Systematic Review and Cumulative Analysis of the Combination of Mitomycin C plus Bacillus Calmette-Guérin (BCG) for Non–Muscle-Invasive Bladder Cancer

This systematic review and cumulative analysis aimed to explore the efficacy and safety of the combination of intravesical mitomycin C (MMC) plus bacillus Calmette-Guerin (BCG) for non-muscle-invasive bladder cancer (NMIBC) patients. A comprehensive literature search using Pubmed, Embase, Medline, Cochrane Library, CBM, CNKI and VIP databases was performed to identify studies applying intravesical MMC plus BCG therapy on NMIBC patients up to June 2016. Summarized unadjusted odds ratios (ORs) with 95% confidence intervals (CIs) were calculated to assess the efficacy and safety of the combination therapy. A total of 25 studies containing 2749 NMIBC patients were included in this systematic review. Compared with BCG monotherapy, the combination therapy could significantly reduce the tumor recurrence rate (OR = 0.64, 95% CI: 0.44–0.94, P = 0.02) and cancer-specific mortality (OR = 0.54, 95% CI: 0.34–0.87, P = 0.01), without more toxicities (OR = 0.58, 95% CI: 0.17–1.94, P = 0.37). The combination therapy could also lead to significant lower tumor recurrence rate than MMC monotherapy (OR = 0.41, 95% CI: 0.24–0.69, P = 0.0009). Our study indicates that the combination of MMC plus BCG instillation is an effective and safe adjuvant treatment for NMIBC patients.


Discussion
This systematic review aimed to evaluate the efficacy and safety of combined intravesical MMC plus BCG instillation as a novel adjuvant therapy for NMIBC. Our analyses concluded that, compared with BCG or MMC monotherapy, the combination therapy could reduce the recurrence rate of NMIBC significantly without causing more toxicities. As a result, all evidences we have achieved till now support that combined intravesical MMC plus BCG instillation may be a better choice for NMIBC patients.
Previous studies have shown that the adherence to bladder wall of BCG is an important step for immunotherapy 35,36 . Chemical disruption of the bladder urothelium induced by MMC could enable BCG to attach more efficiently to bladder wall and then improve the immune response and antitumor activity 37 . Furthermore, MMC instillation could also promote BCG uptake and activate related immune effector cells [38][39][40] . Therefore, an enhanced antitumor effect could be achieved by combined intravesical MMC and BCG instillation. So far, several studies [8][9][10][11][12][13] have investigated the antitumor effect of combined intravesical MMC plus BCG instillation. Lan et al. 41 recently conducted a meta-analysis including only RCTs, having compared the efficacy of combined BCG and MMC therapy with each monotherapy on NMIBC patients. Results from 8 RCTs in their study showed a significant decreased recurrence rate in patients receiving combination therapy compared with monotherapy. However, since a lot of comparative and cohort studies were not included, their conclusions appeared to be rigorous to some extent. Some animal experiments also drew a similar conclusion with us 22, 42 . Matsushima et al. 42 found MMC plus BCG treatment could inhibit tumor growth and cellular proliferation, and prolong the survival period compared to the BCG-alone therapy through an orthotopic bladder cancer model. Moreover, Svatek et al. 22 identified macrophages were polarized toward a beneficial M1 phenotype after MMC plus BCG instillation in a murine model of bladder cancer, which indicated the antitumor effect of combination instillation could be improved by increased number of beneficial cells.
In this systematic review, we recognized that different combination regimens were carried out in these studies, which might have caused varied effects. Table 3 showed that combination regimen 4 could reduce recurrence but lead to more severe side-effects than others. While considering delaying tumor progression and reducing cancer-specific mortality in long-term follow-up, combination regimen 2 might be a better choice. Several courses of MMC before sequential BCG instillation could not only improve the antitumor function, but also promote the activation and production of immune effector cells [38][39][40] . Nevertheless, since these findings were not obtained by statistical comparisons and cumulative analyses of different combination regimens, they should only represent our own opinions and could not be regarded as evidential results.
Solsona et al. 13 conducted a RCT demonstrating that combined MMC plus BCG therapy was more toxic than BCG alone with severe side-effects rate of 9.5%. However, our analysis indicated that taking all clinical trials into  consideration, combination therapy did not cause more toxicities than BCG or MMC monotherapy. Therefore, combination of MMC plus BCG treatment seems to be safe, while more clinical studies are still needed for further evaluation. Several potential limitations should be addressed about this analysis. First, included studies lasted a time span as long as 21 years, during which the living environment and quality of life might change. Second, data of some studies was incomplete even by contacting authors. Third, most high-quality trials were conducted in Europeans   and Asians, which might restrict the application of our results on other populations. At last, insufficient numbers of related studies might bring some potential bias to our results.

