Observer variation in the interpretation of xeromammograms.

We have examined variation in the interpretation of xeromammograms among radiologists designated to take part in a Canadian multicenter randomized controlled trial of screening for breast cancer. Radiologists read 100 xeromammograms comprising 10 histologically proved cancers, 40 benign abnormalities, and 50 normal films. Radiologists' opinions differed widely on the frequency of suspected or identified cancer. The diagnostic category "suspicion of cancer" or "cancer" was selected by radiologists for 10-55% of the films, and biopsy or aspiration was recommended for 21 to 53% of patients whose films were examined. Agreement on specific diagnostic categories was greatest for the diagnosis of cancer; agreement was least for the diagnosis of benign abnormalities and intermediate for the diagnosis of normality. Known cancers were in general correctly identified. These results indicate a need for development of methods to reduce observer variation in a interpretation of xeromammograms while preserving diagnostic sensitivity and validity. Results also emphasize the importance of developing strategies to ensure quality control in multicenter trials.

In situ transcription and detection of CD1a mRNA in epidermal cells: an alternative to standard in situ hybridization techniques.

To develop methods for the investigation of mRNA transcription in rare epidermal cells, we used in situ transcription to study CD1a mRNA in isolated CD1a positive cells. We chose to study this Langerhans cell marker because it is not known which epidermal cells actually produce the CD1a protein and because there is evidence that CD1a mRNA is alternately spliced, a situation which could lead to truncated or alternate protein products in CD1a surface protein negative cells. Disaggregated epidermal cells were resolved into CD1a surface protein positive and negative groups by fluorescence activated cell sorting and cytocentrifuged onto glass slides. A synthetic 52 base, CD1a specific anti-sense oligomer was hybridized to CD1a gene transcripts in these cells, and radiolabeled cDNA synthesized in situ on the oligomerprimed CD1a transcripts. The labeled cDNA fragments were visualized in the cells of origin by autoradiography, and grains per cell were counted. Sixty-eight percent of cells expressing CD1a protein contained CD1a mRNA, as evidenced by grain counts more than two standard deviations above the mean value for similar cells carried through the same procedure with a control oligomer, or the mean value of CD1a surface protein negative cells treated with the CD1a specific oligomer. Thus, it seems likely that the CD1a protein positive epidermal cells use CD1a mRNA to make their own CD1a protein, and that a truncated or masked CD1a protein is not made by CD1a negative neonatal foreskin epidermal cells. In our hands, in situ transcription is simpler and faster than standard methods of in situ hybridization with prelabeled cDNA or RNA probes. Furthermore, it can be applied to the detection of any message of known sequence. The combination of cell sorting and in situ transcription can be used to localize and quantify the expression of specific mRNA by individual cells, allowing the study of rare and difficult-to-obtain cells.

Molecular cloning of CD1a (T6), a human epidermal dendritic cell marker related to class I MHC molecules.

To investigate the structure, function, and control of CD1a, we have cloned a 1.6-kbp cDNA which encodes the expressed CD1a protein and includes untranslated 5' and 3' sequences and the poly-A tail. As the protein recognized by the monoclonal antibody OKT6, CD1a is a useful marker for Langerhans cells (LC). CD1a is found on these cells and on thymocytes, suggesting an important immunologic role for this molecule. We constructed a cDNA library in lambda gt10 using mRNA from MOLT-4, a cell line that expresses the CD1a surface antigen. We then screened the library with an oligonucleotide synthesized according to a known partial sequence for CD1a, and subcloned the cDNA and its restriction fragments into pGEM for sequencing and probe production. Based on this sequence the CD1a protein is predicted to consist of three extracellular domains (alpha 1-3), a hydrophobic transmembrane region, and a cytoplasmic tail. DNA 5' to the alpha 1 region may undergo alternative exon splicing. There is high sequence identity between the beta-2 microglobulin binding region of MHC I molecules and CD1a. The secondary structure predicted for CD1a is very similar to the actual structure of HLA-A2, a classical MHC I molecule. The similarity includes the beta pleated sheets and alpha helices which form the antigen binding groove of the alpha-1 and alpha-2 domains. The homology predicted between CD1a and HLA-A2 in these regions appears to exist on the level of secondary structure despite low primary nucleotide and amino acid sequence identity. The structural data and probes we have developed should facilitate studies of the function of CD1a as well as novel investigations of LC.

