Search

Identification of loss of heterozygosity (LOH) at specific genetic loci in cancer cells suggests the presence of a tumor suppressor gene within the deleted region. A basal cell carcinoma (BCC) susceptibility gene, human homolog of drosophila patched (PTC), has been recently cloned and localized on chromosome 9q22.3. Mutation and deletion of this region has been reported in BCCs using frozen tumor tissue. The objective of this study was to test whether LOH of human PTC on chromosome 9q22 could be detected in archival sporadic BCCs. We studied 20 randomly selected sporadic BCCs by microdissection and polymerase chain reaction using paraffin-embedded, formalin-fixed material on glass slides. In all cases, analysis was performed with the polymorphic markers D9S53, D9S15, D9S287, and D9S303. The LOH frequencies were 30%, 42%, 56%, and 75% with D9S15, D9S287, D9S53, and D9S303, respectively. LOH at 9q22 was identified in 12 of 20 cases (60%) with at least one marker. Seven cases showed LOH with two markers, two cases with three markers, and one case showed LOH with all four markers. The results indicate that BCC LOH can be frequently identified in paraffin-embedded BCC after routine processing.

The balanced translocation t(11;22)(q24;q12) is specific for the Ewing's sarcoma/peripheral primitive neuroectodermal tumors (ES/PNETs) and results in the EWS/FLI-1 fusion transcript, which can be detected by reverse transcription polymerase chain reaction (RT-PCR). Recent studies also have used fluorescence in situ hybridization (FISH) to show the translocation; however, most of these have been performed on cell lines or touch preparations and short-term cultures of tumors. Moreover, the existing probes generally have shown only the break in the specific chromosomes rather than the translocation itself. We describe our findings with a new set of probes that localize to 22q12 (EWS) and 11q24 (FLI-1) and directly show the translocation as juxtaposed red-green signals on der(22) in nuclei extracted from formalin-fixed, paraffm-embedded tissues. After establishing the specificity of the probes (on metaphase spreads and interphase nuclei in two translocation-positive cell lines and normal peripheral blood lymphocytes), we evaluated 11 ES/PNETs and 10 other tumors (four alveolar rhabdomyosarcomas, three neuroblastomas, two lymphomas, one extramedullary myeloid tumor) using a two-color FISH assay. All 11 ES/PNETs showed fusion signals in 20% to 80% of evaluable nuclei. In two lymphoma cases, random overlapping signals were present in 2% and 4% of nuclei, whereas the remaining eight tumors were negative. The presence of t(11;22) was confirmed by RT-PCR in 10 of 11 ES/PNETs. We conclude that FISH analysis with this newly designed probe pair is a specific and sensitive method of detecting t(11;22) on routinely processed tissue and can be useful in the differential diagnosis of ES/PNETs from other small round blue cell tumors when only fixed tissue is available.

Papillary carcinoma of the breast is a variant of predominantly intraductal carcinoma characterized by a papillary growth pattern with fibrovascular support. Loss of heterozygosity (LOH) was evaluated at multiple chromosomal loci (including loci reported to show frequent genetic alterations in breast cancer) to determine the frequency of genetic mutations in these tumors and their precursors. Thirty-three papillary lesions of the breast (6 papillary carcinomas, 12 carcinomas arising in a papilloma, and 15 intraductal papillomas with florid epithelial hyperplasia) were retrieved from the files of the Armed Forces Institute of Pathology (AFIP). Tumor cells and normal tissue were microdissected in each case and screened for LOH at INT-2 and p53 as well as several loci on chromosome 16p13 in the TSC2/PKD1 gene region (D16S423, D16S663, D16S665). LOH on chromosome 16p13 was present in 10 of 16 (63%) informative cases of either papillary carcinoma or carcinoma arising in a papilloma as well as in 6 of 10 (60%) informative cases of intraductal papilloma with florid epithelial hyperplasia (IDH). One case showed simultaneous LOH in both the florid IDH and carcinoma components of a papilloma. LOH was not observed at either INT-2 or p53 in any of the papillary carcinomas or papillomas with florid IDH. In conclusion, a high frequency of LOH at chromosome 16p13 (the TSC2/PKD1 gene region) is in both papillary carcinomas of the breast as well as in papillomas with florid IDH, including a case with LOH present simultaneously in both components. These findings suggest that chromosome 16p contains a tumor suppressor gene that frequently is mutated early in papillary neoplasia.

