IMPACT OF SEX ON RESPONSE TO BCG IN NON-MUSCLE INVASIVE BLADDER CANCER PATIENTS: A CONTEMPORARY REVIEW FROM A TERTIARY CARE CENTER

Abstract

Purpose:

Female sex has been implicated with higher stage at diagnosis and as a negative prognostic factor amongst patients with non-muscle invasive bladder cancer(NMIBC). Whether this holds true with contemporary management paradigms is unknown. We analyzed a cohort of patients treated with adequate bacillus Calmette-Guerin(BCG) for NMIBC in an effort to identify sex-specific influence on BCG response.

Methods:

An IRB-approved review of patients with NMIBC treated at our institution with at least ‘adequate BCG’, as defined by the US FDA and EAU, from 2000–2018 was performed. Patients were then stratified by sex and response to BCG. Non-parametric tests were used to summarize the data overall and by groups. The Kaplan-Meier product limit method was used to calculate median survival endpoints.

Results:

Of the 541 patients treated with adequate BCG, 111(20.5%) were female and 430(79.5%) were male. Female patients were younger (median 66 vs. 69, p=0.071), had a lower BMI (median 27.3 vs. 28.8, p=0.010) and were more likely to have no smoking history (49.5% vs 27.0%, p<0.001). Tumor characteristics with respect to stage, size, multifocality, presence of carcinoma in situ, and presence of variant histology were similar between sexes. While rates of recurrence were higher in females than in males this, was not statistically significant (44.1% vs. 34.7%, p=0.064) and Kaplan-Meier estimates of recurrence-free, progression-free and overall survival demonstrated no significant difference between sexes (p=0.409, p=0.253, p=0.171, respectively).

Conclusion:

In a contemporary cohort of patients with NMIBC treated with adequate BCG, female sex was not associated with adverse oncologic outcomes.

INTRODUCTION

Bladder cancer (BCa) is the second most common genitourinary malignancy with 81,4000 new diagnoses of BCa expected in the United States in 2020¹. Of these, 19,3000 (23.7%) will occur in females with 4,930 expected female BCa-related deaths. Amongst those diagnosed with BCa, approximately 75% will present with non-muscle invasive bladder cancer (NMIBC)².

Despite women representing a minority of new BCa diagnoses, significant sex-related disparities in presentation and outcomes have been reported with women often presenting with more aggressive disease and having worse disease-specific mortality^3–5. Research efforts in recent years have focus on sex-related differences in patients with muscle invasive bladder cancer (MIBC) with a paucity of data specifically focused on sex-related outcomes in NMIBC. Currently published literature is conflicting on the sex-specific differences in disease recurrence after treatment of NMIBC; some studies have found females are at increased risk of recurrence^6,7 and progression⁷, while others have found female sex to be protective for recurrence⁸. Further complicating interpretation of the data is the lack of standard bacillus Calmette-Guerin (BCG) utilization and outcomes reporting in many of these retrospective studies. Palou et al. evaluated 146 patients with high-grade(HG) T1 disease, however, none received maintenance BCG therapy⁷. Similarly, in the study by Jancke and colleagues only 80 (21.3%) of 374 intermediate and high-risk patients received BCG⁸.

In this study, we sought to evaluate the impact of sex on BCG response of NMIBC in a contemporary cohort at a tertiary care center.

METHODS

We used an institutional review board approved retrospectively collected database of patients diagnosed with NMIBC to select those who underwent at least one complete induction course (five of six planned instillations) of intravesical BCG at the University of Texas MD Anderson Cancer Center between 2000–2018. Analysis was further restricted to only those patients who received ‘adequate’ BCG therapy according to the definition proposed by the US Food and Drug Administration (FDA), the European Association of Urology (EAU), and the International Bladder Cancer Group (IBCG), defined as at least five of six induction instillations plus at least two additional instillations as a component of either maintenance or re-induction within a 6-month time period^9,10. Surveillance schedules were standardized among all providers and based on available NMIBC guidelines. The use of peri-operative chemotherapy, restaging transurethral resection (TUR), BCG maintenance or re-induction were left to the discretion of individual providers. Peri-operative chemotherapy was defined as a one-time instillation of gemcitabine or mitomycin-C.

