Health-related quality of life in cardiac sarcoidosis: a systematic review

Abstract People living with cardiac sarcoidosis (CS) are likely to have worse clinical outcomes and greater impairment on health-related quality of life (HRQoL) than other sarcoidosis manifestations. CS can result in a constellation of intrusive symptoms (such as palpitations, dizziness, syncope/pre-syncope, chest pain, dyspnoea, orthopnoea, or peripheral oedema) and/or life-threatening episodes, requiring consideration of invasive cardiac procedures for diagnosis and for the management of acute events. Additionally, the presence of multisystemic involvement and persistent non-specific sarcoidosis symptoms negatively affect HRQoL. A systematic review was undertaken to explore the impact of CS on HRQoL in adults with CS. Multiple bibliographic databases were searched for studies with HRQoL as primary or secondary outcomes in CS (PROSPERO registration: CRD42019119752). Data extraction and quality assessments were undertaken independently by two authors. From the initial 1609 identified records, only 11 studies included CS patients but none specifically reported HRQoL scores for CS patients. The average representation of CS patients was 14.5% within these cohorts (range 2–22%). The majority (73%) was conducted in single-centre tertiary care settings, and only one study (9%) included longitudinal HRQoL data. CS patients were among those sarcoidosis patients with impaired HRQoL and worse outcomes, requiring higher doses of sarcoidosis-specific therapy which contribute to further deterioration of HRQoL. Sarcoidosis studies do not incorporate stratified HRQoL scores for CS patients. While there is a need for longitudinal and multicentre studies assessing HRQoL outcomes in CS cohorts, the development of CS-specific tools is also needed.


Introduction
Cardiac sarcoidosis (CS) is a rare and potentially life-threatening condition characterized by formation of non-caseating granulomas in the heart, which could be detected as part of multi-organ sarcoidosis involvement or may occur in isolation. 1 The most feared manifestation of CS is sudden cardiac death, 2 which may be due to ventricular arrhythmias or conduction abnormalities causing complete heart block. 3,4 Previous studies have shown that CS is clinically manifested in 5% of patients with sarcoidosis. 5,6 However, imaging and autopsy studies estimate the prevalence of cardiac involvement ∼20-35%. [7][8][9][10][11] Diagnosis of CS can be challenging as patients can be asymptomatic or present with a constellation of intrusive symptoms such as palpitations, dizziness, syncope/pre-syncope, chest pain, dyspnoea, orthopnoea, or peripheral oedema. [11][12][13][14] Management of CS requires careful consideration of cardiac medication and invasive cardiac procedures [e.g. endomyocardial biopsy, arrhythmia ablation, implantation of cardiac devices (loop recorder, pacemaker, and/or defibrillator), and in very rare cases heart transplant] in addition to higher doses of immunosuppressive medication with associated potential side effects. Therefore, some people living with CS face additional burden of cardiac symptoms and uncertainties compared with those without cardiac involvement.
Health-related quality of life (HRQoL) is a multi-dimensional concept that reflects the impact of a health condition on involvement and fulfilment of important life areas. 15 It includes general health, physical symptoms and toxicity, and existential issues and domains related to physical, cognitive, emotional, sexual, social, and role functioning. 16 Incorporating HRQoL assessments in routine practice facilitates the patient-clinician communication and enables healthcare professionals to identify moderate-to-severe problems related to HRQoL. 17 The use of patient-reported outcome measures (PROMs) to assess HRQoL has become increasingly important in the management of patients with cardiac diseases. Generic and disease-specific PROMs are essential tools to measure the quality of cardiovascular care, capturing 'what truly matters to our patients'. 18 The use of PROMS provides several benefits for individual patient care. For example, PROMS can facilitate patient-centred care, promote early identification of health issues, and improve outcomes. 19 Moreover, PROMS can reduce unnecessary clinical appointments for stable patients, optimizing the use of limited healthcare resources. 20 Equally, PROMS have shown benefits for clinical teams, service managers, commissioners, regulators, and policy makers. 21 Currently, NHS Digital is using PROMS to calculate health gains on patients undergoing surgical treatment (hip or knee replacement). 22 The burden of sarcoidosis could affect somatic, psychosocial, and economic aspects of life. 23,24 Considering that the indications for treating sarcoidosis are aimed either to 'avoid danger or to improve HRQoL', 25 this approach is particularly pertinent in CS. Firstly, CS is potentially dangerous and, without treatment, could lead to progressive cardiac dysfunction or sudden cardiac death. 2,26 Therefore, prompt initiation of systemic treatment is recommended in patients with CS, 27 which includes steroid and steroid-sparing medication. However, this therapy may itself worsen the overall HRQoL due to side effects and need for frequent monitoring visits for avoiding and early identification of drug toxicity. Secondly, sarcoidosis patients consider HRQoL as the main treatment outcome. 28 In addition to organ-related symptoms, sarcoidosis patients may suffer from a wide range of persistent nonspecific ('parasarcoidosis') symptoms which could impact on HRQoL, such as fatigue, weight reduction, pain, general weakness, exercise limitation, everyday cognitive impairment, poor subjective sleep quality, and depressive symptoms. [29][30][31] It has been suggested that the impact on HRQoL is greater in sarcoidosis patients with cardiac involvement. 32 However, the sarcoidosis-specific HRQoL PROMs were developed without representation from patients with CS, and there is a lack of information related to the impact of CS in HRQoL which limits clinical decisions and future research focus. Thus, we sought to undertake a systematic review to identify and synthetize existing studies exploring the impact of CS on HRQoL in adults.

