Patients with cervical adenopathy suspicious for malignancy are often referred to the Otolaryngology Service for tissue diagnosis. Confirmation of nodal involvement by upper aero-digestive tract tumors (UADT) is best obtained by fine needle aspiration (FNA). Reported studies of FNA for lymphoma diagnosis have yielded conflicting results. Retrospective review of charts and pathology of 161 patients diagnosed with lymphomas yielded 53 patients with cervical adenopathy without apparent UADT. FNA's were performed on 28, and were repeated nine times, for a total of 37. Eleven had Hodgkin's disease and 17 other types of lymphomas. Seven of 37 specimens contained only blood; 15 contained lymphoid cells, nine of which were designated "reactive." Lymphoid cells designated as "atypical" or "suspicious for lymphoma" were found in 13 of the 37 aspirates. Two were diagnostic of lymphoma. Lymphoma was confirmed by histopathologic specimens in all patients, obtained 0-941 days (median 15, mean 73 days) after initial FNA. In lymphoma patients with cervical lymphadenopathy, FNA does not usually suffice for, and often leads to significant delays in diagnosis.

e18568 Background: Genomic aberrations are associated with MM survival and disease evolution. The combination of cytogenetics and Fluorescence in situ hybridization (FISH) reveals more anomalies than CG alone. The incidence of CG abnormalities in African American MM patients (AA) is not known. Methods: Records of patients with MM seen at Kings County Hospital and Downstate Medical Center from 2004 through 2010 were reviewed. Results: Of the 185 records reviewed, 181 (98%) patients had MM; 3 had plasma cell leukemia. Their age ranged from 36-89 years (median 66). There were 84 men (46%) and 97 women. CG analysis was done in 86 patients, 41 (48%) female and 45 (52%) male; of whom 82 were AA. CG was abnormal in 14 male and 6 female patients. Aneuploidy was present in 15 of 20 (75%); 11 of the 15 (73%) were hyperdiploid and 4 (26%) hypodiploid. Hyperdiploid chromosome number ranged from 49 to 67 (median 53). The most frequently aberrant chromosomes were 1, 4, 5 and 15. FISH was done in 42 patients for abnormal chromosomes 13, 14 and 17. One of 37 chromosomes 13 (2.6%), 3 of 28 (11%) chromosomes 14, and 1 of 14 (7.1%) chromosomes 17 displayed molecular abnormalities. In no case was more than one abnormality found. FISH on 28 patients with normal CG yielded only 1 (3.6%) abnormality. Conclusions: Abnormal CG was not more frequent in this AA population than in MM patients from other ethnic groups, but was in male AA patients. Fewer molecular abnormalities were found in this population.

e11004 Background: Locally advanced breast cancers are staged as T3 if larger than 5 cm, or T4 if they involve the overlying skin or underlying chest wall. The purpose of this paper is to evaluate any differences in pathologic attributes of T3 and T4 lesions. Methods: All patients with locally advanced breast cancer without evidence of distant metastasis that presented to the KCHC, Breast Clinic from 1982 to 2010, were examined to determine whether their BC’s involved ≥50% of their breast volumes, defined by gross replacement of at least one hemisphere. Core needle biopsy or post-mastectomy specimens from tumors involving a known per cent of breast volume were evaluated for: 1) pathological grades and lympho-vascular invasion (LVI); 2) hormone receptor (ER/PR) expression >0; 3) epidermoid growth factor 2 (her2) overexpression (3+), determined by immunohistochemical staining or fluorescent in situ hybridization. Results: The data base included 98 patients with T3N1 and 120 patients with T4a-c disease. The tumor mass involved more than 50% of the breast in 23/98 (24%) of T3 lesions and 65/120(54%) T4 lesions (p<0.001). One and 23% of T3 and T4a-c patients respectively presented with total replacement of the breast (p<0.001). The table lists the pathologic attributes of the T3 and T4a-c tumors for which the tests were conducted. There were no significant differences between any of the pathologic attributes of the T3 and T4a-c groups. Conclusions: These data are consistent with the hypothesis that erosion of the overlying skin or underlying chest wall in breast cancer may be simply due to neglect and delay, rather than an inherent biologic aggressiveness. P calculated by Fisher’s exact test (2-sided). T3 % T4a-c % P Grade 3 51/82 62 51/84 61 =0.874 ER>0 36/63 57 38/67 57 =1.000 PR>0 29/63 46 31/67 46 =1.000 ER/PR>0 27/63 43 33/67 49 HER2+ 25/58 42 21/60 35 =0.455 Triple neg 13/59 22 17/60 28 LVI 6/20 30 11/29 38 =0.761

