A systematic review of the patient reported outcomes that affect patients with muscle invasive bladder cancer after radical cystectomy and urinary diversion

Abstract Objectives To determine the functional domains and symptom scales that affect patients most following radical cystectomy (RC) and urinary diversion (UD), and if a single instrument (or combination) adequately captures these bothersome symptoms. It is unclear whether current patient reported outcome (PRO) instruments that have been used to assess quality of life in patients following RC and UD adequately cover the most bothersome symptoms affecting patients. Materials and methods A systematic search of MEDLINE, EMBASE, PubMed, Cinahl and Cochrane was conducted from January 2000 to May 2023 for original articles of patients who had RC and UD since 2000 for muscle invasive bladder cancer. The Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) process was followed. Extracted data included the PRO measures used, domains reported and scores in the first 12 months post‐surgery (short‐term) and after 12 months (long‐term). A conservative threshold of <70 for functional domains and >30 for symptom domains was used to determine which PRO domains were potentially concerning to patients in each study. Quality assessment was performed using the QUALSYST appraisal tool. Results Thirty‐five studies met the inclusion criteria, including a total of eight unique PRO instruments. The main findings indicated that physical function was the most concerning PRO for patients with both neobladder (NB) and ileal conduit (IC) in the short and long term. Additionally, bowel, urinary and sexual bother were concerning symptoms for patients with NB in the long‐term, but only in the short‐term for those with IC. Conclusions The main issues are adequately addressed using the combination of EORTC QLQ‐C30 and QLQ‐BLM30 instruments.


| INTRODUCTION
Radical cystectomy (RC) with urinary diversions (UDs) such as ileal conduit (IC) or neobladder (NB) formation remains the standard for curative treatment of localised muscle invasive bladder cancer (MIBC).
Surgery can be associated with high procedure-related complications, re-admission rates and significant psychological distress to the patient. 1 To optimise clinical outcomes effects on health-related quality of life (HRQoL) should be considered.Therefore, evaluation of patient reported outcomes (PROs) is an imperative component in quality care. 2 Several instruments have been designed to better understand and evaluate PRO in patients' post-surgery.Common instruments used include the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30) 3 accompanied with the Bladder Cancer Muscle Invasive module (EORTC QLQ-BLM30) 4 ; Functional Assessment of Cancer Therapy Bladder Cystectomy (FACT-Bl-Cys), formerly known as the FACT-Vanderbilt Cystectomy Index 5 ; Ileal Orthotopic Neobladder-Patient Reported Outcome (IONB-PRO) 6,7 ; and the Short Form Surveys (e.g.SF-36). 8e aim of this systematic review was to evaluate the PRO data following RC and UD for MIBC studies since 2000 to ascertain the aspects of functioning and symptoms that affected patients the most, in the short and long term.Furthermore, our aim was to determine if a single instrument (or combination) adequately captures the symptoms that matter most to patients, and therefore can be used alone to assist patients and clinicians in better understanding the effects of RC and UD on HRQoL.

| Eligibility criteria
We included full-text primary research studies published since 2000 that reported PRO data for patients who underwent RC and UD (IC and NB) for localised MIBC.Studies that did not report PRO data by domain or in ways that could be extracted from graphs, tables or text, did not provide separate data for different UD types, including patients with metastatic disease and non-English language were excluded.This was to enable the description of important outcomes by type of UD.The study was registered with the International Prospective Register of Systematic Reviews (PROSPERO) (ID CRD42021272159).

| Information sources
A systematic search of Medline, Embase, PubMed, Web of Science, Cinahl and Cochrane (EBM Reviews) databases was conducted from 1 January 2000 to 5 May 2023.We limited studies to after 2000.

| Search strategy
Our search strategy was developed in consultation with an academic librarian and pilot-tested.The final search terms are listed in Appendix A. Reference lists of relevant studies and journals were also searched.

| Selection process
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) process was followed (Figure 1).Two reviewers (J.L. and S.N.) independently screened titles and abstracts, and when relevant, assessed full text for eligibility.In the case of a disagreement, mediation by a third author (R.M.B. or M.I.P.) was required (Figure 1).

| Data collection process
Study and patient characteristics, type of surgery (open or minimally invasive) and UD, PRO measure, PRO domains assessed, PRO data, adverse events, and functional and symptom scores were all extracted from studies and tabulated (Appendix B).

| Study risk of bias assessment
Quality assessment of studies was performed independently by two authors (J.L. and S.N.) using the QUALSYST appraisal tool. 9Again, in the case of disagreement, mediation by a third author (R.M.B) was required.We did not exclude any studies based on overall quality or risk of bias as our aim was to review the PRO instruments used and domains assessed rather than to extract any measure of clinical benefit.Rather, we used the scores to assist our overall interpretation of our results.

