Non-muscle invasive BC recurs frequently. According to a widely accepted model by Millan-Rodriguez and a more recent model by Sylvester [29, 30] patients included into the present study were at intermediate risk for recurrence requiring adjuvant therapy but at low risk for progression according to respective guidelines . BCG is most commonly used in patients at high risk for progression but is also justified in patients at sufficient risk of recurrence as patients included in the present trial. As the efficacy of BEXIDEM could not be reliably judged, no patients at high risk of progression, e.g. CIS, were included in order to avoid undue risks.
While local immunotherapy with BCG is the most effective agent in preventing recurrence, frequent AEs (e.g., cystitis, mild fever) and less common but severe complications (e.g., fever, granulomatous prostatitis) occur [2–4]. The feasibility to develop novel adjuvant agents in addition to chemotherapy or BCG is twofold; for one it would be ideal to combine the efficacy of BCG with a more advantageous AE profile and secondly therapeutic options are limited following failure of one substance .
As BCG is an immunotherapeutic and BC is viewed as susceptible to respective targeting, it is feasible to pursue further immunotherapeutical approaches. Autologous MAK cell therapy has been reported as a promising treatment modality including BC [24, 17]. While BCG is mediating activation of the immune system via T-cells and its action is not tumour specific, MAK-cells are targeting tumour cells. Their mode of action may be considerably more specific resulting in an improved safety profile [26, 19]. However, safety is a concern in treating patients with MAK, as these central agents of immunoresponse could trigger widespread immunological reactions. Hence safety was chosen as the primary endpoint for the present trial and accordingly the protocol defined adverse events in a strict manner, explaining the rather high overall rates of any AEs in both arms (BEXIDEM: 45%; BCG: 85%). Even taking the present and rather strict approach in mind, BEXIDEM appeared safe.
In comparison to BCG, the incidence of SAEs was significantly lower in the BEXIDEM arm, as 14% versus 26% of BEXIDEM and BCG patients experienced serious AEs, and 1 and 6 patients discontinued the protocol due to BEXIDEM and BCG related SAEs, respectively. The safety profile demonstrated in the BCG group was consistent with what would be expected with this approved product and is described in the product labelling and literature .
While this phase II trial returned the results expected based on the previous study  with respect to its primary objectives, a significantly higher proportion of BEXIDEM versus BCG-treated patients experienced BC recurrence. The agents applied were applied correctly, i.e. BCG in accordance to the current EAU- guidelines and BEXIDEM dosing and regimen in accordance to prior phase I experience [1, 26]. Viability of BEXIDEM was routinely assessed and no breech of protocol was noted in processing, handling or administering the product ruling out reduced activity by mishandling.
The efficacy of BEXIDEM is uncertain and three aspects are noteworthy suggesting a careful interpretation of the results. For one, the rather low numbers and events in the prior phase I trial may not reflect the true, i.e. potentially low efficacy of BEXIDEM. Secondly, this phase II trial was underpowered for the secondary clinical end-point and the sample size calculation was apt to reflect safety only in accordance to the primary endpoint. Thirdly the overall numbers of recurrence events were low which has to be attributed to the risk profile of most patients, which in retrospect was inadequately advantageous for assessing efficacy. Fourthly, even if a certain prophylactic effect was present, it may have been overruled by the close to optimal prophylactic effect of BCG. Unfortunately no information on the frequency of previous tumour occurrences was obtained upon inclusion into this trial.
Thus the efficacy of BEXIDEM remains uncertain. Planning the present trial marker lesion studies were extensively discussed but decided against due to concerns that larger tumor burden is a known challenge for immunotherapeutic approaches, which rely on immune cell numbers to overwhelm tumor cell numbers. Further trials are warranted and should adopt the marker lesion concept by observing rather small tumour. Future trials should furthermore include assessment of efficacy by histopathological analysis of bladder tissue biopsies following the application of BEXIDEM immunological panels.