Are renal reference intervals required when screening for plasma cell disorders with serum free light chains and serum protein electrophoresis?

Renal impairment and polyclonal hypergammaglobulinemia may abnormally increase the serum free light chain (sFLC) ratio, giving false-positive results with current reference intervals. We measured sFLCs with concomitant serum protein electrophoresis (SPEP) and urine protein electrophoresis (UPEP) in 281 patients. Results were interpreted relative to renal function (serum creatinine concentrations) and polyclonal hypergammaglobulinemia. Overall, 78 plasma cell disorders (PCDs) were detected with the serum panel of SPEP/sFLC vs 76 with SPEP/UPEP. In 13 samples with negative SPEP/UPEP, mildly increased ratios up to 3.1 (normal, 0.26-1.65) were observed: 10 were associated with increased serum creatinine and 1 with polyclonal hypergammaglobulinemia; 2 were unassociated with either condition. In 2 samples, decreased kappa/lambda ratios were identified that were clinically significant despite normal SPEP/UPEP. Two monoclonal gammopathies were identified with UPEP and sFLC, but samples were normal with SPEP. Screening for PCDs with a serum panel consisting of SPEP and the sFLC assays is a highly sensitive approach that could eliminate the need for UPEP. A mildly increased kappa/lambda ratio up to 3.1 was observed with increased serum creatinine and/or polyclonal hypergammaglobulinemia that was consistent with pathophysiologic changes, and, therefore, renal reference intervals are recommended.

Association between a shift in vaginal flora on Papanicolaou smear and acute chorioamnionitis and preterm delivery.

Bacterial vaginosis has been implicated as a cause of acute chorioamnionitis and preterm delivery. This study was designed to determine any association between the detection of bacterial vaginosis on a prenatal Papanicolaou (Pap) smear, defined as a shift in vaginal flora, and the subsequent occurrence of acute chorioamnionitis or preterm labor. A 47-mo retrospective case-control analysis comparing 186 patients with histologically-proven acute chorioamnionitis (cases) and 186 controls was performed. Initial prenatal Pap smears were evaluated for the presence of altered vaginal flora. Pap smears from women with acute chorioamnionitis were more likely to have altered vaginal flora than those without chorioamnionitis (P < 0.01). Preterm delivery was also more common among women with a shift in vaginal flora (P < 0.01). In conclusion, women at risk for chorioamnionitis and/or preterm delivery may be able to be identified by the results of their prenatal Pap smear.

Immunocytochemical evaluation of HBME-1, CA 19-9, and CD-15 (Leu-M1) in fine-needle aspirates of thyroid nodules.

Overlapping morphologic patterns that may be observed in goiter, follicular adenoma, and papillary carcinoma can limit the cytologic evaluation of the thyroid gland. In an attempt to develop a useful adjunctive test, the immunocytochemical reactivity of HBME-1, carcinoma antigen 19-9 (CA 19-9), and CD-15 (Leu-M1) was tested on 59 cell block preparations from fine-needle aspirations of the thyroid gland. HBME-1 monoclonal antibody was reactive in all 21 papillary carcinomas, in 4 of 18 adenomas, and in 5 of 20 goiters. CA 19-9 was identified in 13 of 21 carcinomas, 1 goiter, but none of the adenomas. CD-15 was present in 15 of 21 carcinomas, 1 goiter, and 1 adenoma. We conclude that HBME-1 is a sensitive marker of papillary thyroid carcinoma. CD-15 and CA 19-9 are less sensitive but more specific. This panel can be useful to help classify morphologically equivocal lesions. As with all immunocytochemical testing, caution must be exercised in the interpretation of results, and correlation made with morphologic and clinical data.

Cytologic diagnosis of fat emboli in peripheral blood during sickle cell infarctive crisis.