Protocol and registration
This review was registered on PROSPERO, an international prospective register of systematic review (February 2019, reference CRD42019119752) and is reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. 33

Eligibility criteria
All studies which included HRQoL as primary or secondary outcomes in adults with a diagnosis of cardiac sarcoidosis were eligible for inclusion. Studies were included if they met the following inclusion and exclusion criteria: population, adults (aged ≥18 years) living with cardiac sarcoidosis, and outcome, HRQoL as primary or secondary outcome but excluding quantitative studies and those using a non-validated HRQoL tool. Studies were not excluded on the basis of publication date or quality assessment but were required to be available in English language.

Information sources
The search was conducted on 24 April 2022 including the following elec-

Study selection
After removal of duplicates, the review team (J.C.Q-C., J.S., M.T., and N.S.) completed two rounds of screening. First, titles and abstracts from the initial search (electronic databases) were independently screened, followed by full-text review of potentially relevant records from the first screening. All screening was undertaken independently by the first reviewer (J.C.Q-C.) and a second reviewer (J.S., M.T., and N.S.). Discordances were resolved by consensus. Due to the high specificity of cardiac sarcoidosis, a second stage of study selection included supplementary searching strategies to optimize identification of potential studies, following the above screening process. A detailed search string for each database is listed in the Supplementary material online, Table S1.

Data extraction (data collection process and data items)
Data from eligible studies were extracted independently by two reviewers (J.C.Q-C. first review and second reviewers one of J.S., M.T., and N.S.) on to a standardized proforma, with discordances resolved by consensus. Extracted data included primary author, publication date, settings and country of study, study recruitment dates, study design, population and sample size, eligibility criteria, PROMs used and assessment timepoints, and sociodemographic and clinical characteristics. HRQoL information included results for domains related to generic or sarcoidosis-specific HRQoL tools at each timepoint, adjusted HRQoL results for cardiac involvement, and any specific subgroups with significant differences for HRQoL.