Granulocyte macrophage--colony-stimulating factor (GM-CSF) decreases CD1a expression by human Langerhans cells and increases proliferation in the mixed epidermal cell-lymphocyte reaction (MELR).

Langerhans cells (LC) undergo a variety of phenotypic and functional changes in vitro. To determine the effects of granulocyte macrophage--colony-stimulating factor (GM-CSF), tumor necrosis factor-alpha (TNF-alpha), and interleukin 1-alpha (IL-1) on LC phenotype in vitro, epidermal cell suspensions were enriched for LC by density-gradient centrifugation and cultured in the presence of 10 ng/ml of these cytokines. The percentage of cells expressing the surface protein CD1a was determined by flow cytometry. This percentage typically dropped after 48 h culture in both control and cytokine-treated medium to less than half that of the starting value. By the fifth day, the percentage of cells expressing CD1a in TNF-alpha and IL-1--treated cultures was still near half of the starting value, slightly above that of control cultures. Treatment with GM-CSF caused large and consistent decreases in the percentage of epidermal cells expressing CD1a. Cell viability in each of the three cytokine-treated cultures was identical to the control cultures, with essentially all cells having died by the sixth day after isolation. To determine the functional effects of these cytokines, the cytokine-containing medium was replaced after 72 h with medium containing purified allogeneic T cells and proliferation measured. Preliminary experiments showed no increased proliferation induced by IL-1 or TNF-alpha--treated epidermal cells. GM-CSF-treated epidermal cells induced 2-3 times more T-cell proliferation than epidermal cells cultured without additional cytokines. We conclude that GM-CSF, a cytokine known to be produced by keratinocytes in vitro, decreases CD1a expression by human LC and increases their ability to stimulate proliferation by allogeneic T cells.

Human dermal endothelial cells express membrane-associated mast cell growth factor.

Mast cell growth factor (MGF), a molecule that serves as a ligand for the receptor tyrosine kinase c-kit, is important in mast cell differentiation, migration, and activation. Previous studies of paraffin-embedded human skin using antibody to murine MGF and reverse transcription-polymerase chain reaction have demonstrated MGF protein and mRNA expression in keratinocytes and isolated dermal cells. We utilized a monoclonal antibody to human MGF to further define patterns of immunoreactivity in frozen specimens of neonatal and adult skin from normal individuals and from patients with urticaria pigmentosa. In addition to keratinocytes and isolated dermal cells in normal and urticaria pigmentosa skin, MGF was detected in cells lining superficial and mid-dermal vessels. Co-expression of MGF and the vascular antigen CD31, and immunoelectron microscopy, identified MGF-positive cells as endothelial cells. Patterns of endothelial MGF expression were not influenced by mast cell degranulation and endothelial E-selectin induction in vitro. By ultrastructure, unfixed specimens demonstrated MGF expression both within the endothelial cytoplasm and in association with lumenal, but not ablumenal, surfaces. Specimens fixed with Nakane's solution had diminished endothelial cytoplasmic MGF reactivity, but lumenal expression was maintained, suggesting persistence of a membrane-associated reactivity. MGF mRNA was also detected in cultured dermal microvascular endothelial cells using reverse transcription-polymerase chain reaction. These data establish human dermal endothelial cells as sites of MGF production and expression in human skin. Mast cell precursors must home to skin via vascular channels and differentiate in the immediate perivascular space. Thus, endothelial MGF may be an important determinant of adhesion and differentiation of mast cell progenitors expressing receptors for MGF.

The immune response to class I-associated tumor-specific cutaneous T-cell lymphoma antigens.

