Improving compliance with guidelines may lead to favorable clinical outcomes for patients with non‐muscle‐invasive bladder cancer: A retrospective multicenter study

Clinical guidelines recommend that patients with non‐muscle‐invasive bladder cancer (NMIBC) should be treated with appropriate adjuvant therapy. However, compliance with guideline recommendations is insufficient, and this may lead to unfavorable outcomes. We aimed to investigate the level of adherence to guideline recommendations in patients with NMIBC and evaluate the outcomes of those who did and did not receive guideline‐recommended therapies.


INTRODUCTION
Non-muscle-invasive bladder cancer (NMIBC) accounts for approximately 75% of newly diagnosed bladder cancer cases and has a 5-year recurrence rate of 30%-80%. 1 Tumor recurrence has a negative impact on health-related quality of life and imposes heavy economic burdens on medical care expenditure; additionally, progression to muscle-invasive bladder cancer (MIBC) is associated with death resulting from bladder cancer. 2,3o reduce the risk of disease recurrence, various treatments, including intravesical bacillus Calmette-Gu erin (BCG), immediate postoperative instillation of chemotherapy (IPIC), additional adjuvant intravesical chemotherapy, and second transurethral resection (TUR), have been performed after initial transurethral resection of bladder tumor (TURBT). 4Many large-scale and high-quality studies have been performed to assess the treatment efficacy of each therapy.The results of multiple randomized controlled trials (RCTs) revealed that maintenance BCG reduced the probability of disease recurrence and might contribute to decreasing the risk of progression. 4,5In a recent systematic review and individual patient data meta-analysis, IPIC reduced the 5-year recurrence rate by 14%. 6A systematic review showed that the second TUR facilitated the detection of residual tumors after the initial TURBT and had the potential to prevent disease recurrence in patients with T1 NMIBC. 7The clinical benefit of each treatment has been well demonstrated.However, in clinical practice, two or more different types of therapies (e.g., second TUR and intravesical BCG therapy) are performed after the initial TURBT, and the outcomes in patients with NMIBC who received them have not been fully clarified.
Clinical practice guidelines for NMIBC recommend stratification of patients with NMIBC into low-, intermediate-, or high-risk groups and provide risk-adapted optimal surveillance and treatment strategies to minimalize the probability of disease recurrence and progression. 4,8However, in real-world clinical practice, a large percentage of patients are not optimally managed according to guidelines, and poor adherence to guideline recommendations may be correlated with unfavorable outcomes. 9Therefore, the present study aimed to investigate the level of adherence to guideline-recommended therapies in patients with NMIBC and analyze the outcomes of those who did and did not undergo them.

Ethics statements
This retrospective study was approved by the Institutional Review Board of Tohoku University Graduate School of Medicine (approval number: 2018-1-450), and it conforms to the provisions of the Declaration of Helsinki.Informed consent was obtained through an online opt-out process.

Study design and population
The records of individual patients who underwent TURBT between January 2009 and December 2016 at nine institutions participating in the Tohoku Urological Evidence-based Medicine Study Group and Kyoto University Hospital were analyzed.The inclusion criteria were as follows: histologically confirmed NMIBC, minimum follow-up of 3 years when patients did not experience disease recurrence or progression, and a minimum follow-up period of 3 months if patients experienced recurrence or progression.The exclusion criteria were as follows: patients with a history of MIBC, non-urothelial carcinoma, and NMIBC who underwent definitive treatment (radiotherapy or radical cystectomy).This study included patients who developed upper urinary tract urothelial carcinoma, underwent radical nephroureterectomy, and then developed intravesical recurrence; however, the European Association of Urology (EAU) risk-stratification system does not include patients with newly diagnosed NMIBC with upper urinary tract recurrence.Thus, we excluded these patients.

