Phakomatous choristoma: a case report and review of the literature

Phakomatous choristoma is a rare congenital lesion of the eyelid that can be clinically and/or histologically mistaken for a cyst, cutaneous adnexal neoplasm, or an ocular adnexal oncocytoma. Only 13 such cases have been previously described, mostly in the English language ophthalmic literature. Zimmerman reported the first case in 1971 and proposed the lesion to be of lenticular anlage origin, a theory that has been widely accepted. We report an additional case occurring in an 8-week-old male infant with a firm nodule of the right lower eyelid that was present since birth. A 15 x 12 x 2 mm circumscribed solid nodule with a homogenously white cut surface was surgically excised. Histologically, this lesion was comprised of cuboidal cells forming cystically dilated and irregularly branched ducts and cords within a densely fibrotic stroma. Also present were eosinophilic basement membranelike material, psammoma body-like calcifications and intraluminal degenerated ghost cells. The immunohistochemical profile of the epithelial cells included strong immunoreactivity for vimentin, focal weak staining for S-100, and negative staining for cytokeratin, epithelial membrane antigen, synaptophysin, and chromogranin. The irregularity of the ducts and cords of epithelial cells within the densely fibrotic stroma resembled an infiltrative neoplasm of cutaneous adnexal or lacrimal duct origin. However, the site of involvement, the peculiar basement membrane material, ghost cells, and immunohistochemical profile were features that helped to distinguish phakomatous choristoma from an infiltrative carcinoma. The correct identification of this lesion is essential to avoid an aggressive surgical excision, thus sparing the eyelid and lacrimal system. The purpose of this article is to bring attention to this rare entity, because it has not been described in either the dermatology or dermatopathology literature and furthermore, is not mentioned in any of the major dermatopathology texts.

Loss of 18q predicts poor survival of patients with squamous cell carcinoma of the head and neck

Tumor suppressor genes play an important role in normal growth regulation. Loss or inactivation of these genes has been implicated in the development of squamous cell cancer and progression of neoplasia. Previous studies in our laboratories have implicated chromosome 18 long-arm deletions as a possible marker of progression in head and neck squamous cell cancer (HNSCC). To test this hypothesis, we evaluated DNA from 67 HNSCC patients for loss of heterozygosity (LOH) at 18q loci, and for association of LOH with survival. Tumor and normal DNA were extracted from fresh tissue and paraffin blocks and were amplified by PCR using primers for three microsatellite repeat polymorphisms in 18q (D18S336, D18S34, and MBP). A total of 27 (40%) patients had LOH of 18q, and these patients had a statistically significantly poorer two-year survival compared to those without 18q LOH (30% vs. 63%; P = 0.008). In a Cox proportional hazards model in which time from diagnosis to death was the outcome variable, patients with 18q LOH had an unadjusted relative risk (RR) of death of 2.46 (P = 0.005). When 18q LOH was placed in a multivariate model controlling for possible confounders in the study, the RR for death was still elevated (RR = 2.10; P = 0.025). The observation of a prognostic association between 18q LOH and poor patient survival suggests that loss of an 18q tumor suppressor gene or genes is important in the progression of HNSCC.

Comparison of two methods of mechanical disaggregation of scirrhous breast adenocarcinomas for DNA flow cytometric analysis of whole cells

Mechanical release of single cell suspensions from solid tumors for DNA analysis by flow cytometry has been shown to be optimal for adenocarcinomas in general. In the breast, many adenocarcinomas are fibrotic or scirrhous, creating difficulty in separating the neoplastic cells from the stroma. The authors have conducted a parallel study in 20 consecutive cases of scirrhous adenocarcinomas of the breast using two mechanical methods of cell disaggregation by: (1) mincing tumor tissue in culture media (MINCE); and (2) scraping the surface of the solid tumor with a scalpel blade and rinsing the blade in culture media (SCRAPE). Both methods were compared with PI staining and quantitation of percent debris and coefficient of variation (CV), as an index of resolution of the G0/G1 tumor peak. Debris was quantitated using Cell-FIT (Becton Dickinson Immunocytometry Systems, San Jose, CA) and Multicycle (Phoenix Flow Systems, San Diego, CA) software programs. With Multicycle, in addition to debris and aggregate determination, an index that included background, aggregates and debris (%BAD) occurring between the boundaries of the tumor G1 and G2 region was evaluated. In the 4 DNA diploid and 16 DNA aneuploid tumors, there was no significant difference in histogram quality measured by CV or amount of debris, aggregates, and %BAD between MINCE and SCRAPE methods. The authors conclude that either method of mechanical disaggregation will produce DNA histograms of comparable quality and degree of resolution. However, the scrape method may be advantageous in the mechanical disaggregation in scirrhous tumors. The scrape method also may be useful in small tumors, where tissue preservation for histology is paramount, and there is insufficient material to separately submit for flow cytometry. Combination of both MINCE and SCRAPE may provide higher cell yields, than using only one of these dissociation techniques. In addition, DNA analysis methods using intact cells obtained with the SCRAPE method result in percent CV values of similar resolution to those reported for methods producing bare nuclear suspensions from fresh tumors.

