According to the EAU guidelines for NMIBC, BCG instillation is the standard treatment strategy for high-risk NMIBC[7]. As many as 25–45% of patients do not benefit from intravesical BCG therapy[8]. Two earlier studies revealed that within a median follow-up time of approximately 2 years, the recurrence and progression rates were relatively high in NMIBC, even though BCG instillation was utilized[18, 19]. Radical cystectomy (RC) is another recommended strategy for high-risk NMIBC, which carries a high risk of perioperative morbidity and mortality, especially in an old and infirm population[5]. It has been suggested that RC can impair patients’ gut function and metabolism, leading to malnutrition, sarcopenia and frailty[20]. All these factors impact the outcomes of patients with bladder cancer. Moreover, owing to urinary diversion, radical cystectomy greatly affects the quality of life of patients. As a result, many patients prefer to reject surgery because of the concern of decline in living standards. To decrease the recurrence rate and delay the progression of NMIBC, so that eventual RC surgery could be avoided, combination therapies and checkpoint inhibitors (CPI) are current research focuses. Pembrolizumab, a CPI, has recently received FDA approval in the treatment of BCG-resistant NMIBC[21], but not for all high-risk NMIBC patients. There are also several ongoing phase I and II trials on CPI for NMIBC (NCT04738630, NCT03317158, NCT05120622, NCT04640623)[22], but no outcome was published so far. The key respect and innovative point of combination therapies is to improve survival rate and prognosis of high-risk NMIBC patients with relatively less costs through ameliorating the standard local treatment of NMIBC rather than systemic treatment.
Among BCG combination therapies, intra-arterial chemotherapy (IAC) is rarely mentioned, which was first used in muscle-invasive bladder cancer and has proven to preserve the bladder or prolong life[9–12]. Our previous studies reported that IAC + IVC was superior to IVC alone in preventing tumor recurrence and progression in NMIBC[14–16]. Another study revealed that, compared to BCG immunotherapy, IVC + IAC has almost equal effects in treating NMIBC[17]. The mechanism of action of BCG is currently considered to be the inflammatory response induced by BCG [23]. Cisplatin and anthracyclines are commonly used in IAC, the former kills tumor cells via oxidative stress-mediated cytotoxicity, cell apoptosis [24], while the latter achieve anti-tumor activity by preventing DNA replication in cancer cells. Certainly, these agents could supplement the anti-tumor effect of BCG owing to distinct mechanisms. Therefore, it is worth exploring whether IAC could bring a gain effect on the therapeutic effect of BCG immunotherapy and ascertaining potentially additional side effects.
In this study, we compared the parameters representing efficacy and toxicity of the two different treatment protocols. The recurrence rate in the BCG alone group was close to a previous study reported by Tom J.H et al. (35.8% vs. 35.2%)[25]. However, it was slightly lower than the rate of 26.4% reported by our previous study[17]. We attribute this to the fact that the follow-up time of both groups in current study were much longer than our previous study (35 and 37 months, respectively. vs. 28 and 25 months). It is rational that tumor recurrence raised with follow-up time extended. The Kaplan-Meier survival analysis showed that the difference in recurrence rate between the two groups was remarkable (p < 0.05). Multivariate analysis for recurrence and progression also revealed that BCG + IAC therapy has better performance in preventing recurrence and progression than BCG alone therapy (p = 0.027 and 0.024, respectively). The hazard ratios which were below 1 indicated that BCG + IAC intervention were probably negatively correlated with tumor recurrence and progression. These outcomes suggest that IAC could further reduce the recurrence rate of BCG immunotherapy. It could be an effective reinforcement for the weakness of BCG immunotherapy. To explain the mechanism of IAC, Hoshi et al.[26] performed a series of animal surgeries on rabbits and found that due to direct arterial administration, antitumor drug concentrations in the plasma and bladder tumors were much higher in the IAC group. Higher drug concentrations could lead to the reduction of potential micrometastases into smooth muscle or pelvic lymph nodes and, therefore, decrease the incidence of recurrence.
Compared to two earlier studies, the progression rate of the IAC + BCG group in our study was quite lower than those of BCG groups (8.9% vs. 25.3% and 36%)[27, 28]. Besides, a significant difference was found in the progression rate between the two experimental groups (p = 0.019). It is widely accepted that long-term accumulation of genetic alterations in cancer cells plays a crucial role in tumor progression[29]. Eradicating more residual mutated cells after TURBT at an early stage will most likely delay NMIBC progression. The above mentioned could probably be the mechanism of IAC treatment in preventing tumor progression. Additionally, according to a systematic review[30], progression in high-risk NMIBC mainly occurs within 48 months, and the median follow-up times of our study haven’t reach the time limit yet. Hence, continuation of follow-up and further prospective studies are still required to verify the more authentic effect of IAC treatment in slowing tumor progression.
Regarding the safety, routine blood changes and gastrointestinal distress were the most common complications. Approximately 22.2% of patients (10/45) experienced adverse effects of IAC. According to the CTCAE v4.0 grading system, the grades of adverse reactions were limited to grade 1 to 2, meaning that they were mild enough to be managed at home or outpatient. These results consist with previous studies.[15, 17] On the contrary, including patients who previously withdrew, 81.5% patients (88/108) suffered from adverse reactions of BCG immunotherapy. Notably, four patients in the BCG + IAC group discontinued treatment because of serious side effects of BCG rather than IAC. No statistically significant difference in side effects was observed between the two groups, indicating that IAC causes reversible and tolerable adverse reactions and brings no additional toxicity to BCG instillation.
Our study has a few limitations. The study had a retrospective design, leading to inevitable bias within the data collection procedure. The follow-up time was not long enough, probably limiting the strength of the conclusion. Inadequate follow-up time and sample size also caused unsatisfactory results of univariate analyses. It is widely known that tumor grade, stage, CIS, history of recurrence, etc. are all risk factors for recurrence and progression. However, in our current study only multifocality was found to be related to recurrence, while history of recurrence was the only significantly associated factor of progression. To solve these problems, further multicenter, large-scale prospective studies are recommended.