**3. Discussion**

This study demonstrated that intravesical injections of either PRP or BoNT-A are safe and effective in IC/BPS symptoms improvement without significantly increasing the PVR. Bladder pain reduction was not significant between groups at 6 months. Patients who received PRP injections did not experience UTI or dysuria after treatment. However, only half of the study cohort had a GRA of ≥2 at the six-month follow-up, and the remaining patients sought additional bladder therapy to improve symptoms. Although the therapeutic efficacy for IC/BPS at 6 months was similar between PRP and BoNT-A, the patients who received the BoNT-A injection had a greater potential risk for UTI after treatment.

The pathophysiology of IC/BPS is complicated and not completely understood. The recent IC Data Base study noted that loss of epithelial integrity is a predominant histopathologic finding in patients with IC/BPS. The epithelial damage may precede other histopathologic findings, such as suburothelial inflammation and sensory nervous activation in the bladder wall, leading to increases in the sensation of bladder fullness and pain in response to bladder inflammation [9,15]. The inflammation of IC/BPS might increase the sensory neuropeptides release from the suburothelial nervous network and integrate the signal transmission from urothelium to the detrusor muscles. In animal models of chemical cystitis or human studies of IC/BPS, detrusor injection of BoNT-A has been shown facilitate an increase in bladder capacity and relief of bladder pain [16,17]. Inhibition of sensory fiber neuroplasticity and inflammation in the suburothelial space by BoNT-A injections provides good therapeutic efficacy in patients with IC/BPS [18]. However, a single BoNT-A injection might not be adequate to provide long-term durability for IC/BPS. Our previous study also demonstrated that four consecutive BoNT-A injections provide a longer therapeutic duration than that of treatment with fewer injections [19].

An intravesical injection of 100–200 U of BoNT-A followed by cystoscopic hydrodistention has been reported effective in decreasing bladder pain and nocturia in patients with IC/BPS [20]. A randomized, double-blind clinical trial demonstrated that 100 U of BoNT-A

is safe and effective for treatment of IC/BPS [21]. Patients who received BoNT-A treatment had a significantly greater improvement in bladder pain reduction and bladder capacity increase than that of those in the placebo group. The therapeutic effects of BoNT-A on IC/BPS are result from inhibition of the noxious neurotransmitter releases [17]. Based on the above clinical and basic science evidence, BoNT-A injection was listed as the fourth-line therapeutic option in the treatment of IC/BPS in the AUA guidelines [22].

Although the BoNT-A injection seems promising for treating symptoms of IC/PBS, long-term results have not revealed a successful outcome [21]. The limited successful result is possibly due to inadequate control of chronic inflammation inside the urinary bladder [17]. Repeated intravesical BoNT-A injections were recently performed for patients with refractory IC/PBS, and the therapeutic effects appear promising. Approximately 70% of patients with non-ulcer type IC/PBS may benefit from repeated BoNT-A injections every six months [23]. A previous immunohistochemistry study also confirmed the reduction in inflammatory biomarkers and pro-apoptotic proteins, such as Bax and Bad expressions, after repeated BoNT-A injections. Furthermore, the adhesive protein E-cadherin and junction protein zonula occludens were increased after repeat BoNT-A injections. These immunohistochemistry changes correlated well with the improvement in clinical symptoms [24].

Autologous PRP is growing in the treatment to augmen<sup>t</sup> wound healing, fasten the recovery speed of muscle and joint injuries, and enhance surgical repair recovery [25]. PRP is extremely rich in several essential growth factors and cytokines, which regulate tissue reconstruction, and has been studied extensively among trauma patients and trauma experimental models [26]. Tissue regeneration can be improved by the local application of autologous bone marrow derived progenitor cells and PRP. In addition, PRP eliminates neuropathic pain primarily by platelet- and stem cell-released factors. These factors initiate the complex cascade of wound healing events, starting with the induction of enhanced inflammation and its complete resolution, including tissue remodeling, wound repair, and axon regeneration, resulting in neuropathic pain elimination; some of these same factors also act directly on neurons to promote axon regeneration, thereby eliminating neuropathic pain [27].

PRP injecting into the bladder wall could initiate the wound healing process, induce a new inflammation, complete the wound healing, resolve previous inflammation, and promote the relief of neuropathic pain. The cytokines and growth factors released from PRP could induce a new inflammation, which might override the residual inflammation, and increase tissue regeneration [25]. Our previous clinical trial demonstrated that multiple low-dose PRP injections were effective in patients with IC/BPS. The PRP injections could effectively decrease IC symptoms and VAS bladder pain scores from 3.38 ± 2.89 at baseline to 1.10 ± 1.85 at 3 months after PRP injections [28]. The elevated urinary cytokines and inflammatory proteins could also be reduced after repeated PRP injections [29]. Moreover, the ultrastructural deficits in IC/BPS urothelial could also recovered after intravesical PRP injections [30].