A diagnosis of fat emboli can be suspected in a patient presenting with the typical symptoms of the fat embolism syndrome, but is rarely proved pathologically, except at autopsy. We described a 25-yr-old man with sickle cell anemia who developed an infarctive crisis complicated by unexplained fever, neurologic change, and respiratory abnormalities. Blood drawn from the femoral vein and examined cytopathologically yielded necrotic bone marrow elements admixed with fat. The cytologic finding of fat emboli from necrotic bone marrow provided the diagnosis and helped guide subsequent medical intervention. This sample test is recommended for patients at risk for fat emboli to aid in the clinical diagnosis.

Distinction between bronchioloalveolar carcinoma and hyperplastic pulmonary proliferations: a cytologic and morphometric analysis.

The objectives of this study were to identify key cytologic features for the morphologic differentiation of bronchioloalveolar carcinoma from reactive pulmonary proliferations, and to determine if morphometry of the cytologic specimens could provide additional data to distinguish the two processes. We analyzed 15 morphologic criteria in pulmonary cytologic specimens from 17 histologically proven cases of bronchioloalveolar carcinoma and 13 cases with reactive cellular changes and compared the findings using univariate analysis. This revealed four statistically significant features more commonly associated with bronchioloalveolar carcinoma: (i) predominance of two- and three-dimensional tissue fragments, (ii) tenacious intercytoplasmic connections between cells, (iii) intranuclear cytoplasmic inclusions, and (iv) paucity of multinucleated cellular forms. Morphometric measurement revealed significant differences between the mean of the nuclear area of benign reactive cells and that of the malignant cells. The utilization of these criteria is helpful to diagnose bronchioloalveolar carcinoma and to distinguish it from reactive pulmonary processes.

Cytologic findings in lichen sclerosus.

OBJECTIVE: To describe the findings in gynecologic smears from 29 patients with biopsy-proven vulvar, vaginal or perineal lichen sclerosus. STUDY DESIGN: The computerized database of the surgical pathology division of the Hospital of the University of Pennsylvania was searched for the diagnosis of lichen sclerosus from the years 1989-1992. Once these patients were identified, a search for all associated gynecologic material, including cytologic smears, biopsies and resections, was conducted. All the corresponding slides were obtained from the files and reviewed. RESULTS: Parakeratosis was present in smears from 14 patients (48%) and hyperkeratosis in 18 (62%). Small parabasal cells with an elongated configuration were identified in smears from 17 patients (59%), and squamous atypia was noted in 12 (41%) of them. Following comparison with the biopsy material, the small parabasal cells appeared morphologically similar to the disorganized squamous cells that have been observed in histologic preparations. Frank dysplasia was identified in only one smear, from a patient with lichen sclerosus who had had a previously resected vulvar carcinoma; the vulvar biopsy disclosed carcinoma in situ. CONCLUSION: Parakeratosis, hyperkeratosis and small, elongated parabasal cells are common features in gynecologic smears from patients with lichen sclerosus. Focal squamous atypia may also be present but is not necessarily reflective of a preneoplastic process. The presence of these four features in routine gynecologic smears should prompt a correlation with the clinical and histologic diagnoses.

Ki-67 antigen staining as an adjunct to identifying cervical intraepithelial neoplasia.

The purpose of the work was to determine the feasibility and predictive value of Ki-67 immunostaining of cervical cytology and the detection of cervical dysplasia. Air-dried cervical smears were stained with MIB-1 antibody to identify the Ki-67 antigen. Nuclear decoration in abnormal squamous nuclei determined immunoreactivity. One hundred twenty-four nonpregnant patients underwent colposcopy and directed biopsies for abnormal cytology. Sensitivity (0.89), specificity (0.65), positive predictive value (0.60), and negative predictive value (0.91) were found for Ki-67 immunostaining in detection of high-grade cervical intraepithelial neoplasia (CIN) in 124 patients and positive Ki-67 staining was a significant predictor of high-grade CIN in both univariate (odds ratio 15.5 (95% CI 5.5-43.8) and multivariable (odds ratio 21.5 (95% CI 5.0-92.0) analysis. In 101 patients with ASCUS and LGSIL, Ki-67 immunostaining demonstrated the following in detection of high-grade CIN: sensitivity (0.96), specificity (0.67), positive predictive value (0.49), and negative predictive value (0.98). Ki-67 immunostaining of cervical cytology is a predictor of significant cervical pathology with high sensitivity and negative predictive value. Ki-67 immunostaining of cervical cytology may represent a new and cost-effective triage tool for patients with minor abnormalities on cytology.