Data collection
Study variables at the initial TURBT included age, sex, tumor stage, grade, multiplicity (single or multiple), largest tumor diameter (<3 or ≥3 cm), 2019 EAU risk group, presence of carcinoma in situ (CIS), and tumor status (primary or recurrent).The clinicopathological characteristics of patients with NMIBC who received intravesical BCG therapy or a second TUR were investigated.Information on BCG therapy (the dose, total number of instillations, and whether the initially predetermined dosage and duration of treatment were completed), second TUR (the period between the initial TURBT and second TUR, and pathological findings at the second TUR), and IPIC (the type of chemotherapeutic agents administered intravesically immediately after the initial TURBT) was also evaluated.Recurrence was defined as intravesical recurrence after the initial TURBT.Progression was defined as the development of muscle-invasive disease (T2 or higher) and metastasis to lymph nodes or other distant organs.Patients with NMIBC were stratified into low-, intermediate-, and high-risk groups according to the 2019 EAU risk-stratification system. 10To evaluate the clinical benefit of each therapy, we investigated recurrence-free survival (RFS) in individual patients with intermediate-and high-risk tumors who underwent a second TUR, BCG therapy, and IPIC.RFS was calculated as the period from the initial TURBT to the first recurrence, and patients who developed progression after the initial TURBT were excluded when RFS was evaluated.To assess the level of compliance with the guideline, the proportions of patients with each risk group of NMIBC who underwent the aforementioned therapies recommended in the 2019 EAU guideline were calculated; additionally, those at academic and non-academic facilities were compared.Patients with T1 NMIBC were identified as complying with the guideline only when they received both maintenance BCG therapy and a second TUR (compliance group); those who underwent only one of both therapies and induction-only BCG (including both induction-only BCG and a second TUR) were classified as not complying with the guideline (non-compliance group).Patients with intermediaterisk tumors were considered to adhere to the guideline only when they received either IPIC followed by maintenance BCG therapy or IPIC followed by additional adjuvant intravesical chemotherapy (compliance group); and those who received only IPIC, induction-only BCG, maintenance BCG (without IPIC), additional adjuvant intravesical chemotherapy (without IPIC), or none of the aforementioned therapies were classified as not complying with the guideline (noncompliance group).This study compared RFS between compliance and non-compliance groups by using propensity score-matched analysis.When patients received additional BCG therapy following induction BCG, they were considered to be treated with "maintenance BCG therapy" even if the duration of treatment or dose failed to meet the standards recommended by the 2019 EAU guideline. 10

Statistical analysis
Continuous variables were expressed as median and interquartile range (IQR), and categorical variables were presented as absolute numbers and percentages.The Kaplan-Meier method with the log-rank test was used to estimate and compare RFS between the groups.Categorical variables were compared using the Pearson chi-squared test.Clinicopathological variables including age, sex, tumor status (primary or recurrent), tumor multiplicity, and largest tumor diameter were used for propensity score matching in patients with intermediate-risk NMIBC; and age, sex, tumor status, stage, presence of CIS, grade, tumor multiplicity, and largest tumor diameter were used for that in patients with T1 NMIBC.
Patients were matched at a 1:1 ratio.Differences were considered statistically significant at p < 0.05.When multiple comparisons between three and four groups were performed, the Bonferroni method was used, with p-values <0.017 and 0.008 being considered statistically significant, respectively.All statistical analyses were performed using EZR (version 1.61; Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R. 11

