Mehregan and Lee, in 1987,⁶⁶ addressed the subject of "malignant proliferating tricholemmal tumors" in three patients and interpreted of the process thus:

"We have observed three such patients with scalp lesions in which benign and malignant areas coexisted suggesting that malignant transformation may have occurred within the pre-existing proliferating tricholemmal cysts . . . Histologic sections revealed areas of relatively benign proliferation of pilar sheath epithelium and foci of tricholemmal keratinization characteristic of a proliferating tricholemmal cyst. There were also extensive areas of severe cellular dysplasia and surrounding dermal invasion indicating malignant transformation."⁶⁶

In 1995, Sau and coauthors²¹ published findings in regard to 96 specimens of "proliferating epithelial cyst," among which there were 63 cases of proliferating tricholemmal cystic carcinoma and 33 of "proliferating epidermoid cyst." They believed that the 10 proliferating tricholemmal cystic carcinomas showed transition from "benign appearing areas to infiltrating tumors morphologically indistinguishable from carcinoma."

Jaworski raised "the question of how atypical a proliferating tricholemmal cyst can become before a label of squamous-cell carcinoma is indicated?"⁶⁷ She performed studies of ploidy on the neoplasm that had prompted her inquiry and found that many cells contained aneuploid DNA.⁶⁸ Based on this, she concluded that "rather than mimicking a squamous-cell carcinoma, it is in fact a malignant proliferating tricholemmal tumor [tricholemmal carcinoma]." Jaworski, however, thought that the carcinoma did not begin de novo, but as conversion of a benign neoplasm, and she phrased that in these lines: "In my opinion, a diagnosis of carcinomatous transformation of a proliferating tricholemmal tumor can be made in the presence of (a) abnormal mitoses; (b) marked cellular pleomorphism; (c) infiltrating margins; and (d) ploidy."

Sleater and coworkers also performed studies of flow cytometric ploidy in four examples of "proliferating tricholemmal cystic carcinoma" and assessed the proliferation index as measured by Ki-67 immunostaining in conjunction with mitotic counts.⁶⁹ To them, the neoplasms showed "benign" features by conventional microscopy. Two of the four were diploid, and two were nondiploid (aneuploid and tetraploid). The proliferative indices and mitotic counts were higher in the nondiploid than in the diploid lesions. From those observations, the authors offered these opinions:

"Proliferating tricholemmal cystic carcinoma [proliferating tricholemmal cyst] can be confused both grossly and microscopically with squamous-cell carcinoma and malignant proliferating tricholemmal cystic carcinoma. Distinction between these lesions has historically been made on a histologic basis . . . [however,] histologic evaluation of proliferating tricholemmal cystic carcinoma alone may be inadequate in determining biologic behavior, and in the absence of overt malignant areas in proliferating tricholemmal cystic carcinoma with increased numbers of mitoses, DNA ploidy and cell proliferation analysis may be warranted."

DNA ploidy analysis by flow cytometry also was performed by Hashimoto and colleagues⁷⁰ on 10 specimens of tricholemmal carcinoma, 5 specimens of proliferating tricholemmal cystic carcinoma, and 6 tricholemmal cysts. The authors summarized the results of their study as follows:

"Nine out of 10 cases of tricholemmal carcinoma and 2 out of 5 cases of proliferating tricholemmal cyst showed aneuploidy. No aneuploidy was detected in 6 cases of benign tricholemmal cyst . . . proliferating tricholemmal cystic carcinoma has been considered to be essentially a benign tumor even in the presence of a slight degree of nuclear atypia. In the present study, DNA aneuploidy was detected in 2 out of 5 cases of proliferating tricholemmal cystic carcinomas, suggesting its pre-malignant nature."

For us, proliferating tricholemmal cystic carcinoma is not a "pre-malignant" lesion; it is malignant because it has the capability to metastasize and cause death. The course of the eight patients who died of proliferating tricholemmal cystic carcinoma confirms that assessment.

Takata and associates⁷¹ recorded what they thought was a tricholemmal carcinoma that arose in the wall of a proliferating tricholemmal cyst, and they isolated DNA "from microdissected areas of the proliferating tricholemmal cyst and the carcinoma, respectively, and [performed] PCR [polymerase chain reaction]-based microsatellite loss of heterozygosity [LOH] analysis as well as p53 gene sequencing." This is what they found genetically in the proliferating tricholemmal cyst and the carcinoma, and the conclusions they drew from it: "A CGA to TGA nonsense mutation at codon 306 in exon 8 of the p53 gene was found in both samples. LOH analysis showed that the proliferating tricholemmal cyst retained chromosome arm 17p (where the p53 gene resides), whereas the carcinoma was associated with the loss of this allele. All the other loci examined were retained including 3p, 9q, 13q and 17q in both tumor parts."

On the basis of those results, Takata and coworkers inferred that "both of the tumor cell populations from proliferating tricholemmal cyst and tricholemmal carcinoma within the same tumor carried an identical p53 mutation, providing direct evidence of common clonal origin." Their findings seem to support our current view that so-called proliferating tricholemmal cyst and the carcinoma that develops in it are the same genetically, an indication that, conceptually, they represent one and the same pathologic process. Takata and coworkers interpreted preservation of the 17p allele in the region of "proliferating tricholemmal cyst" and loss of that allele in the region of the carcinoma to be an indication that the neoplasm was a proliferating tricholemmal cyst that underwent "malignant transformation," becoming thereby a tricholemmal carcinoma, a supposition they put forth as follows: "The LOH analyses suggested that a critical event responsible for progression from the pre-existing proliferating tricholemmal cyst to the tricholemmal carcinoma may be the complete loss of wild-type p53 because of the allelic loss of chromosome arm 17p . . . a critical event responsible for malignant transformation in this particular case."