In order to determine whether the neoplastic T cells from patients with cutaneous T-cell lymphoma express tumor-specific antigens that can serve as the targets of an immune response, we took advantage of family-specific monoclonal antibodies, magnetic bead technology, and recombinant cytokines, which provided the previously precluded ability to isolate and expand populations of purified tumor and autologous CD8 cytotoxic T cells. Four patients with advanced cutaneous T-cell lymphoma had CD8 cells that specifically killed autologous tumor in a class I limited fashion. Tumor cell cytolysis could be specifically enhanced by pre-culture with autologous gamma-irradiated tumor. The cytolytic T cells produced tumor necrosis factor-alpha in response to stimulation with autologous tumor. The presence of tumor-specific cytotoxic T cells recognizing distinctive class I associated molecules on cutaneous T-cell lymphoma tumor cells suggests that infiltration of early lesions by CD8 cells reflects host immunity to the neoplasm. These studies provide the foundation for the development of tumor vaccines through the use of cytotoxic T cells to isolate and characterize tumor-associated cutaneous T-cell lymphoma peptides.

Regulation of transgenic class II major histocompatibility genes in murine Langerhans cells.

I-E is a class II major histocompatibility complex molecule normally expressed by Langerhans cells. A series of transgenic mice were developed previously that carry E alpha d gene constructs with promoter-region deletions that cause expression of I-E by different cell types when maintained on a B6 (I-E[-]) genetic background. To study cis-acting gene sequences that regulate expression of class II proteins by Langerhans cells, we identified transgenic I-E expression by tissue immunoperoxidase staining and by epidermal cell suspension immunofluorescence cytometry. Mice with a transgene containing 1.4 kilobase pairs (kb) of flanking sequence 5' to the E alpha initiation site expressed barely detectable levels of I-E on a tiny percentage of Langerhans cells, indicating that sequences promoting Langerhans cell expression of E alpha exist between 2.0 and 1.4 kb 5' of the E alpha initiation site. Removal of an additional 170 bp of 5' flanking sequence caused near-normal levels of expression by approximately one third of epidermal Langerhans cells, which contrasts with studies that showed minimal transgene expression by splenic dendritic cells in these animals. Thus, sequences between 1.4 and 1.23 kb 5' of the E alpha initiation site decrease expression of I-E by epidermal Langerhans cells, but enable I-E expression by splenic dendritic cells. These studies identify Langerhans cell-specific regulatory sequences and genetic regions controlling major histocompatibility complex class II gene expression in Langerhans cells and splenic dendritic cells. The genetic regions identified may be particularly important because differential regulation of class II major histocompatibility complex protein synthesis by Langerhans cells and dendritic cells may be crucial to immune functions of intact animals.

Isolation, detection, and amplification of intact mRNA from dermatome strips, epidermal sheets, and sorted epidermal cells.

Three different strategies for isolating RNA from epidermal cells were compared. Starting with dermatome sections frozen or disaggregated epidermal cells purified by fluorescence activated cell sorting (FACS), RNA was isolated with a guanidinium thiocyanate technique. Specific mRNA were detected by Northern blot analysis (involucrin, keratin 5, actin), or by reverse transcription and amplification with the polymerase chain reaction (PCR), using primers specific for keratinocyte products (keratins 1 and 14) and Langerhans cells (CD1a). Messenger RNA's characteristic of Langerhans cells and of keratinocytes at different stages of differentiation were detected in dermatome and epidermal sheet preparations as well as in FACS-separated cells. The use of snap-frozen dermatome sections allows the isolation of RNA from epidermis that has undergone minimal trauma and is very close to its in vivo state, but that includes RNA from some dermal cells. Extraction of RNA from Dispase-separated sheets involves slightly more manipulation of the epidermis but provides a sample free from dermal contaminants. PCR analysis of sorted epidermal cells is both sensitive and specific, but involves still greater manipulation. This final technique, however, allows the investigation of mRNA produced by small groups of epidermal cells that are still much closer to their in vivo state than if they had been cultured. By combining these techniques it is possible to determine the baseline production of specific mRNA in the skin in vivo and to assign their production to specific groups of cells with a sensitivity and specificity greater than any approach previously described.

Malignant and normal T cells show random use of T-cell receptor alpha chain variable regions in patients with cutaneous T-cell lymphoma.