Distinction between benign adrenal cortical proliferative lesions and adrenal cortical carcinoma has been approached by a combination of histological, immunohistochemical, and macroscopical parameters. Modern imaging studies allow detection of small adrenal cortical lesions that may be incorrectly diagnosed. Differentiation between benign and malignant tumors of the adrenal cortex was attempted by microdissection of nine cases of adrenal cortical hyperplasia, 10 cortical adenomas, and 18 adrenal cortical carcinomas with subsequent polymerase chain reaction (PCR) amplification for loss of heterozygosity (LOH) of five microsatellites of putative tumor suppressor gene loci: p53 gene (17p), the neuroblastoma candidate gene (1p), the p16 gene (9p), the von Hippel Lindau gene (3p), and the retinoblastoma gene (13q). None of the hyperplastic lesions or cortical adenomas showed LOH of any of the gene markers used. Conversely, genetic changes were observed in 61% (11 of 18) of the cases of carcinoma. Forty-four percent of the lesions showed LOH for p53 (7 of 16). LOH of 1p, 3p, and 9p were seen in 22%, 22%, and 26%, respectively. LOH of the retinoblastoma gene was seen in 80% or four of five of the informative cases studied. We conclude that LOH studies may be used to distinguish malignant from nonmalignant adrenal cortical proliferations. Relative infrequency of LOH in 3p may furthermore help to differentiate adrenal lesions from clear cell carcinomas of the kidney.

A critical area of chromosomal loss at region p16(9p21-22) and p53(17p13) has been implicated in the genesis of malignant melanoma. It is still unclear whether the genetic alterations can be detected in dysplastic nevus, a premalignant lesion of malignant melanoma. We have searched the frequency of p16 and p53 deletion in nine dysplastic nevi and 13 benign intradermal nevi with five microsatellite markers. Hemizygous deletion was detected in seven of nine (78%) dysplastic nevi at one or more loci for p16 and three of seven (43%) for p53, respectively. No loss of heterozygosity (LOH) was detected in any of the benign intradermal nevi. All three dysplastic nevi with LOH for p53 also showed LOH at p16. However, not all dysplastic nevi showing p16 deletion showed p53 gene deletion. Therefore, these data suggest that deletion of p16 may play an important role in the development of dysplastic nevus as an early event and that the changes may represent an early event in the development of malignant melanoma.

Central nervous system hemangioblastoma is a neoplasm with characteristic and well-described histopathological features, including proliferation of vascular and stromal cells. yet, the histogenesis of the stromal cell component and its neoplastic capacity as compared with the vascular component are still controversial. Stromal cells were selectively procured from formalin-fixed, paraffin-embedded archival tissue from a von Hippel-Lindau (VHL) disease patient with a cerebellar hemangioblastoma and studied for loss of heterozygosity (LOH) of the VHL gene locus and associated microsatellite regions. The stromal cells consistently showed LOH. Analysis of mixed stromal anti vascular areas of this tumor and four other hemangioblastomas of VHL patients showed that loss of heterozygosity was partially obscured. These preliminary results suggest that the stromal component of hemangioblastomas contains genetic alterations consistent with a neoplastic nature. Additional samples of pure stromal cells need to be analyzed to establish the prevalence of VHL gene deletion in stromal cells of capillary hemangioblastoma and, hence, its pathogenetic significance.