Although this was not a head-to-head randomized trial, the results of this comparative case series study reveal that PRP and BoNT-A are equally effective in reducing IC symptoms. Based on the therapeutic mechanism of PRP and BoNT-A on IC/BPS, either treatment can reduce chronic inflammation and improve urothelial regeneration [12]. Therefore, IC symptoms and urinary inflammatory cytokines can decrease after treatment [29]. However, BoNT-A has a more potent inhibitory effect on the release of noxious inflammatory neuropeptides and could better reduce bladder pain, compared to the effect of the PRP injection. On the other hand, BoNT-A reduces detrusor contractility by inhibiting acetylcholine from efferent nerves; thus, a larger PVR may develop after treatment [17]. Nevertheless, the inhibitory effect of BoNT-A on detrusor contractility in IC/BPS bladders was less than that observed in patients with overactive bladders [31]. Although the increased PVR is not clinically significant, this adverse event might lead to a higher rate of difficult urination and increase the risk of UTI after BoNT-A injections.

Both PRP and BoNT-A treatments are novel and currently off-labeled therapies for IC/BPS refractory to conventional therapy. The long-term efficacy has not been well investigated. Although fewer adverse events were reported following treatment with PRP than with BoNT-A, the therapeutic effect of PRP on chronic inflammation and for promoting recovery of defective urothelium was limited. As shown in this study, the GRA increased with the increasing number of PRP injections, suggesting the therapeutic efficacy needs an adequate PRP dose to achieve a therapeutic level. Therefore, repeated intravesical PRP injections every month are necessary to achieve a therapeutic efficacy similar to that of BoNT-A at six months. However, repeated monthly PRP injection treatment requires frequent anesthesia, which might place a tremendous burden on the patients. Four monthly injections might also increase the complications related with bladder injections, including hematuria and UTI. Although therapy with a single high-dose injection of PRP has been attempted, the efficacy on symptom improvement was inferior to that of four monthly low-dose PRP injections [14]. In addition, the economic burden of frequent admission, anesthesia, and the cost of preparing the PRP might be much higher than that of a single BoNT-A injection within a six-month treatment period. Therefore, although the BoNT-A injection bears a greater potential for detrusor underactivity and UTI, repeated BoNT-A injections every six months have the advantage of greater bladder pain improvement and less of an anesthesia burden, with the same efficacy as that of PRP treatment. Single BoNT-A injection only provides short-term therapeutic effect, whereas repeat injections provide a higher success rate in long-term follow-up. Before recommending either treatment to patients with refractory IC/BPS, patients should be thoroughly informed as to the efficacy and potential adverse events of both treatments to allow for shared decision-making.

Limitations of this study are small case numbers, the lack of a control arm, and the non-randomization of the study design. The treatment option was based on patients' choice after informing advantage and disadvantages of the treatment outcome and potential adverse events. Because all patients were chronic IC/BPS and had received many different conventional therapies, they might have had high expectations to BoNT-A or PRP injection. Further, all patients were not selected with a strict inclusion criterion as in the clinical trials but were diagnosed according to their present symptoms and past history. Therefore, the patients in this retrospective study are highly heterogeneous with varying severity of disease. Finally, although a GRA ≥2 was reported in around 50% of patients treated with BoNT-A or PRP and the ICSI and ICPI also showed improvement, the decrease in bladder pain VAS was limited. The higher VAS in the baseline of the BoNT-A group could have resulted in bias in the VAS reduction between PRP and BoNT-A group. Because this was a retrospective analysis, patients were informed of the advantages and disadvantages of treatment at baseline, and patients were allowed to choose treatment; therefore, more patients who had a higher bladder pain VAS might have chosen the BoNT-A injection. Although a significant reduction in bladder pain VAS was observed in the BoNT-A group, the difference of VAS changes with time between PRP and BoNT-A groups was not significant. Nevertheless, the results of this study may reflect the treatment outcome of BoNT-A and PRP injection in a real-life practice.

In clinical trials for functional urology such as in IC/BPS, OAB, or LUTS, the subjective primary endpoint and objective secondary endpoints are usually not equally improved. Although the primary endpoint can reach a significant result, a placebo effect of up to 30% can usually be found in clinical trial. In this study we found the treatment success was limited in GRA improvement and the other objective variables such as voiding diary parameters, bladder volume increment, and bladder pain reduction were very mild. These results bring a message that, in real-life practice, treatment with a single small dose of BoNT-A or PRP might not be as successful as that in clinical trials. Nevertheless, the safety endpoints also revealed that both treatments were safe and tolerable. With repeated injections of BoNT-A or PRP, the treatment success might be improved. In the future, randomized controlled trials with different doses might be necessary and a search for

prognostic factors could help urologists select IC/BPS patients for a good response to PRP or BoNT-A injections.