| Synthesis methods
Functional and symptom scales and their respective scores for each instrument were collected from each study and tabulated according to UD type and period of follow-up (3-12 months; over 12 months, corresponding to 'short-term' and 'long-term' outcomes, respectively) (Table 2).For the 'long-term' assessment, we used the 12-month assessment data, or the next and closest available timepoint reported (Table 2).We then assessed the clinical importance of data extracted for each domain at the two assessment points.For studies that used the EORTC QLQ-C30 instrument, thresholds for clinical importance established by Giesinger et al. were used. 10The remaining instruments were all standardised on a 0-100 scale, with higher functional scores indicating better functioning and higher symptom scores indicating worse symptoms.None of these remaining scales had formal clinical interpretation guidelines, so we considered functional domain scores below 70 to indicate impaired function, and symptom scale scores above 30 as bothersome symptoms (except in the case of sexual function for EORTC QLQ-BLM30, where a score below 30 was considered bothersome).Additionally, where individual studies had data contributing to both our timepoints of interest (short-and long-term HRQoL), we compared the two assessments at the domain level to describe the clinical significance of any improvements or regression.We considered 10-point differences in functional domains and symptom scales to be clinically important.

| Study selection
We identified 4769 records from the literature search for initial assessment.After a full-text review of 164 articles, 35 articles were included in the systematic review: 25 included patient-reported outcomes data for IC and 32 for NB (Figure 1).

| Risk of bias in studies
Most studies (n = 29, 83%) scored above 55% on the QUALSYST appraisal tool, which is considered a liberal threshold for quality. 9The range in scores was 31%-91% (Appendix C). and 24 unique PRO domains (Table 2).Physical, social and emotional functioning were the most frequently reported domains, reported in 11, 10 and 10 of the 13 studies, respectively (Table 2).
According to our criteria for patient-perceived domains of concern or clinical relevance, 10 physical functioning was a concern in all 11 studies that reported this domain (Table 2).Similarly, functional wellbeing was a concerning domain in all studies for which it was reported (n = 5/5), as was bowel functioning (n = 4/4), bowel bother (n = 3/3), urinary bother (n = 3/3) and sexual bother (n = 3/3) (Table 2).
Long term (greater than 12 months) following radical cystectomy with ileal conduit There were 21 studies with PRO data reported 12 or more months following IC, which used seven different PRO measures and 28 unique PRO domains (Table 2).Physical functioning/mobility was the most frequently reported domain, bowel function and bother, urinary bother and sexual bother were least commonly reported (Table 2).
According to the criteria for patient-perceived domains of concern or clinical relevance, 10 consistent with previous, physical functioning remained a domain of most concern with 77% of 22 studies with PRO data reporting concerns (Table 2).General health, mental functioning, functional wellbeing and vitality domains reached concerning levels in all four studies (100%) in which the domains were reported (Table 2).

Short term (3-12 months) following radical cystectomy with neobladder
There were 18 studies with PRO data reported in the short term following NB, comprising nine different PRO measures and 30 unique PRO domains (Table 2).Across the 18 studies, there were 16 reported values for physical functioning or mobility, and levels of this domain were deemed concerning in 75% of the 16 instances (Table 2).For the symptom scales, pain met our level for a concerning symptom in n = 2/8 (25%) of instances, diarrhoea n = 2/5 (40%), financial impact n = 2/5 (40%) and dyspnoea n = 2/5 (40%).Bowel bother, urinary bother and sexual bother each reached concerning levels in n = 3/3 (100%) of studies reporting in the short term (Table 2).

Long term (greater than 12 months) following radical cystectomy with neobladder
There were 29 studies that reported PRO data at 12 or more months following NB, which used nine different PRO measures and 29 unique PRO domains (Table 2).Physical functioning was the most reported domain (Table 2).

Coverage of patient reported outcomes of importance across patient reported outcome instruments not identified with EORTC QLQ-C30
and EORTC QLQ-BLM30 for patients with ileal conduit and neobladder After reviewing all the symptom and functional domains that reached clinical importance, and after accounting for semantic differences in labelling domains between instruments, the combination of the EORTC QLQ-C30 and EORTC QLQ-BLM30 addressed all the issues

| Change over time
Change in patient reported outcomes before and after 12-month follow up for ileal conduit Eight studies reported data for both time points for IC (Appendix B).A clinically relevant improvement in emotional, 11 role 12 and cognitive 13 was reported for one study each.Among the studies that had data at both short-and long-term follow-up, symptom scores were only clinically worse (decrement of greater than 10 points) in one study each for fatigue n = 1/4 (25%) studies, constipation n = 1/4 (25%) and body image n = 1/4 (25%) (Table 2).
Change in patient reported outcomes before and after 12-month follow up for neobladder Fourteen studies reported data for both time points for NB (Table 2).
Across all function domains, a clinically relevant improvement in social, 14 emotional, 14 role 12 and cognitive functioning 13 was reported for one study each, respectively.Only one study by D'Agostino et al.
showed a clinically relevant decrement of greater than 10 points for emotional functioning when the pre-and post-surgery-12-month data was compared (Table 2). 14One study each reported a greater than 10-point decrement in future worries, sleep and fatigue. 12,14D'Agostino et al. reported data for 171 patients using the IONB-PRO, demonstrating a 25-point improvement in the social domain for patients who have had NB for more than 12 months (Table 2). 14

| Reporting biases
Most studies included in this study reported data from all domains.