Quality assessment and risk of bias
The quality assessment was performed in parallel with the data extraction process. The relevant Critical Appraisal Skills Programme (CASP) template was used by two reviewers. 34 The CASP was subsequently aggregated to develop a score for each publication as a percentage of the number of met criteria out of the number of applicable. These quality assessment scores were organized into categorized previously reported. 35,36 Studies rated above 80% were considered high quality, 60-79% were moderate quality, and lower than 59% were appraised as poor quality. No studies were excluded on the basis of the quality assessment to avoid the potential exclusion of insightful studies. Graphs were generated to summarize the CASP checklists and the risk of bias summary. 37 Analysis results were summarized using descriptive statistics, tables, and narrative synthesis, as appropriate. Interpretation of the analysis was discussed and agreed by all reviewers.

Patient and public involvement
The patient and public involvement group (five members with cardiac sarcoidosis diagnosis) helped to inform the scope of the research questions and contributed to the PROSPERO protocol. Figure 1 shows the adapted PRISMA flowchart. 33 In brief, a total of 1609 potentially relevant records were identified. Following title and abstract review, 81 went on to full-text review resulting in 11 studies being included for analysis. [38][39][40][41][42][43][44][45][46][47][48] Study characteristics Table 1 summarizes the study characteristics, and Table 2 describes clinical information and the HRQoL results for the 11 included studies.

Study selection
The majority of studies was conducted in centres from the USA [n = 6 (55%)], with an equal proportion published in the last 3 years (2019-2022) ( Table 1). The majority of samples was conducted in a single centre (n = 8, 73%), with just under half conducted between 2005 and 2009 (n = 5, 45%). None of the studies focused solely on CS patients or reported stratified HRQoL scores for CS patients, but four papers presented some adjusted analysis related to HRQoL and cardiac involvement. 41,45,46,48 Almost half of studies (n = 5, 45%) used combinations of generic and sarcoidosisspecific HRQoL tools, and sarcoidosis-specific HRQoL were used individually in over a third of studies (n = 4, 36%). Only one study included longitudinal data for HRQoL assessments (follow-up every 12 months for 3 years) ( Table 1).
Overall, 5334 sarcoidosis participants with mean age of 50 years were recruited across the 11 studies, including a total of 775 CS patients.

Quality assessment and risk of bias
The percentage scores of CASP checklists and the appraised levels of evidence are presented in Table 1. Over half of the included studies (n = 6, 54%) were categorized as poor quality and the remaining publications (n = 5, 46%) as moderate quality. Figure 2 illustrates the quality assessment for each of the included studies. None of the studies fulfilled the full CASP checklist for cohort studies. Considering study designs, cross-sectional studies and retrospective studies scored lower compared with observational studies. The poorest-rated section was the adequacy of follow-up of cohorts, which met the criteria only in one study. Furthermore, only 18% (n = 2) of the studies met the 'consideration of confounding factors in the design' and 'application of results to local population'. Figure 3 demonstrates the predominant 'unclear risk of bias' due to the lack of information provided across all the manuscripts.

HRQoL outcome by measurement tool
There was a lack of standardization in the HRQoL tools used across studies. The majority of the studies (n = 9, 81%) included sarcoidosisspecific HRQoL PROMs: the sarcoidosis health questionnaire (SHQ) was used in five (45%), the King's sarcoidosis questionnaire (KSQ) in three (27%), and the sarcoidosis assessment tool (SAT) in two (18%). In addition, some studies were complemented by a generic HRQoL PROM such as the 36-item short form (SF-36) (n = 3, 27%) and/or the respiratory-specific HRQoL PROM St. George's respiratory questionnaire (SGRQ) (n = 3, 27%).

HRQoL measured by the sarcoidosis health questionnaire
Among the two studies that solely used the SHQ, 43,46 the Serbian study found that patients with multi-organ involvement had the lowest mean  SHQ scores for daily functioning (3.72 ± 0.91), physical functioning (4.00 ± 0.98), and emotional functioning (3.74 ± 0.95). 46 The CS patients (n = 23, 7%) had lower scores for physical functioning (ANOVA F = 1.523, P = 0.041, SHQ scores were not provided). Moreover, patients receiving monotherapy with methotrexate (MTX, n = 21, 6%) had better SHQ scores (P < 0.05 for all domains) in comparison with those with combined therapy (MTX + prednisolone, n = 46, 13%). 46 Three studies combined the SHQ with SF-36. 38,39,43 Statistically significant better HRQoL was reported in men according to the SHQ physical functioning and SHQ total. 38 The Japanese study 48 showed that 23% (n = 20) of those who completed follow-up investigations (n = 88/122) had a 5-year deterioration, which was significantly associated with the SHQ total (adjusted OR 0.54, 0.29-0.99) and the SHQ domain physical functioning scores (OR 0.48, 0.27-0.87) after adjusted for CS at enrolment.

