Not all “mucinous carcinomas” are equal: time to redefine and reinvestigate the biologic significance of mucin types and patterns in the GI tract

Abstract

We read with great interest the study by Solcia et al. [14], published in the December 2004 issue of Virchows Archives. Their results reaffirm that the muconodular (or colloid) type of mucinous carcinoma is associated with a significantly better prognosis than other types of mucinous carcinomas (or the conventional carcinomas of the respective organs), not only in the breast [8] and pancreas [2], but in the stomach as well. As such, this study adds another major blow to a long-held dogma that mucinous carcinomas of any type are particularly aggressive in the gastrointestinal (GI) tract. It has long been known that “pure mucinous” carcinomas of the breast, also referred to as colloid carcinomas [7, 9, 10, 13], are associated with an excellent prognosis, even better than that of some in-situ carcinomas of this organ [7]. It is also noted that, interestingly, even a small component of the usual ductal carcinoma negates this survival benefit [4, 8]. The same seems to apply to the pancreas; pancreatic carcinomas composed almost entirely of the colloid pattern have a very good prognosis [2], whereas mixed-mucinous carcinomas are highly aggressive tumors [12]. Why is pure colloid carcinoma indolent, but not mixedmucinous carcinoma? We believe the answer lies in the pathogenesis of the colloid carcinoma pattern. It has been speculated that in colloid carcinoma, the mucin acts as a containing factor, exerting a function similar to that of the basement membrane in in-situ carcinomas [1, 7]. Indeed, the survival advantage and indolent behavior of colloid carcinomas appear to be most evident when most of the cells are floating within the mucin, without any nonmucinous glands infiltrating into the stroma. What causes the tumor cells to float within the mucin rather than invading the stroma? One characteristic feature of the colloid pattern is that the cells have reversed polarity (as evidenced by both immunohistochemical and ultrastructural studies) [1], in which the secretory pole of the cells is oriented toward the stroma-facing surface of the cells, leading to the accumulation of mucin in the stroma, which in turn acts as a barrier to the spread of the cells. Furthermore, the mucin secreted by these cells appears to be distinctive, with inhibitory properties [1]. MUC2, which is a tumor suppressor, is highly expressed in this type of mucinous carcinoma but not others. Why then is it that even a small component of ordinary carcinoma negates the survival advantage of the colloid pattern? One can speculate that once the cells of a colloid carcinoma acquire the properties that allow them to overcome the mucin and independently invade the stroma, these cells may represent a more aggressive phenotype than even ordinary invasive carcinomas. Does the colloid (muconodular) pattern have the same implication in the GI tract as in the breast and pancreas? The studies by Solcia et al. [14] and others [3] show that it probably does. But in fact, there were already several questions regarding the significance of mucin production in the GI tract. First, appendiceal mucinous neoplasms that have a predominantly colloid pattern are associated with a protracted clinical course even if they give rise to pseudomyxoma peritonei. There are several observations that support this impression [5, 6]. The invasive (disseminated) component of the cases, referred to as “disseminated peritoneal adenomucinosis” by Ronnett et al. [6] or “low-grade mucinous neoplasm” by Misdraji et al. [5], typically shows a predominant colloid pattern and has a much better prognosis than those with less of a colloid pattern. There is yet N. V. Adsay (*) Department of Pathology, The Karmanos Cancer Institute and Harper University Hospital, Wayne State University, Detroit, MI, USA e-mail: adsayv@med.wayne.edu Tel.: +1-313-9932965 Fax: +1-313-7459299