RESULTS
In total, 1204 patients from 10 institutions were included in this study.The median age of the entire cohort at the initial TURBT was 71 years (IQR, 64-79 years).Table 1 shows the clinicopathological characteristics of patients treated and not treated with intravesical BCG therapy or a second TUR.Among patients with T1 NMIBC (n = 330), 184 (55.8%) did not undergo a second TUR.Among the patients with highrisk tumors, 478 (74.0%) were not treated with maintenance Over a median follow-up duration of 56 months, 429 (37.0%) patients experienced disease recurrence after the initial TURBT.Figure 1a,b shows RFS in the entire cohort and patients with low-, intermediate-, and high-risk tumors.The 3-year RFS rate in the entire cohort was 68.4%.Significant differences in RFS were found between the low-and intermediate-risk groups (p = 0.001) and between the lowand high-risk groups (p = 0.002).Figure 2a,b shows RFS in high-and intermediate-risk patients who did and did not receive adjuvant intravesical therapy, including induction-only BCG therapy, maintenance BCG therapy, or additional adjuvant intravesical chemotherapy, respectively.Patients with high-and intermediate-risk tumors who received maintenance BCG therapy had significantly better RFS than those treated with other therapies.Figure 3 shows RFS for patients with T1 NMIBC who did and did not undergo a second TUR.Significantly better RFS was found in patients with T1 NMIBC who underwent a second TUR than in those who did not (p < 0.001) (Figure 3a).Patients who were diagnosed with T1 NMIBC at the initial TURBT and received a second TUR were divided into four groups as follows: patients without residual tumors at the second TUR; patients with Ta tumors (without CIS) at the second TUR; patients with T1 tumors (with or without CIS) at the second TUR; and patients with CIS (with or without Ta tumors) at the second TUR.Significant differences in RFS were found only between patients with T1 tumors at the second TUR and those without residual tumors (p = 0.003; Figure 3b).Figure 4 shows RFS in intermediate-and high-risk patients treated with IPIC.There was no significant difference in RFS in the high-risk group between treatments (Figure 4a, p = 0.94).Additionally, the difference in RFS in patients with    intermediate-risk tumors between treatments was not significant (Figure 4b, p = 0.39).Figure 5a shows RFS in patients with T1 tumors who did or did not receive a second TUR and maintenance BCG therapy.Patients with T1 NMIBC who received them (compliance group) showed significantly better RFS than propensity-matched patients who did not (non-compliance group) (p = 0.031).As shown in Figure 5b, patients with intermediate-risk tumors who received either IPIC followed by maintenance BCG therapy or IPIC followed by additional adjuvant intravesical chemotherapy (compliance group) exhibited significantly greater RFS than propensitymatched patients who did not receive either of the aforementioned therapies (non-compliance group) (p = 0.011).