| Certainty of evidence
From the studies that did not include data for all domains of the PRO  Note: Percentages were calculated by taking the number of trials that reported a concerning level of functioning or symptom burden according to predefined thresholds (numerator) as a percentage of the number of trials reporting that domain (denominator).

| Interpretation of results in context of other evidence
This review is the first to identify patient-reported symptoms and aspects of functioning that have reached concerning levels according to patients with MIBC following UD with IC or NB.The principal findings of this study indicate that physical function was the most concerning PRO for patients with both IC and NB in the short and long term.Additionally, bowel, urinary and sexual bother were concerning symptoms for patients with NB in the long term, but in the short term for those with IC.These issues are covered by using the combination of the EORTC QLQ-C30 and QLQ-BLM30 instruments, which collectively address 18 of 28 concerning issues, therefore we recommend this combination in future studies or for use in clinical practice.
To the best of our knowledge, this is the first systematic review to describe patient-reported functional domains and symptom scales of concern as rated by patients who have undergone IC and NB.Among the studies that assessed physical function in the short and long term, none reported a clinically significant improvement over time, for either UD type (Table 2).A recent prospective study by Abozaid et al. in 2022 used the EORTC QLQ-C30 and EORTC QLQ-BLM30 questionnaires and found that physical function remained below baseline in patients with IC up to 12 months post-RC. 12Contrary to this, Tostivinit et al. in 2021 found in a multicentre observational study of 73 patients using the same questionnaires found that physical function for both NB and IC was associated with global health status and quality of life improvement. 16rthermore, for most scales, patients with IC and NB experienced a higher symptom burden, particularly with sexual function and bother.In the context of the current evidence, Clements et al. in 2022 in a single-centre prospective study demonstrated favourable outcomes in most HRQoL areas except sexual function for IC and NB, and body image for IC. 15 Whilst our findings have found that sexual function was affected in patients with IC, Goldberg et al. in 2016 found as patients grew older, they were less likely to be bothered by decline in sexual function. 17

| Limitations
Our results are limited by the nature of the data reported.Firstly, studies had varied sample sizes, with some studies having less than 20 patients for analysis. 11In some instances, subgroup sizes, for example, gender and operative technique were missing, and therefore, an accurate overall sample size was not completely obtained.Most studies were retrospective (Appendix B).In addition to the presence of heterogenous data within studies utilising the same PRO instrument, the use of diverse measures at different time points made it difficult to compare PRO data directly.Lastly, for scales other than the EORTC QLQ-C30, there were no guidelines or studies present to determine appropriate clinical thresholds of importance, and therefore, we were required to consider our own.

| Implications of results in practice
Our systematic review has demonstrated that the EORTC QLQ-C30 and EORTC QLQ-BLM30 effectively capture the symptoms that affect patients most, therefore, future studies can utilise the combination of these two instruments only to ascertain patient HRQoL.This results in more studies reporting data in a similar manner, and thus, future systematic reviews and trials are less likely to be limited by data heterogeneity.Moreover, having less surveys and instruments delivered to patients can place less burden on the patient, lead to less rates of missing data, and produce higher quality data. 18Furthermore, the practical implication of our results in clinical practice will assist clinicians in identifying areas of need for supportive care interventions for certain patient subgroups.

| Future research
Future studies should aim at assessing the issues that bother patients most in larger cohorts of patients to accurately describe PRO over time, and thus provide clinicians with further insight into ways of optimising patient outcomes throughout the perioperative course.
PRO measures are essential in optimising patient outcomes, as they provide both the patient and clinician with insight into issues that can burden patients in the peri-operative course.Our systematic review has captured the issues that reach clinical thresholds for clinical importance post RC and UD for MIBC.Furthermore, we recommend the EORTC QLQ-C30 and EORTC QLQ-BLM30 instruments as a sufficient PRO measure to identify the issues that impact patients most.The combination of these two instruments only for future studies will make it easier to accumulate prospective data and be less bur-

3. 4 |
Results of syntheses 3.4.1 | Patient reported outcomes of importance Short term (3-12 months) following radical cystectomy with ileal conduit There were 13 studies with PRO data reported in the first 12 months following IC, which used seven different PRO measures F I G U R E 1 Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow chart.IC, ileal conduit; NB, neobladder; UD, urinary diversion.
Study characteristics and patient demographics.Patient reported outcome (PRO) scales that reached clinically relevant thresholds, and studies reporting clinically relevant change in PRO domains over time.