6573 Background: ATLL is a peripheral T cell neoplasm with characteristic clinical and immunopathological features, occurring in some individuals infected with HTLV-1, usually from specific endemic regions. Histopathology does not distinguish ATLL from non-HTLV1-related TCL. HTLV serology (HTLVs) can rule ATLL out, but does not exclude the presence of a non-HTLV-related TCL.OBJECTIVETo compare the clinical and immuno-pathological features of ATLL to those of patients with TCL not associated with HTLV1 infection.METHODSThe charts of patients presenting to the Kings County Hospital Center and University Hospital of Brooklyn between 1999 and 2010, whose biopsies showed ATLL or TCL, were reviewed for the following: HTLV1 serology (HTLVS), circulating lymphocytes with multi-lobed nuclei (LML), serum calcium (Ca) levels, cutaneous involvement (Cut), survival (OS), CD25 and Foxp3 expression, and phenotypic markers determined by flow cytometry.RESULTSOf the 59 patients available for study, 24 were men and 35 women, ranging from 33 to 81 years of age (median 54). All but 2 of the 44 ATLL patients were acute or lymphomatous. HTLVs was available in 30 and positive in 28 of the 44 ATLL patients. Data are presented in the table. Only 4 ATLL patients were neither hypercalcemic nor had LML. All Foxp3+ patients were CD25+. Of the 28 patients known to be serologically positive for HTLV1, 25 were CD25+ (89%). On flow cytometry, 12 of 12 ATLL patients were CD25+ but only 6% were CD7+. None of these parameters were determined for TCL patients.CONCLUSIONSATLL and TCL can usually be diagnosed by clinical and laboratory findings. CD25 and Foxp3 expression may prove helpful in distinguishing between them. [Table: see text].

6622 Background: CG characteristics play a critical role in outcome and therapy of AML. The incidence of normal CG in general population is estimated at 40%. The incidence of abnormal CG in AA population is not known. CG abnormalities help define the pathogenesis of AML subtypes and are prognostically significant; they occur in 40% of all AML cases, but their incidence in African Americans is not known. Methods: Records of adult patients with acute leukemia at Kings County Hospital and Downstate Medical Center from 2004 through 2010 were reviewed. The CG data were stratified into three prognostic groups: good; inv (16), t (8:21) and t (15:17); intermediate: normal CG, +8, t (9:11); poor: ≥ 3 CG abnormalities, including: -5, 5q-, -7, 7q-, inv (3), t (6:9), t (9:22). Results: Of the 52 patients with acute leukemia, 46 had AML; 29 were male and 17 female. Forty-three (93%) of these patients were AA. Ages ranged from 20-82 years (median 53), 15 (33%) having been over 65. Seven (15%) patients had a prior hematological malignancy (HM). CD34 was over-expressed in 23 of 46 (48%) patients with AML. CG data was available in 41 patients, 26 men and 15 women; and was normal in 13 (31%) and abnormal in 28 (69%). Of patients with de-novo AML and known CG; 21 (62%) were abnormal; and 13 (38%) were normal. The distribution of patients in risk stratified groups were: good: 8 (20%); intermediate: 19 (46%); and poor: 14 (34%). In patients with prior HM five (71%) had abnormal CG. The median age of patients with good, intermediate and poor risk CG was 57, 53 and 57 years respectively. Intermediate or poor risk AML occurred in 22 of 26 men (85%) and 11 of 15 women (73%). Conclusions: Our data suggest that CG in AA AML patients are more often abnormal, and of intermediate or poor risk than in the general population.