Cutaneous T-cell lymphoma (CTCL) is a malignancy of mature T lymphocytes, most of which express alpha/beta type T-cell receptors (TCRs). The cause of CTCL is unknown, but hypotheses postulating chronic stimulation of TCRs by superantigen or by a leukemogenic virus have been proposed. Either mechanism might produce bias in the TCR variable (V) region types used by the malignant cells. To determine if TCR alpha use is restricted in CTCL, we used reverse transcription and the polymerase chain reaction to determine V alpha and V beta usage by malignant cells purified from the peripheral blood of leukemic patients with CTCL. Usage of alpha chain V region segments appeared totally random; malignant lymphocytes isolated from each of six patients used different V alpha regions. As has been previously reported, no bias was found in beta chain V region usage either. In addition to productive (in frame) TCR V region mRNAs in malignant cells from each patient, we detected non-productive (out of frame) beta chain transcripts in these cells in two of six patients, and non-productive alpha chain transcripts in five of six. Residual normal peripheral blood lymphocytes from these patients showed a random, polyclonal or oligoclonal pattern of V region usage. We conclude that there is no bias in V region usage in CTCL, making it unlikely that interactions between superantigen or virus and the TCR V regions play a role in the pathogenesis of CTCL.

Staining properties of aldehyde fuchsin analogs.

This investigation was designed to clarify the role of the aldehyde component of aldehyde fuchsin in its staining reactions. Several aldehyde fuchsin analogs were prepared by using different aldehydes. The staining quality of these analogs and pararosaniline-HCl was compared with that of aldehyde fuchsin prepared with paraldehyde in the usual way. The major findings of this investigation include: 1) Aldehyde fuchsin staining of nonoxidized pancreatic B cells requires a stain prepared with either paraldehyde or acetaldehyde. 2) An aldehyde moiety is required for aldehyde fuchsin staining of strong tissue anions. 3) Staining of elastic tissue with aldehyde fuchsin analogs resembles staining of strong tissue anions more than staining of nonoxidized pancreatic B cells. Possible reaction mechanisms of aldehyde fuchsin with tissue substrates are discussed.

The role of interleukin-2 (IL-2) in the specific unresponsiveness of lepromatous leprosy to Mycobacterium leprae: studies in vitro and in vivo.

The role of IL-2 in the immunological deficiency of lepromatous leprosy patients towards Mycobacterium leprae have been studied further. After initial stimulation with M. leprae + IL-2, lepromatous lymphocytes could be restimulated with M. leprae alone. The specificity of the responses obtained varied. Some patients gave a stronger response to BCG as compared to M. leprae, while in others a stronger response to M. leprae as compared to BCG was obtained. Studies of the composition of lymphocytes in dermal infiltrates subsequent to injection of killed M. leprae revealed that in both tuberculoid and lepromatous patients, early accumulation of cell staining for both IL-2 receptor and IL-2 were seen. However, with time IL-2 receptor and IL-2 staining lymphocytes diminished in lepromatous infiltrates, while these were maintained in tuberculoid lesions.

The clonotypic T cell receptor is a source of tumor-associated antigens in cutaneous T cell lymphoma.

To develop cancer vaccines for the treatment of cutaneous T cell lymphoma (CTCL), immunogenic peptides were identified by two approaches. First, through the use of "reverse immunology" the peptide sequence of the idiotypic region of the beta chain of the T cell receptor (TCR) was determined and a series of overlapping peptides synthesized and tested for CD8 T cell recognition. In two patients, the idiotypic CDR3 region provided immunogenic epitopes that were recognized in a class I-restricted fashion by autologous CD8 T cell lines. In a second strategy, peptides were isolated directly from class I MHC molecules on the CTCL surface and sequenced. A peptide with partial homology to sequences contained in the conserved variable portion of the clonotypic TCR beta chain was recognized as immunogenic by autologous CD8 T cells. Therefore, both approaches demonstrated that the clonotypic TCR in CTCL is a source of immunogenic tumor epitopes. To confirm that recognition of TCR-derived sequences provides immunoprotection against tumor growth, a murine model of T cell lymphoma was studied. The immunogenicity of a thymoma, which lacks cell surface TCR expression, was enhanced by transfection of the beta chain of the TCR. The studies reviewed in this paper demonstrate that the TCR can serve as one source for immunogenic tumor peptides in T cell lymphoma in vitro and in vivo. Presentation of TCR epitopes on dendritic cells that express high levels of MHC, costimulatory, and adhesion molecules may provide an effective means for immunization against T cell malignancy.