Pheochromocytomas and extra-adrenal paragangliomas are tumors of the paraganglia with similar histological characteristics. We examined 12 sporadic pheochromocytomas and 5 extra-adrenal paragangliomas for loss of heterozygosity (LOH) in chromosomes 1p and 3p using a microdissection technique. Chromosomes lp34-36, 3p21 and the von Hippel-Lindau (VHL) gene locus (3p25) were analyzed. LOH of a selected region on chromosome lp was detected in 5 of 11 (45%) informative pheochromocytoma cases and in 0 of 5 (0%) informative extra-adrenal paraganglioma cases. LOH of the chromosome 3p25 VHL gene locus was detected in 5 of 9 (45%) informative pheochromocytoma cases and in 0 of 3 (0%) informative extra-adrenal paraganglioma cases. LOH of 3p21 was detected in 2 of 4 (50%) informative extra-adrenal paraganglioma cases. The allelic deletions in chromosomes 1p and 3p appear to be separate events. In conclusion, significant deletions were found at lp34-36 and 3p25 in sporadic pheochromocytomas but not in extra-adrenal paragangliomas. These findings suggest (1) that multiple genetic factors may be involved in pheochromocytoma tumorigenesis, and (2) extra-adrenal paragangliomas may have a different genetic mechanism of tumorigenesis compared with pheochromocytomas.

Disseminated disease is very important in the clinical assessment of pediatric sarcomas. Several reports suggest that reverse transcriptase polymerase chain reaction (RT-PCR) holds great promise in the early staging of cancer patients in general. However, the complexities of these protocols hamper adequate standardization, and their application as routine diagnostic tools has been difficult. The aim of this study is to assess the actual minimal number of tumor cells that may be detected by RT-PCR in a blood sample. Specific tumor cell dilutions from a Ewing's sarcoma cell line reconstituted in peripheral blood from healthy individuals were "ficolled" and submitted to RNA extraction for cDNA preparation and PCR amplification of the t(11–22) (q24;q12) fusion transcript. After PCR amplification, we were able to detect the EWS/FLI-1 chimeric gene product at a dilution of 10 tumor cells per 1 or 2 mL of blood. Our simple method supports a role for routine clinical use of RT-PCR in the detection of circulating Ewing's sarcoma cells.

Carcinoma is an important complication of ulcerative colitis (UC) and develops from dysplastic precursor lesions. Genetic changes involved in the malignant transformation have not been fully characterized. We studied 19 cases of UC with high-grade dysplasia (HGD) and eight samples of associated carcinoma (CA). Microdissection of normal epithelium, epithelium at the site of chronic inflammation, HGD, and CA was performed. Polymerase chain reaction (PCR) amplification for loss of heterozygosity (LOH) of the following polymorphic microsatellites of putative tumor suppressor gene loci was done: APC (5q), DCC (18q), p16 (9p), p53 (17p), and 8p12. To compare genetic alterations, 22 typical adenomas of the colon were studied with the markers for APC and pl6 gene loci. The results indicated that LOH of p16 and p53 were present in nondysplastic epithelium, HGD, and CA. However, the LOH in nondysplastic epithelium was detected in some associated HGD, but not all. Whereas LOH of p16 was present in 7 of 14 cases of HGD (50%), it was noted in only 1 of 22 adenomas (5.0%). LOH in the APC and DCC gene loci in UC was noted in HGD with associated CA, but LOH of APC was not present either in cases of nondysplastic epithelium or in HGD alone. Conversely, LOH in APC was present in 4 of 19 colonic adenomas. We conclude that LOH of p53 and p16 in nondysplastic epithelium may be associated with chronic reparative processes. These changes may lead to susceptibility to further genetic damage involving the APC and DCC gene loci in the development of dysplasia and progression of CA in UC. The low frequency of LOH in the p16 gene (9p) in adenomas compared with dysplasia in UC combined with infrequent LOH in APC gene loci in cases of pure dysplasia in UC may support this combination of markers as a clinical test for the differentiation of polypoid dysplasia from adenomas in UC.