Review and hypothesis: the eosinophil and peripartum heart disease (myocarditis and coronary artery dissection)--coincidence or pathogenetic significance?

OBJECTIVES: To examine a possible relationship between peripartum heart disease (myocarditis and spontaneous coronary dissection) and the presence of eosinophils. BACKGROUND: Eosinophils have been shown to have potential collagenolytic and cytotoxic activity. Eosinophils may play a role in postpartum uterine involution. The presence of eosinophils in spontaneous coronary dissection and myocarditis in the postpartum period raises the possibility of a role for eosinophils in these diseases. METHODS: We reviewed the files of one of us (S.M.F.) for cases of peripartum myocarditis and spontaneous coronary dissection and assessed the frequency of eosinophilic inflammation. Seventeen postpartum myocarditis and/or cardiomyopathy cases were found and two spontaneous coronary dissections. Fifteen sex- and age-matched controls on non-postpartum myocarditis and borderline myocarditis were evaluated and eosinophil counts per unit area compared. Also, a Medline search of all previously published cases of spontaneous coronary dissection was performed back to 1966. RESULTS: Of the 16 heart biopsies and one autopsy in the peripartum period, 10 were shown to contain easily identified eosinophils (6 myocarditis, 1 borderline, 3 cardiomyopathy). When presence of eosinophils was compared with the control group, a statistically significant difference was obtained (P = 0.036). The two new spontaneous coronary dissection cases had eosinophils along the dissection plane; the literature search produced 13 of 24 autopsied peripartum spontaneous coronary dissections with eosinophils for a total of 15 of 26 with our cases. CONCLUSIONS: An association exists between eosinophils and peripartum cardiac disease (myocarditis and spontaneous coronary dissection). The role of eosinophils in labor, uterine involution and collagenolysis and the possible relation to cardiac disease are discussed.

Quantitative image analysis of MIB-1 reactivity in inflammatory, hyperplastic, and neoplastic endocervical lesions.

The diagnosis of endocervical neoplasia can be difficult as it is sometimes mimicked by proliferative or reactive glands. MIB-1 is a proliferation marker that can aid in the diagnosis of squamous intraepithelial lesions (SIL) of the cervix and vulva, but its potential value in the diagnosis of endocervical lesions has not been fully explored. Ten formalin-fixed, paraffin-embedded cases of each of the following were obtained: morphologically normal endocervical glands from patients with cervical SIL, endocervicitis, microglandular hyperplasia (MGH), and endocervical adenocarcinomas (eight in situ, two invasive). Microwave unmasking of antigens was performed prior to immunohistochemical staining for MIB-1 using the avidin/biotin peroxidase method. Labeling indexes were calculated for 34 specimens (10 adenocarcinoma. 8 each of the other diagnoses) using image analysis (Samba 4000). There was diffuse MIB-1 reactivity in adenocarcinoma (labeling index 57-96%, mean 80%), minimal focal reactivity in normal glands underlying SIL (labeling index 0.8-4.3%, mean 2.4%), moderate spotty reactivity in MGH (labeling index 2.9-18.4%, mean 8.5%), and minimal to focally diffuse reactivity in endocervicitis (labeling index 1.0-13.3%, mean 5.7%). These data indicate that the percentage and distribution of MIB-1-reactive endocervical cells can be of diagnostic utility in distinguishing neoplastic glands from those of endocervicitis and MGH.

Cytologic features of pulmonary malakoplakia related to Rhodococcus equi in an immunocompromised host.

