Abstract
Aim: Comparison of the therapeutic effect of treatment of non-muscle invasive bladder carcinoma (NMIBC) after intravesical Interleukin-2 (IL-2) instillations in the presence and absence of a marker tumour. Materials and Methods: Two pilot studies were performed in patients with NMIBC. The first study (10 patients) was performed in Krakow (Poland), the second (26 patients) in Vilnius (Lithuania). In Krakow the tumours were treated with incomplete transurethral resection (TUR) leaving a marker tumour of 0.5-1.0-cm followed by IL-2 instillations (3×106 IU IL-2) on five consecutive days. In Vilnius the tumours were treated with complete TUR, followed by IL-2 instillations (9×106 IU IL-2) on five consecutive days. Results: During 30 months follow-up, the recurrence-free survival was 5/10 (50%) and 6/26 (23%) after incomplete and complete TUR, respectively. So, the ratio of the recurrence-free survival after incomplete/complete TUR of 50/23=2.2. The median of the recurrence-free survival is >20.5 months and 7 months after incomplete and complete TUR, respectively. So, this ratio was >20.5/7= >2.9. The hazard ratio which combines both the chance of the disease recurrence and its timing for both censored and uncensored cases was 0.53, again confirming the better outcome after incomplete TUR. Conclusion: A possible explanation for the better therapeutic effects after incomplete TUR compared with complete TUR is that the marker tumour has tumour-associated antigens (TAA) that could lead to an immune reaction that is stimulated by local application of IL-2. After complete TUR, no TAA are available to initiate and to stimulate an immune reaction; consequently, local IL-2 therapy is less effective after complete TUR. The results of these two pilot studies have led to the recent start of a randomised prospective clinical trial in which therapeutic effects of local IL-2 therapy after complete and incomplete TUR are compared.
Non-muscle invasive bladder carcinoma (NMIBC) should be treated with complete transurethral resection (TUR), in order to obtain optimal staging and to reduce the chance of tumour recurrence. However, recurrence rates at first follow-up cystoscopy at three months varied between 7.4-45.8%, which is a sign that the technique for TUR needs improvement (1). The indication to start adjuvant chemo- or immunotherapy is based on the results of the pathological and clinical findings, because after evaluation of several randomized EORTC-GU-group trials, patients can be classified into different risk group categories for recurrence and progression, if there is no invasion into the muscle layer (2). Adjuvant chemo- and immunotherapy cause side-effects, especially when given for a longer period, and there are no prognostic factors identifying which patients need either of the two treatments and, more importantly, there are no predictive tools to identify those patients in whom disease will progress in spite of or during adjuvant instillation. Improvements in this field of genito-urinary cancer are, thus, still needed.
In order to avoid exposing many patients in a randomized phase III trial to new intravesical agents, the EORTC-GU-group developed the concept of the marker tumour (3-5): during TUR, an untouched marker lesion of 0.5-1.0 cm is deliberately left behind and two months after a course of intravesical instillations, the effect of the investigational agent on the marker tumour is evaluated by cystoscopy under anaesthesia and resection of the tumour or place where the marker lesion was located is carried out; the pathological report has to indicate if there is still tumour or not. If the marker tumour is not present anymore, then it is regarded to be due to the therapeutic activity of the instilled investigational drug. If the marker has not disappeared completely, then the instillation therapy is regarded to be ineffective. The evaluation of response to a marker tumour has been proven to be a safe method to test new drugs (3-5). At the Fourth International Bladder Consensus Conference, it was recommended that marker tumour phase-II studies should be adopted as a standard clinical trial practice for the evaluation of new agents to treat NMIBC before embarking on large phase-III studies (3).
Interleukin-2 (IL-2) has been used for cancer treatment with local (intratumoural or peritumoural) application (6, 7). Therapeutic effects were obtained in tumour models in mice and rabbits (6, 8), and in veterinary (9, 10) and human (11, 12) cancer patients. It was, thus, of interest to treat patients with NMIBC with incomplete TUR, leaving a marker tumour, followed by IL-2 instillation. Applying IL-2 instillation in patients with NMIBC was interesting, because urinary IL-2 levels were increased in patients showing a positive response to immunotherapeutic instillations (13-15).
