This retrospective cohort study is the first report of the long-term clinical course of MC. Over 5 years, 25% of grade I MC cases progressed, and no cases deteriorated in grade after 5-year follow-up. Male sex was confirmed as a predictive factor for progressive MC, and a sex difference in the clinical course was shown. Further, both developing and disappearing cases of MC were detected, which allowed elucidation of their clinical characteristics. These data suggest a minimum time for the development and disappearance of MC, as well as a relevant connection between anthranoid and mirabegron use and the pathogenesis of MC.
To the authors’ best knowledge, only one clinical case report on the development of MC has been published [11]. In an animal experiment, MC developed several days after administration of anthranoids and disappeared soon after drug discontinuation [6]. Willems et al. documented a patient with MC development at 10-month follow-up after colonoscopy [11]. In the current analysis, the mean duration from the last diagnosis without MC to the time of initial diagnosis of MC was 24 months. Although detection of an accurate minimum time to MC development is challenging because of the lack of frequent follow-up, approximately 2 years seems sufficient for a non-MC colon to develop MC. In contrast, cases of MC disappearance were reported to be detected at 1-, 3,- and 10-year follow-up [11, 22, 23]. The present data show 27 months from the last diagnosis of MC to the initial diagnosis of MC absence, which also indicates that at least 2 years is sufficient period for MC to disappear.
As previously known, the sequence of initiation and discontinuation of anthranoids is associated with the development and disappearance of MC, respectively. This relationship was previously documented in several case reports and animal experiments [6]. A case showing MC development and subsequent disappearance in a single patient by initiating and discontinuing anthranoids indicated the strong connection of anthranoids to generation of MC [11]. However, MC cases without anthranoid administration have also been reported [8, 9]. Similarly, MC disappearance with anthranoid continuation has been reported [23]. In fact, the current study revealed a number of MC development and disappearance cases without anthranoid administration and discontinuation, respectively. These findings suggest that various factors, including other agents than anthranoids, can be associated with pathogenesis of MC.
Of interest, this study revealed that several developing cases initiated mirabegron between the last confirmation of a nonpathologic state and the diagnosis of MC. Mirabegron is a selective β3-adrenergic receptor agonist used for patients with overactive bladder [24]. The relationship between MC and mirabegron has not been previously reported. Regarding constipation as adverse effect of mirabegron, which may be also associated with anthranoid use, one study reported that mirabegron users experienced constipation as frequently as placebo patients [25]. Other possible adverse effects seem unrelated to pathogenesis of MC. Additionally, the chemical structure of mirabegron and anthranoids does not share common features [26, 27]. The association between mirabegron and development of MC must be further investigated from wide range of viewpoints.
Only one study to date was available with regard to the clinical course of MC. Gökçe et al reported that only 15.4% of MC patients showed non-MC mucosa after 2-year follow-up observation, despite the fact that most patients discontinued use of laxatives and herbal preparations [12]. The current analysis also confirmed that most patients had stable disease, with 22.8% of grade I MC cases deteriorating in grade and very few cases of grade improvement. The low prevalence of improvement may be in part attributed to the possibility that MC can entirely disappear, rather than only alleviating its grade. In fact, severe-grade MC was reported to convert to non-MC mucosa without gradual improvement of its degree [22]. In the current survey as well, 2 cases demonstrated alteration from severe MC to normal mucosa. Interestingly, the clinical course of MC in humans was proven to differ from that in an animal model, which showed a high rate of immediate development and disappearance after anthranoid initiation and discontinuation, respectively [6]. The authors of the current study hypothesize that this divergence may be ascribed to several factors in humans, such as drugs other than anthranoids, lipofuscin metabolism, and immune status related to macrophage activity. Interventional prospective studies and close inspection of the molecular basis of MC are required to prove these hypotheses.
Of note, a sex difference in the clinical course of MC was first detected in this study, with a greater tendency to progression in male patients. Moreover, the study revealed a higher prevalence of male sex in developing cases and female sex in disappearing cases. Pigment metabolism may explain the sex disparity. Cytochrome P450 is thought to play crucial role in MC pathogenesis, which is proven to be located on colonic mucosa [28] and is associated with accumulation of lipofuscin. The decrease of cytochrome P450 also causes microsomal oxidative damage [29], which results in lipofuscin formation [30]. Actually, expression of the cytochrome P450-related gene was downregulated in MC [31]. The metabolism of lipofuscin pigment influenced by cytochrome P450 can vary between male and female individuals, which results in the different clinical course of MC according to sex. In fact, it was reported that clearance of nifedipine and verapamil, which are metabolized by cytochrome P450 as well, was higher in female than male individuals [32, 33]. Although underlying mechanisms remain unclear, the sex difference in the clinical course of MC should be a consideration in management of patients with MC.
This study has several limitations. First, the retrospective cohort design may lead to information bias and inadequate data quality. Despite careful medical record review, it was not possible to perfectly reflect the medical history because anthranoid use may not have been documented. Some over-the-counter drugs and supplements can contain anthranoids, and these agents are not always reported in the drug history on medical interviews [34]. Although information accuracy regarding comorbidities, social habits, and medication use was influenced by reporting bias, measures were taken to elevate the quality of crucial information such as MC diagnosis by excluding cases with insufficient intubation and preparation for colonoscopy. In addition, the follow-up periods and frequencies were not unified among MC cases. However, because of the low prevalence of MC and no previous reports in this field, the retrospective cohort design was thought to be plausible as the initial step to investigate the clinical course of MC.
Second, with regard to selection bias, data were extracted from a limited region in Japan. However, these data suggest findings that are similar to those of other studies including MC cases [35]; therefore, the results of this survey may not be largely biased. Third, the findings regarding sex difference in the clinical course of MC were confirmed with inadequate statistical power because of small number of cases. Because the sex difference was not primary endpoint of this cohort, the power of the test was not appropriate. However, the results were statistically significant and highly suggestive to enhance an understanding of MC pathogenesis. Based on these results, validation of the relationship between sex and MC status in larger prospective studies is required.
In conclusion, this analysis of the development and disappearance of MC indicated that anthranoid use is related to the incidence of MC. Moreover, other unknown factors, such as agents other than anthranoid, can contribute to pathogenesis of MC. Of the mild MC cases (grade I), 25% experienced progression to a severe state in 5 years. Male patients showed a more detrimental clinical course than female patients in terms of severity. These results provide not only useful guideposts for clinical criteria but also valuable insight into pathogenesis of MC.