Cytologic features are described in bronchial brushings of a large cavitary lung mass from an immunosuppressed patient who had undergone liver transplantation. Scattered histiocytes with abundant eosinophilic, vacuolated cytoplasm were noted in a background of bronchial cells. Within approximately one fourth of the histiocytes, targetoid intracytoplasmic inclusions were present, consistent with Michaelis-Gutmann bodies. They were diffusely positive by histo-chemical staining with von Kossa, Gomori methenamine silver, and periodic acid-Schiff stains, and focally positive with Prussian blue stain. Cultures of the abscess yielded Rhodococcus equi. The characteristic microscopic features of pulmonary malakoplakia can be discerned in bronchial brushings, and should be sought, particularly in immunocompromised patients.

The detection of endocervical gland involvement by high grade squamous intraepithelial lesions in smears prepared from endocervical brush specimens.

The cytologic features of squamous cell carcinoma in situ with endocervical gland involvement have been described in cervical smears. We evaluated the presence of two types of cellular fragments in 43 cervical smears of high grade squamous intraepithelial lesions (HGSIL) to assess their ability to predict glandular involvement by HGSIL in subsequent cone biopsies. An endocervical brush was used to obtain all endocervical specimens. Of 16 cases without glandular involvement, fragments were present in 13 smears. Of 27 cases with glandular involvement, fragments were absent in 11 smears. No statistical association was identified between the presence of abnormal cellular fragments on cervical smears of HGSIL and endocervical gland involvement on cone biopsies.

Efficacy of the Endo-pap sampler in detecting endometrial lesions.

OBJECTIVE: To determine the efficacy of the Endo-pap sampler, a device used to collect endometrial tissue fragments for cytologic examination, in detecting endometrial lesions. STUDY DESIGN: Reports from 1,983 Endo-pap smears collected from 1,248 outpatients over a 44-month period were retrospectively examined in conjunction with corresponding biopsy or hysterectomy reports. Endo-pap smears, histologic sections and clinical information were reviewed in cases originally reported as abnormal. RESULTS: Endometrial abnormalities were reported in 29 Endo-pap smears (1.5%). Among 21 reported as atypical, follow-up was normal in 19 but revealed hyperplasia in 1 and hyperplasia with a focus of carcinoma in 1 on histologic examination. Three were interpreted as malignant and one as suspicious; all had adenocarcinoma at endometrial biopsy. Endometrial abnormalities were reported in 19 of 220 patients who underwent endometrial biopsy following a previous Endo-pap smear reported as negative for malignancy. Their diagnoses included endometrial polyps (14), hyperplasia without atypia (3), atypical hyperplasia (1) and endometrial adenocarcinoma (1) three years after a negative Endo-pap smear. CONCLUSION: Although the sensitivity and specificity for malignancy were 100%, the overall sensitivity for any pathologic lesion was only 28%. We recommend proceeding directly to endometrial biopsy for a conclusive diagnosis in symptomatic patients.

Clinical significance of increasing histologic severity of acute inflammation in the fetal membranes and umbilical cord.

The purpose of this study was to determine the importance of varying histologic stages of inflammation in the placental membranes and cord. Acute inflammation was histologically staged in fetal membranes and umbilical cord sections from 2899 placentas received from consecutive singleton deliveries. Then clinical data were collected for a subset of randomly selected placentas with stage 1 through stage 4 membrane inflammation (n = 212) and without significant inflammation (stage 0, n = 216). Statistical analyses revealed that increasing stage of membrane inflammation was associated with an increasing rate of funisitis, perinatal death, and preterm birth (P < .05). Inflammation permeating the entire trophoblastic layer of the chorion (stage 1) was associated with clinical symptoms of intrauterine infection and thus was an important pathologic finding. Acute necrotizing chorioamnionitis was very strongly associated with perinatal death and preterm birth. Acute funisitis was a more specific but less sensitive marker for perinatal complications than inflammation in the membranes. With increasing stage of funisitis, there was an increased incidence of clinical symptoms of intrauterine infection, preterm birth, and perinatal death. Almost three-fourths of the cases with histologic evidence of membrane inflammation were clinically silent. In conclusion, increasing histologic stages of inflammation of the membranes and cord are associated with an increased rate of perinatal morbidity and mortality. Stage I membrane inflammation provides a clinically acceptable minimum threshold for the reporting of pathologic changes.