Abstract 5077 Background: The incidence of BM fibrosis in MM is low and uncertain, and its causes are not known. Cytogenetic and fluorescence in situ hybridization (FISH) in some MM patients reveals prognostically significant anomalies. Methods: Records of patients with MM seen at Kings County Hospital from 2004 through 2010 were reviewed, the histological sections of patients reported to have fibrosis we re-examined. The degree of fibrosis was graded according to the World Health Organization system. Results: Records of 113 patients were reviewed, 110 (97%) were African American (AA). Of these, 62 (55%) were female and 51 (45%) male. Their age ranged (median 65) from 38 to 89 years. Cytogenetic data (CGD) was available in 46 patients; and abnormal in 10 (22%) and normal in 36 (78%) of those. All patients with abnormalities of chromosomal number were hyperdiploid. Of 113 patients, 62 (55%) were female, 110 were African American. Ages ranged from 38–89 (median 65) years. Cytogenetic data was available for 46 patients and abnormal in 10 (22%). All patients with abnormal chromosome numbers were hyperdiploid. FISH studies to detect abnormalities in chromosomes 13, 14 and 17 were available in 25 and abnormal in 2 (8%). BM fibrosis had been initially noted in 27 of 113 patients (24%), and confirmed by another hematopathologist; 17 (63%) were women. The ages of the patients with fibrosis ranged from 4–79 years: median age was 67, 67 for the women and 62 for the men. Focal and grades 1, 2 and 3 fibrosis were noted in 2 (7%), 12 (44%), 7 (26%) and 6 (22%) of patients. Grade 3 fibrosis was found in 24% of the women and 20% of the men with fibrosis. CG data was available for 17 fibrosis patients and abnormal (hyperdipliod) in 2 (18%). FISH studies for chromosomes 13, 14 and 17 were normal in the 7 patients studied. CGD for all 8 of the female fibrosis patients studied were normal, and abnormal in 2 of the 9 men (22%). Conclusions: Of our 113 AA myeloma patients 25% had detectable BM fibrosis, but it was grade 3 in only 5%. Female preponderance was more marked in the patients with fibrosis than in the whole MM group. CG and FISH data did not distinguish patients with and without fibrosis. Disclosures: No relevant conflicts of interest to declare.

We report four HIV-related lymphoma patients, and review those from eight other published studies, who responded rapidly and durably to highly-active antiretroviral therapy (HAART) alone. Nine of these patients had localized, extra-nodal lymphomas, while two had primary effusion lymphomas (PEL); eight were of B cell origin. Some HIV-related lymphomas may be effectively treated, and perhaps cured with immune reconstitution alone. Because HAART therapy markedly improves the prognosis of HIV-associated lymphomas, it is now often initiated as soon as they are diagnosed. We report four patients who responded to HAART before chemotherapy was initiated, and who have remained in remission with no other treatment. The literature review revealed a total of eight HIVinfected patients whose biopsy proven lymphomas also responded to HAART alone. Here, we report and summarize these 12 patients’ data. In three of our own patients in whom HAART preceded planned chemotherapy, rapid and complete remissions of their lymphomas occurred; a fourth patient remained in remission 6 months after total cutaneous tumor resection and HAART. Table I presents a detailed outline of each patient’s data [1–8]; patients 1–4 are those reported by us. Table II is a summary of the HIV-related data and time intervals. CD4 levels at lymphoma diagnosis were 10–92 in six patients (median 35, mean 41), and 195–256 in four patients; viral loads were 410. HAART was not initiated until lymphoma diagnosis (regimens outlined in Table I), though HIV was diagnosed as long as 27 months earlier. Two patients had primary effusion lymphomas (PEL), one Hodgkin disease (HD) and eight localized extra-nodal lymphomas (77%) (Table I) Patient no. 1 presented with an area of multi-nodular cutaneous infiltration of the abdominal wall, found to be a high grade lymphoblastic neoplasm. HAART was initiated after resection and no recurrent or additional lymphoma had developed in the ensuing 6 months. Lymphoma diagnosis was established by incisional or excisional biopsy of four peripheral masses, core needle biopsy of two lung masses and three brain masses, and cytologic analysis of the two PEL patients’ effusions. The results of marker studies are presented in Table I. CD4 levels increased after HAART by a median of 3.47 fold, while the viral loads markedly diminished (Table II). After HAART therapy, the median CD4 count of the 12 patients rose from 74 to 315/mcl, and the viral load median fell from 259 000 to 875 (Table II). The interval between HAART initiation and lymphoma response ranged from 1 to 150 weeks, with a median of 10 and mean of 24 weeks. Remission duration ranged from 1 to 36 months, with a median of 16 and a mean of 36 months (Table II). Neither our patients’ lymphomas nor those reported by others are known to have relapsed after response to HAART, and none received

