceptual model for HCC that also provides insights per disease stage, from early stage (BCLC stage A) to advanced stage (BCLC stage C) HCC.
Of the 51 signs/symptoms and impacts in the HCC conceptual model, 20 concepts were considered salient, that is, they were prevalent and highly disturbing. However, although these concepts were considered salient based on the interviews conducted with people with HCC, individual experiences may vary and this should be accounted for when developing PRO measurement strategies for clinical trials, as determined by specific context of use. Although the presence or history of cirrhosis was not captured in the present study, a significant proportion of patients diagnosed with HCC have pre-existing liver disease and cirrhosis in addition to their tumor burden [
35]. Some of the reported signs/symptoms and impacts may, thus, have been caused or amplified by underlying liver disease and their presence and magnitude may, therefore, differ in patients with HCC depending on whether cirrhosis is present or absent.
Research across cancer types has shown that fatigue is one of the most frequent and distressing symptoms of cancer and its treatment [
36,
37]. Consistent with these findings, in the current study, fatigue/lack of energy was identified across the various disease stages as being a particularly prevalent experience that individuals with HCC found highly disturbing. Findings per disease stage illustrated that abdominal pain and skin-related issues, which were rated as highly disturbing, were particularly prevalent in individuals with stage C disease, most likely reflecting the increasing tumor burden and liver damage in this disease stage. The most prevalent and disturbing impacts of HCC across the various disease stages were the emotional impacts of the disease, including frustration, fear, and depression, as well as impacts on social life, family and friends, and important activities of daily living. In support of current study findings, the Taiwanese study of 33 individuals treated for HCC participants described feelings of anger, anxiety and depressive mood, and disrupted and strained social and family relationships [
10]. Although the current study did not include individuals with stage D HCC, findings are consistent with interviews conducted with people with terminal stage HCC who described their disease experience as all-consuming and talked about feeling hopeless, annoyed and isolated [
38]. In accordance with the Taiwanese study, in which participants described a lack of physical vigor [
10], physical impacts were frequent in the current study: findings per disease stage showed that a decrease in overall physical activity was common and disturbing for individuals with stage B HCC, and difficulty performing strenuous activities or exercise and decreased working ability were common and disturbing for those with stage C disease.
Women appeared to provide unsolicited information on emotional impacts more often than men, indicating that asking about these impacts is relevant and should be made a routine part of practice.
PRO measures are needed to capture the signs/symptoms and impacts of a disease or condition directly from patients. The EORTC QLQ-C30 and its HCC-specific module QLQ-HCC18, and the FACT-G/FACT-Hep [
34] were identified as the most commonly used disease-specific PRO instruments in HCC studies. The EORTC QLQ-C30/QLQ-HCC18 were shown in the current study to capture the relevant signs/symptoms that are most meaningful to patients with HCC, and to be fit for purpose. Of the 12 salient signs/symptoms elicited from individuals with HCC, 10 are covered by these EORTC instruments. The FACT-G and its hepatobiliary cancer-specific module FACT-Hep were not further evaluated in the current study. While there is commonality between the EORTC QLQ-C30/QLQ-HCC18 and the FACT-G/FACT-Hep in terms of conceptual coverage, there are some gaps in both forms of questionnaires, such as that the FACT-G/FACT-Hep do not cover muscle/strength loss, vomiting, difficulty concentrating or shortness of breath, the EORTC QLQ-C30/QLQ-HCC 18 do not cover dry mouth, and none cover dizziness/vertigo. Researchers should consider disease- and treatment-related criteria, including those of novel targeted therapies, within the context of use when deciding on PRO questionnaire strategy. Additional PRO items should be considered as supplementary tools to evaluate fully the impact of the disease and treatment on patients’ experience given a specific context of use [
39‐
41]. Supplemental items to capture the patient experience can be obtained from the EORTC quality of life item library, the Patient-Reported Outcome Measurement Information System (PROMIS) cancer item bank, and the PRO-Common Terminology Criteria for Adverse Events (PRO-CTCAE) item library.
This study has several important strengths. It had a specific focus on HCC. Rigorous qualitative methodologies were used in the qualitative interviews. Concept saturation was reached, indicating an adequate sample size for the qualitative interviews. Although qualitative study data indicate that a minimum sample size of 12–15 is usually sufficient for concept elicitation [
30,
31], a limitation of this study is the small number of participants specifically in the early disease setting (BCLC stage A;
n = 3). The study did not include individuals with very early stage (stage 0) or terminal stage (stage D) HCC. While significant efforts were made to obtain representation across the HCC spectrum, recruitment proved challenging despite partnering with a patient advocacy organization. This was particularly the case for individuals with early stage HCC, as HCC is often diagnosed only at advanced disease stage. All participants were from the USA and were, thus, not representative of the geography of the disease, which is more common in Asia and Africa than in other parts of the world. Despite being fairly representative of US demographics, the sample was skewed toward white participants, which is a demographic limitation from the perspective of global demographics. Participants were asked about whether they attributed a concept to HCC and/or to its treatment and, although exclusively treatment-related concepts were excluded from the HCC conceptual model, it is likely that the treatment regimens experienced by different individuals also affected the signs/symptoms and impacts of HCC that were reported.
In conclusion, individuals with HCC reported a range of signs/symptoms and impacts, such as emotional and physical distress that negatively affect their daily functioning and quality of life. Obtaining rich qualitative patient experience data directly from people living with the disease guides patient-centered outcome measurement and can help inform patients about what to expect during treatment. The results provide healthcare professionals with important information to contribute to collaborative treatment decision making and patient education. Including patient-reported outcome measures such as the EORTC QLQ-C30 and QLQ-HCC18 in HCC drug development has the potential to provide meaningful patient perspectives, complementing traditional efficacy and safety outcome measures in clinical trials. In light of recent advancements in cancer therapeutics, such as targeted therapies, researchers and healthcare professionals may wish to ensure that relevant treatment-related symptoms are appropriately evaluated.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.