Clinicopathologic data was collected for all patients and included age, sex, body mass index (BMI), smoking history, prior BCG exposure, tumor stage, tumor size and multifocality, presence of variant histology, presence of carcinoma in situ (CIS), performance of re-staging TURBT, peri-operative intravesical chemotherapy, pathologic staging, number of BCG instillations, response to BCG, subsequent treatments, and disease outcomes.

T-tests (or Wilcoxon’s rank sum) and Pearson’s chi-square (or Fisher’s exact test) were used to compare variables by groups. Recurrence-free survival (RFS) and progression-free survival (PFS) were calculated as the number of months from the start of first BCG instillation to event date or death date. Patients without an event at their last follow-up were censored on that date. Overall survival (OS) was calculated as the number of months from start of first BCG instillation to death or last follow-up date. Patients who were alive at their last follow-up were censored on that date. The Kaplan-Meier product limit method was used to estimate the median survival endpoints. Cox proportional hazards models were used to determine the association between RFS and potential risk factors. The full model multivariate analysis included all the variables with p<0.25 in the univariate analyses or important clinical variables such as gender. Backward elimination selection methods were used the final reduced model. For all comparisons, a two-sided p-value of <0.05 was considered statistically significant. Statistical analysis was performed using Stata/SE version 16.1 statistical software (State Corp. LP, College Station, TX).

RESULTS

Table 1 shows the clinicopathologic characteristic of the 111 female patients (20.5%) and 430 male patients (79.5%). Median follow-up was 45.2 months for female patients (interquartile range [IQR] 25.7–95.0) and 45.5 months for males (IQR 23.6–74.3). Female patients were younger at diagnosis (66 vs. 69 years, p=0.071), had lower BMI (27.3 vs. 28.8, p=0.010) and were more likely to have no smoking history (49.5% vs. 28.9%, p<0.001). Rates of prior BCG therapy were similar between sexes (6.3% vs. 8.6%, p=0.430). There were no differences between sexes in tumor characteristics with respect to initial tumor stage, grade, size, presence of carcinoma in situ (CIS) or variant histology. Additionally, males and females demonstrated similar EAU risk group distribution. 86% of females and 84.5% of males with T1 tumors underwent guideline-recommended re-staging TURBT (p=0.791). Receipt of peri-operative chemotherapy was similar between the sexes (13.6% vs. 14.0%, p=0.912).

Table 1:

Patient demographics and tumor characteristics in all NMIBC patients stratified by sex

Variable	Female (n=111)	Male (n=430)	P
Age, median (IQR)	66 (59–74)	69 (62–76)	0.071
BMI, median (IQR)	27.3 (23.3–30.9)	28.8 (25.4–32.6)	0.010
Smoking Status, n(%)			<0.001
Never	55 (49.5)	116 (27.0)
Current/Former	56 (50.5)	314 (73.0)
Prior BCG, n(%)			0.430
No	104 (93.7)	393 (91.4)
Yes	7 (6.3)	37 (8.6)
Tumor Stage			0.872
Ta	52 (46.8)	201 (46.7)
Tis	9 (8.1)	29 (6.7)
T1	50 (45.0)	200 (46.5)
Tumor Grade			0.140
Low	15 (13.5)	38 (8.8)
High	96 (86.5)	392 (91.2)
Tumor Size, n(%)			0.717
≤3 cm	69 (65.1)	261 (63.2)
>3 cm	37 (34.9)	152 (36.8)
Focality, n(%)			0.161
Solitary	48 (43.2)	217 (50.7)
Multifocal	63 (56.8)	211 (49.3)
Variant Histology, n(%)			0.592
No	107 (97.3)	400 (95.5)
Yes	3 (2.7)	19 (4.5)
CIS, n(%)			0.733
Absent	77 (69.4)	291 (67.7)
Present	34 (30.6)	139 (32.3)
T1 with Re-staging TURBT, n(%)			0.791
No	7 (14.0)	31 (15.5)
Yes	43 (86.0)	169 (84.5)
Peri-operative chemotherapy, n(%)			0.912
No	95 (86.4)	361 (86.0)
Yes	15 (13.6)	59 (14.0)
EAU Risk Stratification			0.801
Low	7 (6.3)	18 (4.2)
Intermediate	23 (20.7)	97 (22.6)
High	58 (52.3)	224 (52.1)
Very High	23 (20.7)	91 (21.2)