Clinical and histologic features of pityriasis lichenoides et varioliformis acuta in children.

Pityriasis lichenoides et varioliformis acuta (PLEVA) is commonly thought of as a disease of young adults, yet we identified five cases, involving patients who were 3, 5, 6, 8, and 11 years of age, among 13,000 consecutive specimens submitted to a general dermatopathology laboratory during a 15-week period. The clinical and histologic features of PLEVA in our cases were similar to those reported for adults, except that no lesions were observed on the scalp or mucous membranes of children. A high index of suspicion and biopsy specimens of suspected lesions are often needed to differentiate PLEVA from other papular and crusted eruptions seen in the pediatric age group. These include reactions to arthropods, Gianotti-Crosti syndrome, varicella, and erythema multiforme. Histologically, papular eczema and pityriasis rosea may be misdiagnosed as PLEVA.

Hydatidiform mole: clinicopathologic associations with the development of postevacuation trophoblastic disease.

Clinical information and histopathologic material for 165 patients with hydatidiform mole referred to the John I. Brewer Trophoblastic Disease Center of Northwestern University Medical School during one year were reviewed in order to identify characteristics more likely to be associated with the development of gestational trophoblastic tumors. Twenty-nine patients (18%) required chemotherapy for invasive mole or choriocarcinoma. Patients with uterine enlargement beyond that expected for dates and patients with ovarian theca-lutein cysts were much more likely to require treatment after molar evacuation (47% vs. 18% and 40% vs. 16%, respectively). There was no correlation between the initial human chorionic gonadotropin level, gestational age, uterine size per se, maternal age or gravidity and the subsequent clinical course. Histologically, the following factors were associated with an increased incidence of postmolar gestational trophoblastic tumor: (1) progressive nuclear atypia (26.7% if atypia present vs. 40% if absent); (2) necrosis and hemorrhage (39.1% if extensive vs. 12.8% if limited); (3) decreased trophoblast maturation (48% if less than 20% mature vs. 8.7% if greater than or equal to 20% mature); (4) trophoblast proliferation (50% if marked vs. 13.9% if limited); (5) increased ratio of cytotrophoblast to syncytium (33.3% if greater than 1 vs. 6.4% if less than 1); and (6) absence of Nitabuch's layer (21.4% if absent vs. 11.6% if present). Hydatidiform moles which demonstrate clinical or histopathologic evidence of excessively abnormal proliferative activity, as indicated by these features, are more likely to develop invasive mole or choriocarcinoma and should be considered for prophylactic chemotherapy.

[In situ transcription and determination of mRNA in human thymic antigen 1a positive cells within epidermis].

This paper introduces a method of study on the characteristics of cell surface protein. We used human thymic antigen 1a anti-sense oligonucleotide primer, radiolabeled deoxynucleotide and in situ transcription with cells. The results of in situ transcription and determination of mRNA in CD1a positive cells isolated from the epidermis of the prepuce of normal neonate showed that most of the positive cells contained specific CD1a mRNA, and the specificity and sensitivity of this method were higher and the procedure was relatively simple.

Implementation of pharmaceutical care in acute medical cardiovascular patients.

All inpatients who were admitted to one designated cardiologist at a private community hospital were followed by a pharmacist. The pharmacist prospectively evaluated the patients' medications in order to identify, resolve, and prevent any medication related problems. Recommendations concerning these medication related problems were made to the physician involved. In addition, the pharmacist documented medication information questions, medication dosing consultations, and patient medication counseling consults. Ninety-seven percent of the recommendations were accepted by the prescriber. The most common categories of recommendations were for drugs belonging to cardiovascular (40.3%) and anti-infective (18.4%) classes. Improper medication selection (19.6%), untreated indication (17.4%), overdosage (16.3%), and medication given without an indication (13%) were the most common medication related problems and accounted for over two-thirds of the total accepted recommendations. Sixty-six percent of the recommendations were considered significant and 4% were considered extremely significant. Based on the pharmacist's interventions, an annual patient medication charge savings of $17,576.00 was estimated.