We describe a clinicopathologically distinct subtype of cholecystitis, the extensively
calcific version of which has been presented in the clinical literature as "porcelain
gallbladder (PG)." This cholecystitis, which we propose to refer to as hyalinizing
cholecystitis (HC), is characterized by dense, paucicellular hyaline fibrosis transforming
the gallbladder (GB) wall into a relatively thin and uniform band. The process diffusely
effaces most of the normal structures of GB, and some cases show calcifications. To
determine the clinicopathologic associations of HC, we systematically analyzed 4231
cholecystectomies (606 of which had carcinoma) histopathologically, in addition to
a targeted search in our databases. Ninety-six cases of HC were identified (1.6% of
cholecystectomies). Patients with HC were a decade older than ordinary cholecystitis
patients (56 vs. 47; P<0.001), suggesting that HC may be a long-term complication
of chronic injury in some patients. Calcifications of variable amounts and degrees
were identified in two thirds of the cases. In addition, 10 cases showed diffuse marked
calcifications and were considered separately as "complete porcelain" GB. Thirty-eight
HC cases had carcinoma with a calculated frequency of 15% and an odds ratio of cancer
risk of 4.6. Only 42% of the invasive cases were associated with calcifications; none
of the 10 diffusely calcific cases had carcinoma. HC-related carcinomas were challenging
diagnostically. They did not form distinct masses or any significant thickening (mean
thickness, 2.6 vs. 4.0 mm in ordinary adenocarcinomas; P<0.002). Microscopically,
they had widely scattered and bland-appearing glands embedded in the thin band of
hyaline stroma of HC, commonly showing a disappearing lining, leaving behind the granular,
necrotic intraluminal debris (regression) with or without calcifications, which could
be the only sign of cancer in some sections. The morphologic features that allowed
the recognition of these glands as malignant included their longitudinal axis parallel
to the surface, their irregular contours, clear cytoplasm with distinct borders, nuclear
irregularities, and washed-off chromatin. Surface epithelium, if preserved (and it
was not in most cases), typically showed carcinoma in situ of either denuding or micropapillary
types. HC-associated carcinomas, with a median survival of 7 months, appeared to have
a clinical course at least as aggressive as that of regular carcinomas (median survival
12 months; P=0.02). In conclusion, HC is a distinct clinicopathologic entity, which
is often associated with carcinoma, and the carcinomas arising from this group are
often very subtle and prone to misdiagnosis microscopically. As HC is typically devoid
of epithelium, any glandular elements on the wall of HC should be regarded as a suspect
for carcinoma. This study also confirms recent findings in the radiology literature-it
is not the complete (diffusely calcific) PG that is associated with cancer. Instead,
a distinct, histopathologically defined form of cholecystitis, HC with minimal or
no calcifications (incomplete PG), is associated with invasive carcinoma. Thus, imaging
protocols ought to focus on the correlates of HC rather than fixating on calcifications.
Further studies into the pathogenesis of this process and its mechanisms of progression
to carcinoma are warranted.