DISCUSSION
Various mechanisms, including incomplete resection, tumor re-implantation, and the field change cancerization effect, have been considered responsible for the disease recurrence of NMIBC. 12There are evidence-based, comprehensive strategies to counteract those mechanisms; however, the recurrence rate remains high, as reported previously. 6,7,10,12In a previous study, very few patients with NMIBC were managed in a manner recommended by the guidelines, and a significant improvement in outcome was found in patients with NMIBC who received at least half of the recommended care. 9The result suggested that poor adherence to guideline recommendations was partially responsible for recurrence.Intravesical BCG therapy is superior to TURBT alone and TURBT with additional adjuvant intravesical chemotherapy in preventing disease recurrence, as reported previously. 4In a recent systematic review and meta-analysis of RCTs, a 21% reduction in risk ratios of recurrence was observed in patients with NMIBC treated with maintenance BCG therapy compared with those who underwent induction-only BCG therapy. 135][16] Herein, patients with intermediate-and high-risk NMIBC treated with maintenance BCG therapy demonstrated significantly better RFS than those who underwent induction-only BCG therapy (Figure 2a,b).
A systematic review concluded that a second TUR might improve the outcome of patients diagnosed with T1 NMIBC. 7erein, 55.8% of patients diagnosed with T1 NMIBC at the initial TURBT did not undergo a second TUR, and  significantly poor RFS was found in patients with T1 NMIBC who did not receive a second TUR compared with those who did (Figure 3a).The current study's results indicate that a second TUR has the potential to improve the outcomes in patients with T1 NMIBC.Patients with T1 NMIBC at the initial TURBT with T1 histology at the second TUR may be at high risk of disease progression and are possible candidates for radical cystectomy; our data showed that the RFS in these patients was significantly worse than that in patients with T1 NMIBC at the initial TURBT without residual tumors at the second TUR (Figure 3b). 17Pathological findings at the second TUR may have a significant impact on the prognosis of patients diagnosed with T1 NMIBC after the initial TURBT; however, the guidelines do not provide treatment strategies for these patients based on pathological findings at the second TUR. 4,18aintenance BCG therapy and second TUR are recommended for patients with T1 NMIBC in the guidelines. 10,15n some previous studies, better outcomes were observed in high-risk patients treated with a second TUR followed by BCG therapy than in those treated with BCG therapy alone. 19,20In the current propensity score-matched analysis, patients with T1 NMIBC treated with a second TUR, followed by maintenance BCG therapy, showed significantly superior RFS compared with those not treated with both therapies, suggesting that adherence to guideline-recommended therapies improves patients' outcome.
In a recent systematic review and individual patient data meta-analysis of RCTs, IPIC demonstrated a 14% reduction in the 5-year recurrence rate; however, IPIC cannot provide clinical benefit for patients with a prior recurrence rate of >1 recurrence/year, and the European Organization for Research and Treatment of Cancer (EORTC) recurrence score ≥5. 6Consequently, the 2019 EAU guideline did not recommend IPIC for these patients. 10The current data available for analysis lack information on EORTC recurrence score and the prior recurrence rate per year, and it is possible that intermediate-risk patients with a prior recurrence rate of >1 recurrence/year or an EORTC recurrence score ≥5 are identified as complying with the guideline.IPIC may not bring clinical benefit to patients with high-risk tumors, and accordingly, the guidelines do not strongly recommend IPIC. 10,21In the current study, 73.1% of high-risk patients received IPIC, and that did not produce a significant reduction in the risk of recurrence.A systematic review reported that 53.0% of presumed low-and intermediate-risk patients underwent IPIC. 22Herein, 80.8% of intermediate-risk patients underwent IPIC.However, there was no significant difference in RFS between the treatments (Figure 4b), and it is unclear why IPIC did not show treatment efficacy.
4][25] Multiple factors, including patientrelated, condition-related, therapy-related, cancer-related, and socioeconomic factors, are associated with non-compliance to guidelines. 22However, identifying the primary causes of not providing guideline-recommended therapies is difficult, and detailed approaches to improve guideline adherence have not been shown. 25As an attempt to reduce an evidence-practice gap, theory-informed behavior change interventions have been applied; awareness of the evidence-practice gap is the first important step to achieving behavioral change necessary to improve guideline adherence. 26Herein, a higher level of guideline adherence was generally observed in academic facilities compared with non-academic facilities (Table 3).To our knowledge, in Japan, a large-scale research study on guideline adherence for NMIBC has not been performed. 22n investigation with a sufficient number of cases to fully grasp the current situation (guideline adherence), followed by an assessment of potential barriers to the implementation of guideline recommendations and the development of methods to overcome these barriers, may lead to improvement in guideline adherence. 26n the present study, 46 patients (3.8% of the entire cohort) experienced progression after the initial TURBT.The current study did not include patients treated with definitive therapies after the initial TURBT; additionally, it lacked data on those who experienced disease recurrence after the initial TURBT followed by the development of progression.Therefore, our data could not be used to precisely estimate the patients who developed progression.It is important to investigate the association between guideline-recommended therapies and disease progression.A further study is needed to clarify that.In our study, there was no significant difference in RFS between patients with intermediate-and high-risk tumors.Tumor multiplicity (multiple), larger maximum tumor diameter (≥3 cm), and high prior recurrence rate (>1 recurrence per year), which were factors common to both risk groups, were more strongly associated with disease recurrence than the tumor grade (high), presence of CIS, and T category (T1 or Tis); this might explain the lack of difference in RFS between the high-and intermediate-risk groups. 27his study has some limitations.The primary limitation is its retrospective design, and the patients analyzed herein did not receive a standardized follow-up regimen.Moreover, the correlation between RFS and BCG dose or duration of maintenance therapy has not yet been investigated.9][30] Therefore, patients with NMIBC who complied with the JUA or EAU guidelines at the time of treatment might be classified as not complying with the 2019 EAU guideline.Thus, such cases were considered as non-compliance with the 2019 EAU guideline in this study.We performed a propensity score-matched analysis aimed at comparing RFS between patients with T1 tumors who did and did not receive therapies recommended by the 2019 EAU guideline; however, a similar analysis for patients with high-risk tumors was not carried out.We believe that, regardless of these limitations, our data provide real-world outcomes of patients treated with a second TUR, intravesical therapy, and their combinations and information on the level of adherence to guideline recommendations.
In conclusion, our data revealed that compliance with guideline recommendations was poor, and unfavorable outcomes were observed in patients with NMIBC who did not receive guideline-recommended therapies.The results of the current study suggest that improvements in adherence to guidelines may contribute to improved clinical outcomes.