Time from TURBT to initiation of induction BCG (median 45 days in females vs. 52 days in males, p=0.222) and mean number of BCG doses administered (19.7 doses vs 19.5 doses, p=0.854) were the same between sexes (Table 2). Recurrence rates were higher in females than males (44.1% vs 34.7%, p=0.064) but not statistically significant and rates of any stage progression and progression to muscle invasive or metastatic disease were also similar between sexes (9.0% vs 11.6%, p=0.433 and 5.4% vs. 7.9, p=0.366, respectively). Rates of radical cystectomy were also similar between groups (11.7% vs. 12.3%, p=0.860). When we further stratified female patients by history of smoking there were no differences between the two groups, no smoking history (55 patients, 49.5%) vs. former or current smokers (56 patients, 50.5%), with respect to recurrence, treatment failure or progression (Supplementary Table 1). There were no significant differences in RFS, PFS, and OS when patients were stratified by sex. (Figure 1).

Table 2:

Response to BCG by sex

Variable	Female	Male	P
Total # of BCG Doses (mean, SD)	19.7 (8.4)	19.5 (7.4)	0.854
Time from Diagnosis to Induction BCG (median, IQR)	45 (27–71)	52 (32–71)	0.222
Recurrence, n(%)			0.064
No	62 (55.9)	281 (65.3)
Yes	49 (44.1)	149 (34.7)
BCG Relapse, n(%)			0.832
No	92 (82.9)	360 (83.7)
Yes	19 (17.1)	70 (16.3)
BCG Unresponsive, n(%)			0.264
No	87 (79.1)	357 (83.6)
Yes	23 (20.9)	70 (16.4)
Progression on BCG (any stage), n(%)			0.433
No	101 (91.0)	380 (88.4)
Yes	10 (9.0)	50 (11.6)
Progression to MIBC or distant metastatic disease, n(%)			0.366
No	105 (94.6)	395 (92.1)
Yes	6 (5.4)	34 (7.9)
Systemic chemotherapy, n(%)			0.852
No	101 (91.0)	390 (91.5)
Yes	10 (9.0)	36 (8.5)
Radical cystectomy, n(%)			0.860
No	98 (88.3)	377 (87.7)
Yes	13 (11.7)	53 (12.3)

Figure 1:

Kaplan-Meier survival curves for Recurrence-Free Survival, Progression-Free Survival, and Overall Survival

On univariate and multivariate analysis, sex was not associated with recurrence-free survival (Hazard ratio [HR] 1.14, 95% Confidence Interval [CI] 0.84–1.55, p=0.409 and HR 1.11, 95% CI 0.80–1.53, p=0.532, respectively). Patient characteristics that were significantly associated with recurrence on univariate analysis were advanced age, receipt of prior BCG and lack of re-staging TURBT. On multivariate analysis, prior BCG exposure and lack of re-staging TURBT were the main prognostic factors for recurrence (Table 3).

Table 3:

Univariate and Multivariate Models for Recurrence-Free Survival

	Univariate analysis			Multivariate analysis
Variable	HR	95% CI	P	HR	95% CI	P
Sex (male/female)	1.14	0.84–1.55	0.409	1.12	0.83–1.53	0.457
BMI*	0.98	0.96–1.00	0.129
Prior BCG	2.28	1.55–3.35	<0.001	2.19	1.48–3.21	<0.001
Re-staging TURBT	0.69	0.53–0.90	0.006	0.72	0.56–0.94	0.016
Smoking history (current/former, never)**	1.04	0.79–1.38	0.771
EAU Risk Stratification
Low
Intermediate	1.39	0.63–3.07	0.420
High	1.59	0.74–3.41	0.232
Very High	1.50	0.68–3.32	0.314

^*Cox regression with continuous covariate

^**This variable was not included in the multivariate analysis.

DISCUSSION

In our study of patients with NMIBC treated with adequate BCG in a contemporary fashion, we found that sex does not influence response to BCG therapy.