Demonstration of identical allelic loss on chromosome 11q13 in synchronous in situ (DCIS) and invasive ductal (IDC) breast carcinoma has provided molecular evidence of the progression of DCIS to IDC. We investigated loss of heterozygosity (LOH) at chromosome llql3 in the spectrum of "marker/premalignant" and "malignant" lobular lesions of the breast, including atypical lobular hyperplasia (ALH), lobular carcinoma in situ (LCIS), and infiltrating lobular carcinoma (ILC). Thirty-eight cases with various combinations of ALH, LCIS, and ILC were studied. Synchronous ductal lesions were present in 9 of 38 cases. Areas of interest were specifically isolated by tissue microdissection. The extracted DNA was amplified by polymerase chain reaction (PCR) and analyzed with two polymorphic markers for chromosome 11q13 (INT2 and PYGM). LOH at 11q13 was identified in ILC and LCIS in approximately one third of informative cases. LCIS in association with ILC showed a loss in 50% of cases, whereas pure LCIS in the absence of ILC had a much lower frequency of LOH, which was comparable to that of pure ALH. These results suggest that LOH on chromosome llql3 may play an important role in development of ILC, similar to that of IDC from DCIS/ADH. Additionally, frequent LOH in ILC and LCIS associated with ILC and a significantly lower and comparable frequency of LOH in LCIS without ILC and ALH implies that genetic alteration(s) on chromosome l1ql3 may be important in the transition of LCIS to ILC. LOH was detected in three of nine synchronous ductal lesions (one IDC and two DCIS), confirming our earlier findings and indicating that lobular and ductal neoplasia in the breast show some similar genetic changes. We hypothesize that LOH may help in separating morphologically similar yet genetically different subgroups of ALH and LCIS into one group with genetic changes and an increased potential to progress to invasive cancer and another group, the "marker" lesions of LCIS/ALH, that remain stable or possibly regress.

The progression of human malignancies is thought to involve the inactivation or loss of tumor suppressor genes. Previous studies have suggested that inactivation of tumor suppressor genes on chromosomes 5q, 17p, 18q, and 8p play a role in the development of colorectal carcinoma. However, chromosome 3p at the von Hippel-Lindau disease (VHL) gene locus (3p25-26) has not been previously implicated in the development or progression of sporadic colorectal carcinoma. The authors have analyzed VHL gene alterations on chromosome 3p in sporadic human colon carcinomas and adenomas using modified microdissection techniques. These techniques allow for procurement and analysis of selected subpopulations of cells from both paraffin-embedded and frozen human tumor specimens. VHL disease gene deletion was detected by polymerase chain reaction (PCR) and single-strand conformation polymorphism (SSCP) analysis in microdissected colon carcinoma specimens. Allelic loss of VHL gene was detected in 7 of 11 (64%) informative patients who underwent colectomy for primary sporadic colon carcinoma. However, no allelic loss of VHL gene was shown in colonic adenomas of eight informative patients. These results indicate that VHL disease gene deletion frequently occurs in sporadic colon carcinoma. Because this deletion was not present in adenomas, VHL gene may play a role in colonic carcinogenesis and represent a relatively late event in colonic neoplasia progression. Additionally, microdissection of tissue sections may be especially useful in detecting allelic loss in PCR-based studies of infiltrating tumors, particularly when the tumor cells represent a relatively small percentage of the total cell population.

The coexistence of mucinous ovarian and appendiceal tumors in association with pseudomyxoma peritonei (PP) is well established. However, it has not been determined whether they represent independent or metastatic neoplasms. The authors analyzed microsatellites on chromosome 17q 21.3-22 (nm23), 3p 25-26 (von Hippel Lindau disease [VHL] gene), and 5q 21-22 (D5S346 locus) in 12 synchronous ovarian and appendiceal mucinous lesions. Loss of heterozygosity (LOH) at the nm23 locus has been shown previously in ovarian carcinomas, and genetic alterations at both the 3p and 5q loci have been reported in colorectal carcinomas. The ovarian lesions consisted of nine mucinous tumors of low malignant potential and three invasive adenocarcinomas, and the appendiceal lesions consisted of eight carcinomas without invasion, two invasive carcinomas, and two mucosal hyperplasias. DNA was extracted from microdissected cells obtained from formalin-fixed, paraffin-embedded tissue sections and amplified by polymerase chain reaction. In three specimens, genetic alterations occurred at 17q 21.3-22 in only the ovarian tumors. One of these cases showed LOH on chromosome 5q 21-22 in only the appendiceal tumor. In three other specimens, LOH at the same locus was found in both tumors. Six specimens did not show LOH at any locus. These results suggest that a subset of synchronous mucinos ovarian and appendiceal lesions showing different LOH patterns in both sites most likely represent patients with two separate primary lesions. Another group of specimens with the same allelic loss in both tumors most likely represent patients with a single primary and metastatic spread. Thus, genetic analysis of these lesions may be useful in investigating the origin of histologically similar synchronous tumors.