Myocardial alterations in diabetes and hypertension.

Diabetes mellitus is a complex group of diseases that has hyperglycemia as a common metabolic abnormality. Although it is well-known that diabetic patients are susceptible to the effects of large vessel atherosclerosis with specific cardiac and cerebral complications, the association of diabetes mellitus with cardiac dysfunction caused by cardiomyopathy in the absence of significant coronary artery disease has been recognized for many years. However, the pathogenesis of diabetic cardiomyopathy remains unknown and has been somewhat controversial. Specifically, whether diabetes mellitus with its metabolic effects is sufficient to account for cardiomyopathy remains to be proven. This paper reviews the evidence for and against a metabolic etiology. In addition, we review the clinical and experimental evidence that supports the view that diabetes mellitus acts together with hypertension to produce structural damage in the heart that manifests as ventricular dysfunction and ultimately congestive heart failure. The concomitant effects of the metabolic derangements of diabetes and the vascular abnormalities associated with hypertension may lead to microvascular-induced tissue injury. Findings supporting this hypothesis are presented, along with observations suggesting that treatment with vasodilating calcium channel blockers or angiotensin converting enzyme inhibitors may be beneficial in regard to tissue pathology and mortality in experimental models. Recent clinical studies also support a role for the microcirculation in diabetics. Finally, it is suggested that if the microcirculation is pathogenetically involved in diabetic cardiomyopathy, then agents that improve microcirculatory flow along with tight control of hypertension may be as beneficial in the treatment or prevention of diabetic cardiomyopathy as strict metabolic control of hyperglycemia.

Cytology of small cell carcinoma of the urinary bladder.

To characterize the cytologic spectrum of small cell carcinoma of the urinary bladder, a review of 42 urinary cytology specimens from 13 patients with histologically proven tumors was conducted. Patients ages ranged from 45 to 81 yr (mean 68.9). In four tumors, small cell carcinoma was the sole malignant cellular component: all 11 urinary specimens in these patients harbored cells with features of an undifferentiated small cell carcinoma. In the remaining nine tumors, small cell carcinoma appeared with transitional, squamous cell or adenocarcinoma, but in four, small cell carcinoma was the sole invasive component. Almost one third of urinary specimens in this group lacked a small cell component. Neuroendocrine differentiation was confirmed by immunopathology in eleven cases (neuron specific enolase positive in 11 of 12, synaptophysin in 2/11, chromogranin in 2/13, Leu 7 in 2/7), and by ultrastructural analysis in two. Small cell carcinoma is a cytologically recognizable variant of bladder cancer, but admixture with other malignant components may mask its appearance in urinary specimens.

Cytology of the thyroid gland: pitfalls in aspiration of the fibrotic nodule.

Fibrotic changes can be found in a variety of pathologic processes that affect the thyroid gland, and yet fine-needle aspiration (FNA) correlation of these lesions is not well-known. Cytologic findings are described from three different lesions, all with extensive fibrosis, that occurred in thyroid gland FNA. In each case, fibrotic changes resulted in cytologic changes potentially representing pitfalls in aspiration diagnosis. FNA revealed, solely or predominantly, fibroblasts in 2 patients whose subsequent thyroidectomies disclosed diffusely sclerotic papillary carcinoma and multinodular goiter. In the third case, an initially suspicious FNA was followed by multiple unsuccessful attempts at a second FNA. At thyroidectomy, a follicular adenoma with post-FNA fibrosis and infarction was identified. The differential diagnoses in thyroid FNA of fibrotic nodules can be broad, and this is discussed. Awareness of potential pitfalls may improve diagnostic accuracy.

Prevalence, identification and significance of fiber contaminants in cervical smears.