In Krakow, patients were treated in a pilot study with incomplete TUR followed by IL-2 instillations (11). To test the reproducibility of the data, an identical trial was planned in Vilnius (16, 17). However, the urologists in Vilnius regarded application of the marker tumour concept to be unethical, as they thought that this might harm the patients, due to tumour progression. Hence, they performed complete TUR. This enabled us to compare the effects of intravesical IL-2 therapy in NMIBC in the presence and the absence of a marker tumour.
The therapeutic influence of incomplete and complete TUR was expressed as: the ratio of the % of cases that did not develop a recurrence during the follow-up period after incomplete and complete TUR; the ratio of the median time until tumour recurrence after incomplete and complete TUR; the Kaplan Meier plot combining number and timing of recurrences after incomplete and complete TUR.
Characteristics of patients in Krakow*.
If the ratio after incomplete/complete TUR is 1, then both therapies exert an equal therapeutic effect; if the ratio is 2, then the most effective therapy is twice as effective as the other therapy. Etc.
The Krakow study
Krakow: Materials, Methods and Results
In the first pilot study ten patients who had stage T1, grades 1 or 2 NMIBC were treated with incomplete TUR, leaving a marker tumour. The patients gave their informed consent and the Ethics Committee of the Jagellonian University Medical College in Krakow, Poland, approved the protocol. Patients' characteristics are shown in Table I. At TUR, a tumour of 0.5-1.0 cm was left as a marker tumour. Cancer was histologically-proven using the removed tumour tissue. Two days after TUR, when the diagnosis was histologically-confirmed, 3×106 units IL-2 in 50 ml saline plus 0.1% human serum albumin (HSA) were instilled in the bladder through a catheter. The IL-2 solution remained in the bladder for two hours. IL-2 was instilled on five consecutive days. The effect of IL-2 treatment on the marker tumour was evaluated by cystoscopy and repeat biopsy of the marker site two months after treatment. Any marker tumour tissue present after two months was resected at that time. Thereafter, during follow-up visits, tumour recurrence was assessed with cystoscopy and biopsy. In case of tumour recurrence, the patient was removed from the study. The IL-2 was produced by Chiron/Novartis. Data of this study have been published by Den Otter et al. in 1998 (11). The results are summarized in Table II. Following treatment complete response (CR; no tumour or tumour cells were detectable anymore) was demonstrated in 5/10 (50%) patients during the whole follow-up period of 30-54 months. The median recurrence-free interval was > 20.5 months (>30+11)/2). The findings for patients 1 and 2 in Tables I and II are interesting, since these patients had bladder carcinoma for 7 and 11 years and were treated with 23 and 12 TURs; after incomplete TUR and IL-2 instillations, both patients were tumour-free during the whole follow-up period of 54 months.
Time (months) between treatment and recurrence after incomplete TUR and local IL-2*.
The Vilnius study
Vilnius: Materials, Methods and Results
In the second pilot study, 26 consecutive patients with stage Ta-T1 G1, G2, or G3 NMIBC were treated with complete TUR and intravesical IL-2 instillations. All patients gave their written informed consent. This protocol was approved by the Lithuanian Bioethics Committee and State Medicines Control Agency of Lithuania (trial registration: EudraCT: 2004-002821- 30).
Patients' characteristics are shown in Table III. Two days postoperatively, following pathological confirmation of NMIBC, 9X106 units IL-2 (Chiron/Novartis) in 50 ml saline were instilled in the bladder through a catheter. The IL-2 solution remained in the bladder for two hours. IL-2 was instilled on five consecutive days. The effect of the TUR plus IL-2 therapy was evaluated during the regular follow-up visits (cystoscopy and cytology) until tumour recurrence. If tumour recurrence was demonstrated, the patient was removed from this study. Preliminary data of this study have been published by our group in 2010 and 2011 (16, 17).
The results of the Vilnius study are shown in Table IV. A total of 2/26 of these patients (8 %) did not experience a recurrence within the follow-up period of 86 months. The median recurrence free interval of these 26 patients was 7 months.