Editorial Comment
Editorial Comment to "Improving compliance with guidelines may lead to favorable clinical outcomes for patients with non-muscle-invasive bladder cancer: A retrospective multicenter study" In the present study, Sato et al. 1 focused on the important topic of whether guideline-recommended therapy affects the oncological outcome of non-muscle-invasive bladder cancer (NMIBC) using 1204 patients collected from 10 institutions.They found that, surprisingly, 74% of their high-risk patients did not receive maintenance BCG therapy.Moreover, propensity scorematched analysis of T1 NMIBC patients who received a second TUR as well as maintenance BCG therapy revealed that they had a favorable recurrence-free survival (RFS).
One of the concerns about the study is that the definition of maintenance BCG therapy was ambiguous.The authors defined maintenance BCG therapy as patients who received additional BCG instillation following induction BCG.Although the strict schedule of maintenance BCG instillation has not yet been established in Japan, Hinotsu et al. 2 reported that 18-month BCG maintenance therapy contributed to significant prolongation of RFS compared to BCG induction therapy in patients with recurrent or multiple NMIBC.The Japanese Urological Association guidelines state that the regime of BCG maintenance should be based on the SWOG 8507 3 trial for at least 12 months.Thus, further studies are warranted to clarify the importance of guideline adherence using the BCG maintenance schedule reported by SWOG 8507 or Hinotsu et al.
Some part of initial T1 disease, so-called very high-risk tumor, needs to be treated with radical cystectomy (RC) when additional risk factors, such as concomitant non-UC subtype, presence of lymphovascular invasion, and concurrent carcinoma in situ in either bladder or prostatic urethra, are included.However, the compliance of RC in very high-risk NMIBC patients is low.We previously reported 4 the importance of performing RC during the time course of NMIBC stage in accordance with the guidelines' recommendation.In this report, when NMIBC patients who followed the guidelines from initial transurethral resection of bladder tumors to RC were defined as a guideline-adherent group, guideline nonadherence was independently associated with cancer-specific death.The importance of guideline-adherent therapy in T1 disease, including RC for very high-risk tumor, needs to be clarified in a future study.

FIGURE 1
FIGURE 1 Kaplan-Meier plots displaying RFS.Panel (a) represents RFS in the entire cohort.Panel (b) represents RFS stratified according to the low-, intermediate-, and high-risk groups classified by the 2019 EAU risk-stratification system.EAU, European Association of Urology; RFS, recurrence-free survival.

FIGURE 4
FIGURE 4 Kaplan-Meier curves for recurrencefree survival in patients with (a) high-and (b) intermediate-risk non-muscle-invasive bladder cancer stratified according to treatment (IPIC vs. no IPIC).IPIC, immediate postoperative instillation of chemotherapy.

FIGURE 5
FIGURE 5 (a) Kaplan-Meier plots displaying RFS in patients with stage T1 NMIBC who received both maintenance BCG therapy and a second TUR (compliance group), and propensity-matched patients who did not receive them (non-compliance group).(b) Kaplan-Meier curve for RFS in patients with intermediate-risk NMIBC who underwent either IPIC followed by maintenance BCG therapy or IPIC followed by additional adjuvant intravesical chemotherapy (compliance group), and propensity-matched patients who received only IPIC, induction-only BCG, maintenance BCG (without IPIC), additional adjuvant intravesical chemotherapy (without IPIC), or none of the aforementioned therapies (non-compliance group).BCG, bacillus Calmette-Gu erin; IPIC, immediate postoperative instillation of chemotherapy; NMIBC, non-muscle-invasive bladder cancer; RFS, recurrence-free survival; TUR, transurethral resection.

TABLE 1
Clinicopathological characteristics of patients who did and did not undergo intravesical BCG therapy or a second TUR.BCG therapy.Among patients with intermediate-risk tumors, 41 (9.0%) and 23 (5.1%) received maintenance BCG therapy or additional adjuvant intravesical chemotherapy, respectively.Information on BCG therapy, second TUR, and IPIC is shown in Table2.Of the patients who underwent BCG therapy, approximately 70% were treated with a full dose of BCG and completed the initially intended dosage and duration of BCG therapy.Table3displays the compliance rates in the whole cohort and those in academic or non-academic facilities.Except for IPIC in patients with low-and intermediate-risk tumors, academic facilities exhibited significantly higher compliance rates than non-academic facilities.

TABLE 2
BCG treatment schedule and pathological results of the second TUR and intravesically administered chemotherapeutic agents immediately after the initial TURBT.

TABLE 3
Proportion of patients with each risk group of NMIBC treated with adjuvant intravesical therapy and a second TUR.