OBJECTIVE: To determine the prevalence of fiber contaminants, to provide morphologic descriptions of various fiber types and to try to understand the significance of their presence in cervical smears. STUDY DESIGN: Fibers from cotton swab tips, hair, cardboard slide transporters, Cytobrush bristles and three commercial brands of tampons were smeared onto glass slides, Papanicolaou stained and examined. Then a prospective study of consecutively screened cervicovaginal smears from 1,368 women was undertaken to identify fiber contaminants. RESULTS: Fibers were identified in 178 cases (13%). In 120, the fibers were considered extrinsic, commonly having the microscopic features of cardboard. In 58, the fibers were considered to be intrinsic to the smear because of their presence in the same plane and alignment as the cells, an associated inflammatory cell reaction, and adherent or absorbed cellular products, occasionally hemosiderin. The most common intrinsic fibers were cotton and rayon. No statistically significant relationship was identified for the presence of intrinsic fibers and patients' ages, stated menstrual dates or pathologic lesions. CONCLUSION: Fiber contaminants are commonly present in cervicovaginal smears, usually do not originate in the patient (are "extrinsic") and have no pathologic significance. Knowledge of the microscopic features of fiber contaminants enables them to be identified confidently and distinguished from fungal hyphae and other filamentous pathogens.

Protocol for immunocytochemical detection of SIL in cervical smears using MIB-1 antibody to Ki-67 [corrected].

Ki-67 nuclear antigen is expressed in upper epithelial levels of intraepithelial neoplasia of the cervix and vulva, variably in condyloma, and in basal and parabasal cells of normal squamous mucosa in histologic preparations. The application of antibodies to Ki-67 as a marker of squamous intraepithelial lesions in cervical smears was explored using either air-dried, acetone-fixed cervical smears obtained from 106 consenting patients or a single slide from archival two-slide cases of squamous intraepithelial lesions MIB-1 monoclonal antibody to Ki-67 was tested using two immunocytochemical techniques. In one set of smears, avidin-biotin peroxidase was used for detection and diaminobenzidine with H2O2 as the chromogen. Some specimens were incubated with 0.3% H2O2 and phosphate buffered saline for blockade of endogenous peroxidase. Alternatively, other air-dried smears were stained using alkaline phosphatase antialkaline phosphatase for detection and new fuchsin as the chromogen. Nuclear staining in squamous intraepithelial lesions was identified in air-dried smears using all of the above methods. Slides stained with avidin-biotin peroxidase and blocked with 0.3% H2O2 and phosphate buffered saline showed less background staining from neutrophils and erythrocytes compared with those without blocking. Slides stained using alkaline phosphatase antialkaline phosphatase showed excessive cytoplasmic staining of endocervical cells, making intraepithelial difficult. No nuclear staining of squamous intraepithelial lesions was observed in destained archival smears. Air-dried smears blocked with 0.3% H2O2 and phosphate buffered saline, incubated with MIB-1, and stained using avidin-biotin peroxidase gave the best results for identification of Ki-67 expression in squamous intraepithelial lesions.

Immunohistochemical staining for proliferating cell nuclear antigen, BCL2, and Ki-67 in vulvar tissues.

Dysplastic cells of the uterine cervix commonly react with antibodies to Ki-67 and proliferating cell nuclear antigen (PCNA), but less is known regarding the reactivity of vulvar lesions. Paraffin-embedded slides from seven normal vulvar tissues, eight cases of lichen sclerosus, and 10 cases of vulvar intraepithelial neoplasia (VIN), many with associated condylomatous changes, were immunostained with antibodies to PCNA, Ki-67, and BCL2 protein, a protooncogene product normally expressed only in basal cells of squamous epithelium. Ki-67 and PCNA nuclear staining was largely restricted to basal and parabasal cells in normal tissues and lichen sclerosus. Focal midepithelial staining with PCNA and Ki-67 was seen in one case of lichen sclerosus; a vulvar biopsy of this patient 1 year later was negative for VIN. Both antibodies stained dysplastic cells at higher epithelial levels in VIN, but Ki-67 was more consistently reactive and showed a sharper distinction from adjacent histologically uninvolved epithelium compared to PCNA. Condylomatous changes were variably stained. The pattern of BCL2 staining was identical in normal vulva, lichen sclerosus, and VIN, but BCL2 decorated occasional mitotic figures in VIN. Overall, Ki-67 was a better marker of vulvar dysplasia than PCNA.