Deciding When to Omit Repeat Transurethral Resection of Superficial Bladder Cancer: Do Photodynamic Diagnostics help?

ABBREVIATIONS

INTRODUCTION

Urothelial Cancer of the urinary bladder is the fourth most common malignancy in males around the world [1]. Furthermore, its management is one of the most cost-intensive ones of all malignant diseases [2]. The initial transurethral resection of bladder tumor (iTURBT) is paramount in initial staging and therapy of the disease. Further treatment differs greatly depending on invasiveness, so an adequate primary resection is essential for a correct histopathological staging.

An incomplete initial resection can leave residual tumors or carcinoma in situ untreated, which also yields incorrect histopathological staging and possibly results in understaging. In a combined analysis of over 2400 patients in seven EORTC Phase III studies, Brausi et al. showed that, in follow-up cystoscopy after three months, recurrence rate following unifocal tumors is up to 20% and up to 45% in multifocal disease [3].

To prevent this, a repeat transurethral resection (reTURBT) may be necessary if certain parameters are met. Current EAU guidelines recommend reTURBT in case of incomplete initial resection, when there is no muscle in the specimen (with the exception of Ta low-grade (LG)/G1 tumors and primary carcinoma in situ - CIS) and in T1 tumors) [4].

In this setting, the subgroup of Ta high-grade (HG) tumors represents a peculiarity as they do not necessitate a repeat resection according to current EAU guidelines [4]. Nevertheless, in two retrospective trials, tumor persistence was shown to be as high as up to 41,4% in initial Ta HG [5] resp. 54,6% in initial T1 HG [6]. Current AUA Guidelines leave a repeat resection as optional in case of visually complete primary resection and in small tumors [7].

Multifocality has also been found to be a predictor of tumor persistence and -recurrence, whereas no allowance has been made hereto in current guidelines, even though it has been observed as a significant factor in tumor persistence and recurrence [5, 6]. According to a recent systematic review, the probability of upstaging following reTURBT is between 0,4% in Ta tumors and 8% in T1 tumors [8].

To enhance tumor visualization and, in turn, resection quality of the initial resection, photodynamic diagnostics (PDD) is a viable approach. Here, hexaminolevulinate is instilled into the bladder preoperatively, which enriches especially in tissue which has a high rate of cell turnover, such as in tumor cells or cystitis. When exposed to light of a certain wavelength, suspicious tissue is highlighted during cystoscopy. Multiple systematic reviews highlighted a positive correlation of PDD usage with tumor detection [9–12], residual tumor rate and recurrence, as well as longer recurrence-free survival, whereas a more recent work by Neuzillet et al. showed no significant difference in these regards [13].

While several publications reported a decrease in overall tumor recurrence following reTURBT in Ta tumors [14] and a more pronounced decrease in tumor recurrence and progression in T1 tumors [15], a recent systemic review by Cumberbatch et al. only showed a non-significant trend toward lower progession [8]. A significant influence on overall mortality or recurrence risk was not seen.

MATERIALS AND METHODS

We retrospectively evaluated 373 patients in the timeframe of 2007-2017, in whom a repeat transurethral resection has been performed following initial resection of bladder tumor at our department. Parameters were, among others, residual tumor rate, histopathological staging, multifocality, tumor size and completeness of initial resection, as well as PDD use. During the mentioned timeframe, PDD was gradually introduced in our department as standard procedure in initial bladder tumor resections or in cases of late (>5 years) recurrence.

We evaluated in how far objective clinical parameters (T-stage, grading, muscularis, tumor size and multifocality) as well as surgeon reported completeness of resection had an influence on residual tumor rate, as well as the influence of PDD. Special attention was granted to those histopathologic constellations in which a clear recommendation for or against routine repeat resection is lacking in current guidelines.

This study was reviewed by the Upper Austrian Ethics Committee (approval number: J-1–15).

Informed consent has been obtained by all participants. The study was performed in compliance with the Declaration of Helsinki in its most recent form. (Adopted in 1964 by the 18th World Medical Assembly in Helsinki, Finland, and revised by the 64th World Medical Assembly in Fortaleza in 2013).

STATISTICS

Subgroup data sets of the only continuous variable (age) were checked for normal distribution (test of normality: Kolmogorov-Smirnov with Lilliefors significance correction, type I error = 10%). Accordingly, subgroup comparisons were performed either by the t-test (test for variance homogeneity: Levene test, type I error = 5%) for independent samples or by the Mann-Whitney U test. The latter was also used for subgroup comparisons of ordinal variables. Categorical variables were compared by the exact Chi-square test or by the Fisher’s exact test.

Logistic regression analysis (including stepwise forward approach) was used to investigate the influence of the following variables on tumor on reTURBT [no residual tumor vs. residual tumor]; muscularis [present vs. not present in histopathological staging]; sex [male vs. female]; T-stage on initial resection [pTa vs. pT1]; PDD [not performed vs. performed] as well as age [years].