Comparing both studies. The Kaplan-Meier curve of the Proportion Recurrence-free Patients versus Months from TUR (Figure 1), clearly shows that the therapeutic effects after incomplete TUR are much better than after complete TUR. Table V summarizes the statistical data comparing both studies. The percentage of patients with recurrence-free survival during the follow-up period after incomplete TUR was 2.2-times higher (50% versus 23%). The median of the length of the recurrence-free survival time after incomplete TUR was >2.9-fold that after complete TUR (>20.5 versus 7 months). The hazard ratio combining risk and event in censored and uncensored data, was 0.53 (95% confidence interval 0.23-1.19). Thus, all data indicated much better therapeutic results with regard to recurrences after incomplete TUR and complete TUR.
Patients' characteristics in Vilnius*.
The patients characteristics differ between both studies. Table VI compares the major differences between the patient populations of both studies. The preoperative number of tumours was a little bit lower in Krakow compared to Vilnius, 3.1 versus 3.5, respectively. The recurrence rate per year was 2.2 versus 1.5, respectively. The T-stage in the two trials differ: all cases in the Krakow study and 8% in the Vilnius study were T1. In Krakow 8/10 of the tumours were G1; in Vilnius 6/26 were G1 14/26 were G2 and 6/26 were G3. The patients in Krakow were approximately 10 years younger, e.g. 56 compared to 66 years old in Vilnius. The percentage of males was lower in Krakow, 60% compared to 80% in Vilnius.
Time (months) between treatment and recurrence after complete TUR and local IL-2*.
The underlying question is how do all these differences influence the therapeutic outcome of both studies. Sylvester et al. (2) have provided weight estimates for various factors on the risk of recurrences after surgery for bladder carcinoma. The measured parameter values (columns 2 and 3 of Table VI) are weighed using the Sylvester weight estimates (column 4 and 5). Column 4 and 5 show the average weight of the parameter values of column 2 and 3 on recurrence risk, calculated from for each patient based on individual characteristics. Column 6 shows the predicted RFS after incomplete TUR based on the comparison of data in column 4 and 5; also other data in the literature are taken in account. A difference of more than 0.5 is considered better or worse, depending of its direction; differences between 0.1 and 0.5 are considered ‘slightly’; differences less than 0.1 are considered minimal. So, the result is an educated guess of the influence of the differences in parameter values in both trials on the RFS. The predicted effect of T- and G-stage seem to balance each other between both studies. Sex and marker tumour are not reported to have significant effects. The small difference in preoperative tumours and the difference in age have only small opposite effects. The recurrence rate/year would indicate a worse prognosis for the Polish study. Taken together, the influence of the first seven parameters on the recurrence-free survival seems worse for the Polish study compared to the Lithuanian study, the weights for recurrence are 8.4 and 7.7, respectively. This means that after incomplete TUR more recurrences can be expected; weight 8.4. This cannot explain the obviously opposite effect we found, as shown in Table V.
Discussion
IL-2 is thought to stimulate immune reactivity against tumour associated antigens (TAA). In other words, TAA should initiate an immune reaction, and IL-2 can stimulate this reaction. In line with this view, a recent review showed that local application of immunotherapeutics at the tumour site is much more effective than systemic application (7). However, after a well-performed TUR for NMIBC, no TAAs are left and consequently IL-2 cannot stimulate the immune reaction induced by TAA. This view is supported by the findings for patients 1 and 2 in Krakow: these patients were treated for recurrent bladder carcinoma with complete TUR for many years, but finally they were ‘cured’ after treatment with the antigenic marker tumour in combination with IL-2. The presence or absence of TAA could explain the therapeutic differences observed in these two pilot studies: taken together these data suggest that the presence of a marker tumour leads to improved therapeutic effects.
There are some differences between these two studies that have to be highlighted: Firstly, the difference in results may, partly, be due to differences in the technique of the TUR performed by individual surgeons (2). Secondly, the patients' characteristics differed between both groups (Table VI). These differences in patient characteristics seem to predict more recurrences after incomplete TUR than after complete TUR. However after therapy the opposite effect is obtained.