HRQoL measured by the King's sarcoidosis questionnaire
The KSQ was used solely in a German cohort 41 (n = 200) reporting that patients with multiple organ involvement had significantly lower scores for the KSQ subdomain general health status (GHS) (adjusted R2 = 0.56, P < 0.001) and those receiving sarcoidosis-specific therapy had worse HRQoL (P < 0.002). 41 Cardiac involvement lowered the GHS score by estimate 4.25 points (95% CI −21.20 to 12.70, P = 0.621). 41 Among the studies combining the KSQ with additional HRQoL tools, worse dyspnoea was associated with lower scores for all the KSQ subdomains (P < 0.001), including the strongest correlations for the GHS (rho 0.635) and lung health (LH) (rho 0.724). 47 Similarly, fatigue was strongly correlated with the GHS + LH (rho −0.800). 47 Lower baseline modified extrapulmonary organ assessment tool (MePOST of <3), which incorporated both the number of organs and the severity of organs involved, was associated with statistically better outcomes for 9 of the 15 outcome of interest for sarcoidosis patients (OIPs): unscheduled healthcare visits, workdays missed, steps walked, prednisolone dose, pain, satisfaction, sleep, lung concerns, and global health perception. 45

HRQoL measured by the sarcoidosis assessment tool
The presence of symptoms at the initial visit was statistically and clinically associated with worse HRQoL (P < 0.022 for the SAT modules daily activities, satisfaction, and fatigue), significantly more organ involvement, and more frequent requirement for any sarcoidosisspecific therapy (P < 0.001). 44 Moreover, cardiac involvement at baseline was associated with 3 of the 15 selected OIPs including higher doses of prednisolone (P < 0.0001), more consumed calories (P < 0.0027), and more sleep disturbance based on the SAT subdomain (P < 0.0336) after a 6-month follow-up. 45

HRQoL measured by the patient visual analogue scale
Less than 7 cm (out of a total of 10 cm) on the patient visual analogue scale (patient-VAS) at baseline was statistically associated with worse outcomes in 13 of the 15 selected OIPs (all except steps walked and calories consumed), which was superior to the additional physiologic, radiographic, or HRQoL measures. 45

HRQoL measured by SF-36
Among the three studies using SF-36 in combination with the SHQ, 38      conclusions were that men had better HRQoL (P < 0.02 for SF-36 scores) 38 and the SF-36 was a valid HRQoL tool for sarcoidosis. 39 Moreover, three SF-36 domains (general health perception, vitality/energy, and limitations in usual role activities due to physical health problems) had statistically significant associations with a 5-year clinical deterioration, but only baseline HRQoL scores were collected to evaluate the long-term deterioration. 48

HRQoL by WHOQOL-BREF
The sarcoidosis group had statistically significant better HRQoL compared with the idiopathic pulmonary fibrosis group, including the WHOQoL-BREF domains 'environment' (P < 0.05), 'physical health' (P < 0.01), and 'facet overall QoL' (P < 0.01) (see Table 2). Moreover, the prevalence of type D or 'distressed' personality in both groups did not differ from a general population. 40

HRQoL measured by the 15-dimensional measure of HRQoL
The 15-dimensional (15D) was combined with the SGRQ, but there was not any significant difference for the 15D results when segregated by 'isolated pulmonary' or 'pulmonary plus extrapulmonary' sarcoidosis cases (P = 0.106). 42

Factors influencing HRQoL in cardiac sarcoidosis patients
The 11 identified studies reported various factors influencing HRQoL in sarcoidosis cohorts, with variable CS representation. These associations include potential predictors of HRQoL in CS patients related to socio-demographic, symptomatology, therapy, physiologic, and clinical status ( Table 3 and Supplementary material online, Table S3).