Conclusion
Our study concluded that combination of MMC plus BCG intravesical instillation was an effective and safe adjuvant treatment for NMIBC patients after TUR. This therapy could significantly reduce the tumor recurrence rate and would not bring more toxicities than BCG or MMC monotherapy. However, further high-quality clinical trials are still needed to verify conclusions of our study.

Materials and Methods
Search strategy. A systematic literature search using Pubmed, Embase, Medline, Cochrane Library, CBM, CNKI and VIP databases was performed to identify studies exploring the efficacy of intravesical MMC plus BCG therapy for NMIBC patients up to June 2016. Search terms were "'mitomycin C' or 'MMC'" and "'bacillus Calmette-Guerin' or 'BCG'" in combination with "'non-muscle-invasive bladder cancer' or 'NMIBC' or 'superficial bladder cancer' or 'orthotopic bladder cancer' or 'bladder carcinoma in situ'". The study language was restricted to English and Chinese. Reference lists of relevant studies were also checked.
Inclusion and exclusion criteria. Studies applying intravesical MMC plus BCG therapy on NMIBC patients and providing detailed information were included in this systematic review, and data comparing the efficacy of combination therapy with MMC or BCG monotherapy was pooled in cumulative analysis. Accordingly, we excluded studies involving congress abstracts, conference proceedings, editorials, reviews, animal experiments and repeated publications. Two authors (T.D. and B.L.) independently assessed relevant records, evaluated the quality of included studies and extracted studies' data. Discrepancies were resolved via open discussion.
Study quality assessment and data extraction. GRADE approach was used to assess the LOE of all eligible studies 43 . Furthermore, the Cochrane Collaboration Risk of Bias Tool was applied to evaluate the quality of RCTs 44 . Data was attentively extracted including research methodology, participants' information, tumor stage, surgical procedure, therapeutic regimens of MMC plus BCG (instillation schedule, dose and retaining time), course of treatment, and disease-related outcomes (recurrence, progression, disease-free survival, disease-free interval, cancer-specific survival, overall survival and severe side-effect). In comparative studies, HRs and 95% CIs were also extracted to predict the recurrence-free survival between combined MMC plus BCG and MMC or BCG alone.

Statistics analysis.
In the cumulative analysis, summarized unadjusted ORs and 95% CIs were calculated to assess the efficacy of combined MMC and BCG instillation compared with MMC or BCG alone. Available multivariable adjusted HRs were also pooled as references. Subgroup analyses were conducted according to type of combination regimen, study design, patient ethnicity, number of instillation, therapeutic course, and follow-up time. Statistical heterogeneity among included studies was tested through chi-square test 45 . If no heterogeneity existed with p value > 0.10, the fixed-effect model was used. Otherwise, the random-effect model was applied. A two-sided p value < 0.05 was considered significant for all results in cumulative analysis. Publication bias was assessed by inverted funnel plot and Egger's test 46 . All statistical analyses were conducted by RevMan (version 5.3; Cochrane Collaboration, Oxford, UK) and STATA (version 13.0; StataCorp, College Station, Texas, USA) software.