Since the type I error was not adjusted for multiple testing, the results of inferential statistics are descriptive only and the use of the term “significant” in the description of the study results always reflects only a local p < 0 , 05 but no error probability below 5%.

Statistical analysis was performed using the open-source R statistical software package, version 3.4.1 (The R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

DISCUSSION

When PDD was used, there was significantly more often mention of muscularis in the histopathological in comparison to white-light resections. It remains to be debated if PDD use itself leads to a more careful operative approach by the surgeon. The observation that incomplete resections were reported more often when PDD was used, possibly seconds this. Maybe a more thorough or deeper resection was prompted by PDD-positive residuals in the tumor foundation, but this could not be systematically proven.

Special attention should also be granted to the observation that in up to a quarter of PDD resections, additional and histopathologically proven tumors have been resected which were not visible in white light and would therefore be missed. This “PDD-Benefit” did not lead to a significant impact on residual tumor rate in our study. Even more interesting would be the question of tumor recurrence, which was not the scope of our work. It has partially been answered by Grossman et al. in a prospective trial, where the difference on tumor recurrence rate following PDD- and non-PDD resections was shown to differ up to nine months after a follow-up of more than four years [16].

Residual tumor rate on repeat resection was 8,8% (Ta) and 31,6% (T1) in our collective. This projects our data on the lower margin of distributions which have been reported in comparable literature: a meta-review showed 19–56% (Ta) and 15–55% (T1) [17]. This is possibly a consequence of PDD use and its associated effect in our collective, but as we could only observe a non-significant trend toward less residual tumors under PDD, other effects will play a role here as well.

Upstaging was a rare occurrence in our collective. Apparently, it did not make a difference if muscularis was reported in the primary resection or not, albeit reporting of muscularis was relatively scarce in our collective, potentially owing to an undue brevity of pathological reposts. Our results have to be interpreted in light of this potential limitation. Comparable literature mentions upstaging rates of 9,5% and 23,3% in T1 tumors [18] but it has to be said that this was evaluated using routine biopsy of the tumor foundation, which has not been performed routinely in our study. As the quality of the histopathological diagnosis suffers from cauterization artifacts [19], there will necessarily be a difference in reports of muscularis.

The rate of upstaging was comparably low overall in our collective, but there was a further significant reduction when PDD was used, which we interpret to be an effect of improved initial staging quality.

There was more mention of muscularis in the resection of large (>3 cm) tumors and in tumors, which later turned out to be T1 on histopathological report. There is a possibility that the macroscopic aspect of such tumors triggered a more thorough and radical approach by the surgeon. In addition, such cases are more often performed by senior surgeons. An interesting question in this regard would be if a more radical resection correlates with better outcome parameters when also applied to tumors which appear less malign on first sight.

A possible limitation of our study was the comparably low rate of muscularis in specimens. Least mention of muscularis was seen in women and patients with recurring disease. A previous study could show that muscularis is lacking in up to 51% of all resections [19]. This was more often associated with low-grade tumors and was subject to significant inter-observer variability. This dependence on the pathologist’s report is also an important aspect under which our data has to be interpreted.

The subgroup of Ta HG tumors deserves to be highlighted in context of our findings. Current guidelines are unequivocal on this topic. According to current EAU guidelines, tumor size and multifocality are no determinators which routinely trigger a change in therapeutic approach [4] CUA guidelines categorize large multifocal Ta tumors as ‘high-risk’, where a routine reTURBT is recommended [20]. It is partially reflected in the current NCCN guidelines, where size is a triggering factor in papillary tumors [21]. Residual tumor rate of Ta HG cases (19,1%) itself was lower than that of T1 HG, but in the case of large (>3 cm) and multifocal Ta high-grade tumors, it is significantly higher, and we are not the only ones reporting this. In our opinion, tumor size and multifocality are parameters which deserve justification as determinators triggering an obligatory repeat resection in Ta HG disease.

In Ta LG tumors, our observation of low residual tumor rate supports current guideline recommendations, where a systematical second resection can be omitted.

CONCLUSION

When PDD was used in the initial resection, residual tumor rate was lower and tumor staging was shown to be more precise. Furthermore, muscularis was significantly more often present in resections when PDD was used and more instances of pT1 and CIS tumor was detected.

In about a quarter of resections using PDD, histopathologically proven tumor tissue was resected which was not apparent to the surgeon when not using PDD. More in-depth research into the implications of these apparently more complete resections on long-term recurrence rate and survival, as shown by Grosman et al., should be worthwhile.

In contrast to current guideline recommendations, tumor size and multifocality seems to have an effect on residual tumor rate in repeat resections in our collective. In light of our observations, we argue that Ta high-grade tumors which are over 3 cm of size or multifocal should routinely be subject to a repeat transurethral resection. Further long-term research into overall mortality and tumor progression is warranted.