Discussion
In 2011, the World Association of Sarcoidosis and other Granulomas disease (WASOG) 56 recommended that all research studies should incorporate HRQoL measurement; however, evidence remains limited. 24,57,58 This review summarizes 11 studies that included CS patients exploring HRQoL as primary or secondary outcome, of which only 6 included data collected since the WASOG recommendation. Furthermore, the three current clinical diagnostic tools for CS were introduced between 2014 and 2017. 14,52,59 Despite five studies collected considered these studies as unclear risk of selection bias. Four of the studies did not provide sufficient information to determine if the outcome was measured accurately and were deemed as unclear for detection bias. Based on a review of evidence and determined a priori, significant factors were identified to assess confounding bias. Three out of the four studies with follow-up did not provide sufficient information to determine completeness rates and were therefore deemed as unclear risk of attrition bias. Considering that none of the studies fulfilled the full CASP checklist, all the studies were marked as unclear risk for other bias. Additionally, the included studies did not provide sufficient socio-demographic characteristics. For example, three studies did not report the ethnicity of their samples, 40,42,48 and only two studies provided data for the occupation and/or education level. 41,43 Furthermore, the following reasons were noted to be deemed as unclear risk of other bias: • Limited CS representation (4-5%), socio-demographically imbalanced characteristics, scores for the SF-36 subscales not provided. 38,39 • Limited CS representation (8%), no details were provided for the 25% of the sarcoidosis population who did not return study questionnaires. 40 • Limited CS representation (3%), two references were used to describe the cohort of participants, but the provided samples were not matching with the reported data. 41 • Limited CS representation (2%). 42 • Online survey with no clinical characteristics. Self-reported socio-demographic characteristics. Only the SHQ total score was provided. 43 • Retrospective analysis. Over half (363/660, 55%) of the cohort did not complete the SAT due to the validation date. 44 • Monthly self-completed assessments using a new online platform for 6 months, and the study only reported baseline HRQoL scores. 45 • Limited CS representation (7%). Only patients undergoing pulmonary function test during the clinical routine visits were eligible. No confidence intervals included. 46 • Data analysed retrospectively from a multi-national database (n = 456). Participants with advanced sarcoidosis (n = 231) were followed up every 6 months for 3 years, but only baseline HRQoL assessments were reported. The analysis included over one-half (53%) of the cohort with advanced sarcoidosis and completed a 12-month visit. 47 • Participants completed a follow-up 5-year visit, but HRQoL assessments were included only at baseline. 48 • Other bias was deemed as unclear risk for all the studies considering that none of the studies fulfilled the full CASP checklist for cohort studies as detailed in Figure 2.
data between 2014 and 2020, only two 45,47 used one of the latest tools (WASOG, 2014) 52 which could have limited the accuracy to define cardiac involvement in the included sarcoidosis cohorts ( Table 1). 49 Our findings potentially demonstrate impaired HRQoL in patients with CS; however, data need to be interpreted with caution because of the lack of CS-specific studies, the lack of disease-specific standardized HRQoL tools, and the appraised low to moderate evidence levels. For instance, the higher incidence of CS in Japanese cohorts 60 or the lack of longitudinal assessments to measure changes in HRQoL 61 should be considered when interpreting that certain baseline HRQoL scores (SHQ and SF-36) were associated with a 5-year clinical deterioration. 48 Also, sarcoidosis patients with 'extrapulmonary' 42,46 or 'advanced disease' 47   these findings is limited by the small number of CS patients in these cohorts (range 2-17%). 42,46,47 The majority of the studies (n = 10, 91%) reported HRQoL results for a single timepoint (baseline), and only two studies (18%) included longitudinal HRQoL assessments at 6 months 45 and every 12 months 47 up to 3 years. This is unsatisfactory to determine the impact of a complex disease and HRQoL impairment over time. Previous work has recommended to initiate systemic sarcoidosis therapy in CS, 27 which is usually combined with cardiac-specific treatments and likely to impact on HRQoL at any specific time. Similarly, a post hoc exploratory analysis of HRQoL in sarcoidosis outpatients with corticosteroid therapy suggested that CS patients had a greater impact on HRQoL compared with patients with non-cardiac involvement. 32 Therefore, there is a need to incorporate longitudinal 'snapshot' HRQoL assessments and to describe HRQoL scores thoroughly for patients with cardiac and extracardiac involvement.
Overall, we also identified that there was no consistency in the HRQoL tools used, with contradictory conclusions regarding the relevance of generic and/or sarcoidosis-specific HRQoL tools. 57 Furthermore, these sarcoidosis-specific tools do not include any cardiac-specific subscale/item. Certainly, the development of CS-specific items/tools and/or adaptation of existing sarcoidosisspecific PROMs would be a substantial advance for clinical research and to understand the impact of CS. 23 Indeed, the implementation of these HRQoL tools in clinical practice will help to maximize the quality of care considering the indications for systemic therapy and the need to evaluate HRQoL in CS patients. 21,62,63 A recent study focussing on sarcoidosis healthcare costs found that CS patients had higher healthcare costs compared with those without cardiac involvement. 64 In the same way, CS was reported as a predictor for hospitalizations, development of new sarcoidosis-related comorbidities, and severe impairment on family finances. 43 Therefore, there is an opportunity to quantify the CS-associated costs in detail, assessing the CS-related impact on work ability and productivity, and the burden for disability claims which may impact on HRQoL. 15 Many unmet and urgent needs remain for people with CS. This review provides potential factors impacting on HRQoL based on sarcoidosis cohorts with CS representation. For instance, one study supported psychological screening, and if needed, the incorporation of psychological counselling, coping strategies, and emotional support in the management of sarcoidosis. 40 Similarly, recent reviews have suggested potential factors influencing HRQoL 58 and sarcoidosis-related manifestations to be considered during disability evaluations. 15 For example, stress or anxiety may contribute to palpitations. Thus, multidisciplinary CS teams should ensure comprehensive care and holistic management considering individual needs and concerns. 65