Purpose: To determine the safety and efficacy of treatment with gamma interferon (IFNγ) in patients with metastatic carcinoid tumor. Patients and methods: 51 patients were enrolled on this Phase II Eastern Cooperative Oncology Group (ECOG) study. Seventy five percent of them had hormonally active tumors. Treatment consisted of IFNγ subcutaneously at a daily dose of 0.1 mg/m2. Patents were evaluated for toxicity weekly for the first month and monthly thereafter; response was determined radiologically every 8 weeks. Results: Patients received treatment with IFNγ for a median of 17.9 weeks (range 2–175). Toxicity was generally mild and expected: 61% experienced noninfected fever and 21% developed granulocytopenia. Three patients (6%) had a partial response; there were no complete responses. Median time to progression was 5.5 months (95% confidence interval 3.9–11.1). The 1-year progression free rate was 28% (13.4–43.4%). Median survival was 42 months, with a 1-year survival rate of 67% (53.3–80%). Discussion: This Phase II study demonstrated that therapy with IFNγ in patients with metastatic carcinoid tumor was well-tolerated, but did not produce significant antitumor effects. The overall results were somewhat comparable to those previously seen with alpha interferons as well as cytotoxic drugs.

Massive hepatic necrosis following exposure to phenytoin and trimethoprim-sulfamethoxazole is a rare occurrence and to the best of our knowledge has not been reported previously. Acute hepatic failure following administration of trimethoprim-sulfamethoxazole has rarely been seen, and only 4 cases have been well documented pathologically. We report a case of acute liver failure in a 60-year-old woman following ingestion of phenytoin and trimethoprim-sulfamethoxazole concomitantly over a 9-day period. Autopsy findings revealed acute fulminant hepatic failure. This case demonstrates the effects of chemical-chemical interactions in the potentiation of hepatotoxicity of single agents and specifically illustrates the need for discontinuing trimethoprim-sulfamethoxazole in the presence of early liver injury.

Background: Neoadjuvant chemotherapy facilitates breast conservation in stage II breast cancer patients, whose primary tumors are assumed to be invasive because they are palpable. However, chemotherapy may not be indicated in the minority of patients whose clinically T2 tumors are completely or predominantly in situ. Almost all previous studies of core needle biopsy in breast cancer have been concerned with nonpalpable, mammographically detected tumors, and none have evaluated its ability to quantitatively determine the amounts of in situ and invasive disease.Methods: From September, 1992 to December, 1997, core needle biopsy was performed on all patients presenting to the Kings County Hospital Breast Clinic with palpable breast masses. Carcinoma was present in both core needle biopsy samples and surgical specimens subsequently obtained from 95 of 99 patients. Each specimen was evaluated for tumor type, histologic grade, and the amounts of in situ and invasive carcinoma it contained, and the results from surgical and core needle biopsy specimens from the same patients were then compared.Results: The surgical specimens of 14 patients had completely or predominantly in situ disease. Completely or predominantly invasive disease was present in 67 specimens, and the remaining 14 had significant amounts of both. The high level of agreement between the amounts of in situ and invasive disease in core needle biopsy and surgical specimens is indicated by Pearson and intraclass correlation coefficients of 0.91 (P < .001 and < .00001, respectively). Tumor type was correctly predicted by core needle biopsy in each case. Variables among these patients, including primary tumor size, interval between biopsy and surgery, or administration of neoadjuvant systemic therapy, did not alter agreement between core needle biopsy and surgical specimens.Conclusions: Core needle biopsy can identify palpable breast tumors that are predominantly or completely in situ, and, thus, avoid unnecessary neoadjuvant chemotherapy. It also can demonstrate that a tumor is predominantly invasive, but cannot rule out small invasive foci. For that purpose, complete surgical excision of the tumor is required.

Department of Pathology; State University of New York Health Science Center at Brooklyn; Brooklyn, New York