There are discordant findings in the literature surrounding the impact of female sex on recurrence following treatment for NMIBC. Female sex has previously been identified as a risk factor for disease recurrence and is one of the prognostic factors incorporated into the Spanish Urological Club for Oncological Treatment (CUETO) scoring model derived from 4 CUETO trials in which the majority, but not all, received adequate BCG therapy¹¹. A recent meta-analysis by Uhlig et al. confirmed these findings indicating that females had an 11% higher risk of disease recurrence compared to male patients¹². Subgroup analyses demonstrated that sex-specific disparities were most pronounced in studies administering BCG (HR 1.64); conversely, there were no sex-related differences in studies where patients did not receive BCG. A proposed mechanism for the potential sex-related discrepancies in response to BCG includes the role of estrogen on the downregulation of IL-6 and other key cytokines responsible for BCG’s antitumor activity¹³.

Despite the aforementioned studies, there are also numerous studies suggesting that outcomes of males and females with NMIBC treated with BCG are similar. Sex was not a variable included in the European Organization for Research and Treatment of Cancers (EORTC) risk tables as it was not associated with recurrence or progression in the combined analysis of 2596 NMIBC patients in prior EORTC trials¹⁴. Additionally, a retrospective review by Boorjian et al. of 1021 patients (756 men, 265 women) treated with induction BCG found that on multivariate analysis sex was not associated with increased risk of recurrence (HR 0.94, 95% CI 0.79–1.11, p=0.44) or progression (HR 1.18, 95% CI 0.85–1.63, p=0.33)¹⁵. Similar results have also been demonstrated in a multicenter retrospective analysis of 2,451 patients (2,012 men, 439 women) with high-grade T1 NMIBC in which sex was not identified as a prognostic indicator of recurrence or progression¹⁶. Drawing accurate conclusion from the literature is difficult as there is significant heterogenicity in treatments utilized, variable utilization of restaging TUR, and the majority of studies included patients who underwent induction BCG only rather than the current standard of induction followed by maintenance therapy.

In our present study of NMIBC patients who have received adequate BCG, there was a non-significant increase in recurrence amongst female patients (44.1% vs 35.0%, p=0.076) but there was no difference in progression rates. Additionally, the majority of patients with T1 tumors underwent guideline recommended re-staging TURBT.

Our study is one of few in which sex differences have been evaluated in patients undergoing current standard of care treatment for NMIBC including re-staging TURBT, when indicated, and adequate BCG, both of which have previously been shown to impact recurrence and progression rates^17,18. When receiving the current standard of care treatment for NMIBC at our institution, there were no differences in BCG response rates between sexes. Our analysis did demonstrate a trend towards increased recurrence in females, albeit our sample size was small (111 females received adequate BCG) and the female recurrence rates did not reach statistical significance. Nonetheless, this finding, in combination with the previous meta-analysis by Uhlig et al.¹² highlight that there may in fact be a difference in BCG response amongst female patients, especially those receiving long-term maintenance BCG. These findings should be further explored in future prospective studies as they may have implications on patient counseling, surveillance frequency, and treatment in female patients with NMIBC.

Sex disparities in BCa have previously been attributed to carcinogen exposure, delay in diagnosis and treatment, or higher grade/stage at diagnosis amongst females. Similar to a recent large retrospective study of more than 27,000 patients with MIBC, our study of patients with NMIBC did not identify any significant differences in treatment patterns and quality between sexes that would explain differences in outcomes¹⁹. However, differences in BCa incidence and outcomes are influenced by sex hormones, sex-specific differences in immune response, tumor genetics and the bladder microbiome²⁰– all of which can influence response to immune-based therapies such as BCG in NMIBC and immune checkpoint inhibitors in the NMIBC and muscle-invasive settings.

Limitations of our study include its retrospective design, however, there were no significant differences in clinicopathologic features including tumor stage, presence of CIS or variant histology between sexes. Additionally, treatment decisions were at the discretion of the urologic oncologist accounting for potential differences in use of repeat induction BCG, use of second line agents after BCG failure, etc.. Finally, there is a trend in our data towards inferior recurrence among female patients, this was not statistically significant in our analysis, however, our small sample may be inadequately powered to detect a difference.

CONCLUSION

In the present study of patients treated uniformly with adequate BCG therapy at a large tertiary care center, sex did not influence response to BCG nor long-term rates of recurrence or progression. Amongst those receiving adequate BCG, there was a higher recurrence rate in female patients, although this was not statistically significant. Prospective studies further elucidating the role of sexual dimorphism in disease recurrence, innate immune differences and response to treatment are needed.