Limitations
This review has three main limitations. Firstly, none of the included studies reported specific HRQoL results for CS patients, and only 4 of the 11 presented some adjusted HRQoL results for sarcoidosis patients with cardiac involvement. Therefore, the interpretation of these HRQoL findings and the generalizability is challenging. Secondly, no studies were excluded on the basis of quality assessment or risk of bias. This was to enable full exploration of insightful studies with multiorgan manifestations. However, the data availability bias could have contributed to mask other potential bias, such as selection bias (population) or reporting bias. Thirdly, due to the heterogeneity of the outcome measures and variations in the original designs, this review had limited data to conduct direct comparisons or meta-analysis. Despite these limitations, this methodological robust review is valuable in describing and summarizing the current HRQoL evidence for sarcoidosis cohorts with cardiac involvement representation.

Conclusions
In conclusion, patients with CS are included in sarcoidosis cohorts with impaired HRQoL and worse clinical outcomes. However, sarcoidosis studies with HRQoL as primary or secondary outcome do not incorporate stratified HRQoL scores for patients with cardiac involvement. There is a need for further research, especially for longitudinal and multicentre studies, using complementary HRQoL tools to assess CS cohorts. The development of CS-specific and/or adaptation of existing sarcoidosis tools would be a substantial advance to understand and evaluate the impact of CS on HRQoL, improving the quality of care for a complex disease.

Data availability
The data underlying this article are available in the article and in its online supplementary material. Any additional data will be shared on reasonable request to the corresponding author.