Although IM are reported to occur in 0.1 to 0.9 % of the bladder population, the true incidence is difficult to determine due to the vague, nonspecific symptoms often associated with this condition [3, 4]. In this study, we studied 89 patients from 27,424 bladder tumor patients with long-term follow-up (mean 105 months) after the initial diagnosis of IM. To the best our knowledge, this represents the largest series of IM cases described in a single institution. In the current series most patients reported symptoms of haematuria, irritative voiding symptoms and mucosuria in the fifth to sixth decade of life, which is comparable to that in prior studies [10, 11]. However, the frequent finding of gross or microscopic haematuria in 82 % of our patients is much higher than the rates previously reported in patients with IM .
Although the etiology of IM is not well characterized, many cases are thought to represent an acquired condition that occur secondary to long-standing injurious stimulus [1, 5]. More than half of the patients in our study had associated conditions predisposing to the stasis of urine or exposure to chronic irritants. Persistent inflammation of the bladder has been proposed to lead to adaptation and subsequently transformation of the urothelium, resulting in the development of glandular metaplasia . Of the patients in our series 33 % had a history of chronic UTI, suggesting a possible association between this finding and an IM in a large proportion.
Grossly, IM can be found anywhere in the bladder but usually arises from the trigone . In our cases, 75 % of the IM developed on or near the triangle region, which is almost consistent with the reports of previous investigators . It is important to note that over one-fourth of the cases are multifocal, highlighting the entire urothelium needs to be evaluated if a lesion is found.
In theory, the urothelium may return to a normal pattern of differentiation, if the stimulus that caused metaplasia is removed or ceased. However, patients often require surgical intervention due to ongoing clinical symptoms or concern about dysplastic changes. Because IM has no tendency toward infiltration, and under most circumstances is superficial , nearly all cases in our series were treated by TUR. However, difficulty in differentiating extensive IM from adenocarcinoma and occasional coexistence of IM and carcinoma make cystectomy unavoidable in some cases. In our cases, seven patients underwent partial cystectomy to remove the lesions.
The significance of IM of the bladder has been the subject of numerous debates [1, 5, 13–15]. On the one hand, numerous of molecular changes have been demonstrated to be present in the metaplastic changes which indicate that IM may constitute a putative precursor lesion of adenocarcinoma. By applying the fluorescent in situ hybridization, Morton et al. provided evidence of significant telomere shortening in IM compared with telomere length in adjacent normal urothelial cells. In addition, chromosomal abnormalities associated with urothelial carcinoma were shown to be present in a subset of IM. The authors concluded that these findings suggest a premalignant potential of IM . Bryan et al. investigated the role of tumor necrosis factor-alpha and the adherens junction component beta-catenin in IM, and found nuclear localization of the latter in IM. Since nuclear localization of beta-catenin is also seen in the Barrett metaplasia of the esophagus, which is a preneoplastic condition , the authors concluded that bladder IM may have the same potential to progress to malignancy .
On the other hand, to date only six cases of carcinoma arising in IM have been reported [4, 7–9, 11, 19]. It is noteworthy that the malignancies in the IM appear not to be unique, but one of the standard bladder cancers, ie adenocarcinoma or transitional cell carcinoma, with the majority of tumors reported being adenocarcinomas. However, many documented cases were derived from case reports and have been criticized to have a history of a prior or concurrent carcinoma. Furthermore, in a clinical investigation of 53 patients with IM, Corica et al. found that none of the patients developed bladder carcinoma after a median follow-up of 13 years. The authors concluded that IM is not a strong risk factor for bladder carcinoma . Smith et al. studied 12 cases of IM and found that only one patient developed an urothelial carcinoma 3 months after resection of an IM; they concluded that IM does not seem to increase the future risk of bladder malignancy and surveillance cystoscopy is not recommended in such patients .
In our study we identified carcinoma arising within 1 % of IM and in this case in our series it represented adenocarcinoma. The result is in contrast to the findings in previous studies [10, 11]. Several explanations may account for this discrepancy. First, our cohort includes the largest number of IM patients to date with long-term follow-up. This obviously differs from previous small study with relatively short follow-up, which may not have adequate power to capture the incidence of events . Second, in Corica et al. study, most of recruited patients were children rather than adults in our series . Metaplastic cells in children with extrophy might occur at early stage in the replicative cycle and require decades to progress to carcinoma . Third, all patients with extrophy have been underwent surgery to reconstruct the bladder. It is reasonable to consider that repair of the extrophy likely removed the stimulus and delay or reverse the process of carcinogenesis .
Although we could not rule out the possibility that development within 6 months of a carcinoma in our study was because an overlook at the time of diagnosis, this single case highlights several important clinical features. Extensive lesion was present in the initial cystoscopic findings but, more importantly, the histologic examination indicated severe dysplastic changes. Recently, Gordetsky et al. reported the histologic details and follow-up of patients with dysplasia. With a total of 20 cases included, the authors concluded that dysplastic changes are significantly associated with concurrent adenocarcinoma and suggested that patients with IM exhibiting this feature should undergo close follow-up .
This might raise the question of whether the presence of dysplasia is an indicator for the development of subsequent adenocarcinoma. The metaplasia-dysplasia-adenocarcinoma sequence is a widely recognized event in several epithelial tissues including esophagus and gastric [17, 21]. In supporting this sequence in bladder IM, Srivastava et al. demonstrated positive expression and allelic imbalance of TP53 and loss of heterozygosity for D2S123 in glandular dysplastic foci but not in IM. The authors therefore suggested that the presence of dysplasia might be the early changes in the stepwise progression to adenocarcinoma . Unfortunately, our data could not provide sufficient evidence to answer this question. More studies with patients with dysplasia are needed.
Implication for malignant potential of the IM might also be obtained in the number of recurrences, which has been estimated to be around 6 % in previous study . Of 78 patients for whom follow-up was presented in our study, 4 (5.0 %) were reported to have had recurrent lesions after a follow-up of 9 to 24 months. Although the rates are lower than in transitional cell carcinoma, in three cases the IM recurred two times. In fact, when cases of recurrence are examined IM develops near the origin site in many at the time of recurrence, and cases with a history of chronic UTI before the development of IM are seen frequently.
Our study has some methodologic factors that might affect the accuracy of our estimates. First, it was a retrospective study of a limited number of patients. The patients were not followed up uniformly at regular intervals, and the progression and recurrence may be higher if patients were followed more closely or regularly. Second, immunohistochemical staining was not routinely undertaken during the study period. In contrast to other glandular lesions, IM commonly expresses staining for CDX2 and CK20, which is often seen in colonic mucosa . Third, All lesions were surgical removed before the start of follow-up. Hence, there was no residual lesion in the bladder and all patients could be regarded to have normal urothelium at the beginning of follow-up. Therefore, the nature history of IM remains to be elucidated.