Rapid (one-shot) staining method for two-color multiparametric DNA flow cytometric analysis of carcinomas using staining for cytokeratin and leukocyte common antigen

The authors present an improved method for rapid two-color staining with direct conjugated antibodies to cytokeratin and CD45 antigen (leukocyte common antigen) for whole-cell, ethanol-fixed preparations of human carcinomas. This method was quality controlled with the T24 human bladder tumor cell line and compared in parallel analysis of 24 fresh human carcinomas with the original two-color method of multiparametric analysis that had been published in 1989. This rapid method was designed to achieve comparable staining intensities of both green (phenotype directed monoclonal antibody label) and red (propidium iodide labeled DNA) fluorescence, identical DNA indexes, comparable coefficients of variation, and subjective visual quality of DNA histograms. This is accomplished in a single (one-shot), abbreviated incubation with monoclonal antibody diluted in propidium iodide-RNase, thereby eliminating two incubations and three wash steps required with the original method. The single rinse is done in the propidium iodide-RNase staining solution with resuspension in fresh staining solution before analysis. With the rapid method, the preparation time is reduced by 130 minutes, resulting in a 60% time savings in batch staining mode compared with the original method. The time reduction and fewer wash steps, which should avoid excessive cell loss and cytoplasmic stripping, may advance the adoption of this two-color method in clinical practice.

Two-color multiparametric method for flow cytometric DNA analysis. Standardization of spectral compensation

Spectral overlap of green fluorescence signals into the red detector (red-minus-green compensation) is one potential source of variation in two-color flow cytometric DNA analysis. Suboptimal compensation in a two-color propidium iodide (PI)/fluorescein isothiocyanate (FITC) system may be observed if compensation is adjusted using an inappropriate standard, or if changes to fluorescence detector high-voltage settings are made without corresponding readjustment of fluorescence compensation. To quantitate the influence of red-minus-green compensation on the quality of DNA histograms, data from 60 dual PI/cytokeratin (CK)-FITC stained carcinomas were acquired in parallel using two compensation standards: a PI/CK-FITC-stained T24 cell line calibrator overstained to achieve a high-intensity green fluorescence standard (HIGFS) with manually set compensation and automated compensation settings derived from commercial phycoerythrin/low intensity FITC beads (LIGF). Both compensation standards gave similar DNA hyperdiploidy results (DNA index, 1.1-2.8). However, LIGF standard yielded two falsely hypodiploid peaks (DNA index, .7 and .9). Eight left-skewed peaks became DNA diploid and symmetric, respectively, with the HIGFS. Use of HIGFS lowered the coefficient of variation percentage in 95% of cases, the greatest differences (maximum, 3.4%; mean, 1.81%) in tumors of highest intensity CK-FITC. The authors concluded that use of cell-based compensation standards (HIGFS) with intense green signals that mimic clinical tumor samples will avoid spurious aneuploidy and maximize resolution of near-diploid abnormalities.

Bone infarct-associated osteosarcoma

Reported is a patient with an osteosarcoma arising in a medullary infarct of the humerus. Infarct-associated sarcoma (IAS) of bone is rare. In a collective review of 50 cases reported in the medical literature, only 37 were fully documented. Including our patient, 26 men and 12 women, ranging in age from 18 to 82 years (mean, 53.4 years) have been reported. Black patients appeared to be disproportionately represented, accounting for 36% of the group. In most patients, there was no known cause for the infarct, whereas in the remainder, the most common underlying condition was a prior dysbaric event or alcoholism. Approximately 75% of the patients had multiple bone infarcts. The femur was involved in 21 patients, the tibia in 14, the humerus in 2, and the radius in 1. Among 40 sarcomas in these patients, 7 (18.4%) were osteosarcomas, and 29 (72.5%) were malignant fibrous histiocytomas. The survival rate in patients with IAS is poor: 5 of the 7 patients with osteosarcoma (71%) and 20 of the 31 other patients (65%) died of tumor. Eight patients are alive and well, all for longer than 5 years.