Unfortunately, no pathology review was part of these pilot studies. Thirdly, in Krakow IL-2 was dissolved in a solution of saline with 0.1% HSA, whereas HSA was not used in Vilnius. HSA was used in Krakow because IL-2 binds to HSA; this prevents IL-2 sticking to the wall of syringes, catheters etc. In Vilnius HSA was omitted; the IL-2 was dissolved in a relatively small volume of saline; this prevents precipitation of IL-2. Fourthly, in Krakow, 3×106 IU IL-2 were instilled daily, whereas in Vilnius 9×106 IU were instilled daily. Urologists in Vilnius preferred this higher dose as they expected better therapeutic results. However, a literature review showed that there is no linear relation between IL-2 dose and therapeutic effect (6). Moreover, a negative dose-response effect would be quite unexpected.
In both pilot studies, there were no side-effects observed due to intravesical IL-2 instillations.
Kaplan-Meier curve of the proportion of recurrence-free patients versus time after incomplete and complete TUR. The fraction of patients that were recurrence-free after local IL-2 and incomplete TUR (line; Krakow, Poland) and complete TUR (stripes; Vilnius, Lithuania). The X-axis indicates months after treatment with (in)complete TUR and local IL-2; the Y-axis indicates the fraction of patients having recurrence free survival. Circles are used to indicate censored data. Table V provides statistical analyses. Table VI summarizes differences in the patients populations of both studies and the predicted effect of several parameters values on the predicted RFS after incomplete versus complete TUR.
Therapeutic outcome of after incomplete and complete TUR and local IL-2 application.
Beside the study in Krakow, seven other studies (11, 18-25) suggest that IL-2 instillation in the bladder may exert therapeutic effects. These data are summarized in Table VII. This table shows that direct contact between bladder carcinoma and IL-2 (column 2) leads to an impressive number of complete remissions (CRs) (column 3). In seven studies (that is all mentioned studies except that of Grasso et al, (24)) comprising 66 patients, there were 27 CRs (41%) and 8 partial remissions (PRs) (12%). That is very high, but similar to the proportion of patients without a recurrence during the follow-up that we found in Krakow. Although the number of patients is small, it is clear from the data in Table VII that contact between IL-2 and tumour cells can induce a CR. In the study of Grasso et al. (24) no CR was observed in 27 patients, but the percentage of historical recurrences per year was 95% and this was reduced to 33% in this study. Taken together, each of these eight studies indicates a clear therapeutic effect of local IL-2 in direct contact with bladder carcinoma. Only, the study of IL-2 installation after complete TUR of bladder carcinoma failed to show therapeutic efficacy.
We are aware that IL-2 is regarded as a stimulator of an effective cytotoxic reaction, as well as stimulating T-reg cell expansion leading to immune suppression (7, 25, 26). These seemingly contradictory effects of IL-2 may be due to differences in the doses of IL-2 that are used, the site of IL-2 application (7, 27), or the stage of the immune reaction, such as stimulation at an early or a late stage in the immune reaction. We did not observe any example of IL-2-induced stimulation of tumour growth.
Predicted outcome based on pre-treatment patient values.
Therapeutic effects of direct contact of IL-2 and carcinoma of the bladder.
The Kuhnian paradigm of immunotherapy of cancer is that an antigen, in this case TAA, induces a local immune reaction that leads to a cure (7). If a tumour is removed by a complete TUR, then TAA are not present to induce an immune reaction. The results described in this paper are exactly what the immunotherapeutic paradigm predicts.
It is clear that the therapeutic differences described in this article warrant a new study with the presence/absence of the marker tumour followed by IL-2 instillation. Currently we are performing such a randomized trial in the Netherlands (EudraCT number 2010-020397-42). If the data reported here are confirmed in this new ongoing trial, then the treatment of NMIBC with immunotherapy might be improved.
Acknowledgements
JJLJ was supported by Stichting Nationaal Fonds tegen Kanker (SNFK) and we received a grant from Hoogland Medical Heesch.
Footnotes
-
This article is freely accessible online.
-
Conflicts of Interest
None.
- Received January 14, 2013.
- Revision received March 26, 2013.
- Accepted March 27, 2013.
- Copyright© 2013 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved