Morphologic analyses of positive peritoneal cytology in endometrial carcinoma.

OBJECTIVE: To investigate the relationship between the morphologic features of endometrial adenocarcinoma cells in peritoneal fluids (effusions and washings) and macroscopic intraabdominal adenocarcinoma at laparotomy as well as prognosis. STUDY DESIGN: Seventy-one patients with endometrial adenocarcinoma who showed positive peritoneal cytology at laparotomy were clinically divided into three groups: 25 patients with macroscopic neoplastic seeding in the peritoneal cavity (type 1), 38 patients without macroscopic peritoneal metastasis who survived with no evidence of disease (type 2) and 8 patients without macroscopic peritoneal metastasis who later developed recurrence of adenocarcinoma (type 3). Morphologic features of the adenocarcinoma cells in smears of peritoneal fluids were examined. RESULTS: Most of the smears from type 1 patients showed moderate to high cellularity, scalloped edges of cell clusters and isolated adenocarcinoma cells, whereas these features were seldom observed in type 2 patients. Although not all type 3 patients demonstrated these three features, patients in the series whose specimens exhibited none of the three features did not show any peritoneal lesions or have a recurrence of their disease. CONCLUSION: The finding of endometrial adenocarcinoma cells exhibiting high cellularity, scalloped edge of cell clusters and isolated cells in smears of peritoneal fluid is associated with the presence of intraabdominal macroscopic metastatic lesions and could be regarded as a risk factor for intraabdominal recurrence of carcinoma.     

An approach to diagnosis of T-cell lymphoproliferative disorders by flow cytometry

T-cell lymphoproliferative disorders are among the most challenging diagnoses in hematopathology. Unlike the more common B-cell disorders, in which clonality is often readily discernible by surface immunoglobulin light chain restriction, there is no specific immunophenotypic signature that is diagnostic of a clonal T-cell population. Immunophenotypic criteria that are helpful in the diagnosis of T-cell neoplasms include T-cell subset antigen restriction, anomalous T-cell subset antigen expression, deletion or diminution of one of the pan T-cell antigens, a precursor T-cell phenotype, and expression of additional markers (e.g., CD30, CD20, major myeloid antigens, and TCRgammadelta). Analysis of the inherent forward and orthogonal light scatter properties of the cell can also provide important diagnostic clues. None of these features is 100% specific, however, for aberrant expression of pan-T antigens may be seen in viral infections, B-cell malignancies, or in reactive changes following administration of certain medications. An increased CD4:CD8 ratio is often observed in Hodgkin's lymphoma. Based on the analysis of 87 neoplastic and 80 control cases, we conclude that flow cytometric features that are most suspicious for malignancy include the loss or markedly dim expression of CD45; complete loss of one or more pan-T antigens; diminished expression of more than two pan-T antigens in conjunction with altered light scatter properties; and CD4/CD8 dual-positive or dual-negative expression (except thymic lesions).     

Anaplastic large cell lymphoma presenting as a pleural effusion and mimicking primary effusion lymphoma. A report of 2 cases

BACKGROUND: Systemic anaplastic large cell lymphoma (ALCL) is predominantly a nodal disease, but extranodal involvement can occur during the disease course or as the primary presentation. We report two rare cases of ALCL presenting with a pleural effusion, mimicking primary effusion lymphoma (PEL). CASES: Two patients, a 47-year-old woman and an 81-year-old man, presented with a pleural effusion for investigation. The pleural fluid contained abundant, large, lymphoid cells with marked nuclear atypia. These neoplastic cells strongly expressed CD30 and EMA and showed a T-cell phenotype (CD3+CD45RO+ for case 1 and CD4+ for case 2). Case 1, in addition, showed ALK1 expression. The tumor cells in both cases were negative for human herpes virus type 8 (HHV8) and Epstein-Barr virus (EBV). ALCL shows overlapping cytologic features with PEL, but the T-cell phenotype, ALK1 expression in case 1, lack of association with HHV8 and EBV, HIV seronegativity and subsequent discovery of nodal disease in case 2 were all in favor of ALCL over PEL. CONCLUSION: In rare cases a pleural effusion is the presenting feature of ALCL, and distinction from PEL depends on correlation with clinical findings, detailed immunophenotyping and study of the status of HHV8 and EBV.     

Cytologic features of NK/T-cell lymphoma

OBJECTIVE: To describe the diagnostic cytologic features of NK/T-cell lymphoma. STUDY DESIGN: The cytologic features of 3 cases of natural killer cell (NK)/T-cell lymphoma were studied and correlated with histology. Immunohistochemistry for CD56, T-cell intracellular antigen (TIA-1) and EBV-encoded small nuclear RNAs (EBER) in situ hybridization was reviewed. RESULTS: The lymphomas have mixed-sized cells with eccentric, round to ovoid nuclei; 1 or 2 prominent nucleoli; and abundant, clear to pale, eosinophilic cytoplasm. Mitotic figures, necrotic debris and tingible body macrophages are common accompaniments. In fluid, the lymphoma cells appear more shrunken. A clot section of 1 case was positive for CD56, TIA-1 and EBER. CONCLUSION: Helpful cytologic features for the diagnosis of NK/T-cell lymphoma are described. Immunohistochemistry for CD56, TIA-1 and EBER in situ hybridization are very helpful adjuncts for the diagnosis.     

Peripheral T-cell lymphoma not otherwise specified vs. Hodgkin's lymphoma on fine needle aspiration cytology

OBJECTIVE: To study the morphologic features of peripheral T-cell lymphoma, not otherwise specified (PTCL, NOS) on fine needle aspiration cytology (FNAC) smears and to identify cytomorphologic features that would delineate them from features of cases of Hodgkin's lymphoma (HL). CONCLUSION: Fourteen cases each of PTCL, NOS, and HL with adequate FNAC smears were retrieved. These cases were analyzed for 12 cytomorphologic features: presence of atypical lymphoid cells, percentage of large lymphoid cells, lymphoid cells with cleaved/indented nuclei, typical Reed-Sternberg (RS) cells, mononuclear cells with prominent eosinophilic macronucleoli, mononuclear cells with prominent nucleoli, histiocytes, eosinophils, plasma cells, polymorphonuclear leukocytes, vessels and perivascular clustering of atypical mononuclear/atypical lymphoid cells. Each of these features was evaluated for its presence or absence and was semiquantitated on a scale from 0 to +3. RESULTS: The presence of atypical lymphoid cells with a spectrum ranging from small to intermediate and large was seen exclusively in cases of PTCL. Lymphoid cells with cleaved or indented nuclei, endothelium-lined vessels with perivascular clustering of tumor cells and absence of typical RS cells, and mononuclear cells with prominent eosinophilic macronucleoli emerged as the parameters significant in not only diagnosing cases of PTCL, NOS, but also in their delineation from cases of HL. CONCLUSION: A careful analysis of cytomorphologic features can be useful in at least suggesting a diagnosis of PTCL and help to distinguish that diagnosis from HL, which the features may mimic. Immunophenotyping and molecular studies are important in arriving at a definitive diagnosis.     

Diagnostic considerations in prolymphocytes in pleural fluid: a case report

BACKGROUND: Prolymphocytes are nucleolated cells that are the defining features of the 2 chronic lymphoproliferative disorders, prolymphocytic leukemia (PLL) and chronic lymphocytic leukemia (CLL) with increased prolymphocytes. Prolymphocytes appear relatively unfamiliar in cytopathology practice, and, particularly when present in body fluids, may resemble blasts or adult T-cell leukemia/ lymphoma (ATLL) cells. CASE: A 32-year-old man, referred to us with a diagnosis of acute leukemia, presented with shortness of breath for 2 months and loss of appetite for 3 months. He had enlarged liver and spleen, 6 and 8 cm, respectively, below the costal margin and pleural effusion. The raised total leukocyte count chiefly comprised prolymphocytes that, especially in the pleural fluid, had prominent nucleoli and significant pleomorphism, raising the possibility of blasts or ATLL. CONCLUSION: Prolymphocytes in body fluids can be misinterpreted as blasts or even ATLL cells. Better awareness among cytopathologists about prolymphocytes and the disease states in which they occur, as well as insistence, in a clinical setting of leukemia, on interpreting the pleural fluid in relation to the clinical and laboratory findings, especially those of the peripheral blood and bone marrow, can prevent misdiagnosis. Equally importantly, immunophenotyping must be done in such situations.     

AIDS-related body cavity-based lymphoma. A case report

BACKGROUND: Body cavity-based lymphomas are rare malignancies in human immunodeficiency virus (HIV)-infected patients, but because of their unusual clinical, morphologic and immunophenotypic features, they are recognized as a distinct subgroup of lymphomas connected to human herpesvirus 8 (HHV-8) infection. CASE: A 39-year-old, HIV-positive, homosexual man was admitted to the hospital because of a left-sided pleural effusion that contained malignant lymphoid cells. He responded partially to a low-dose cyclophosphamide/doxorubycin/vincristine/prednisone regimen and died five months after the diagnosis of lymphoma. On cytology, the sediments contained exclusively large, round, neoplastic, lymphoid cells with abundant basophilic cytoplasm and large, round nuclei with prominent nucleoli. Many cells had immunoblastic features, and some had plasmocytoid differentiation. Mitotic figures were numerous. On flow cytometry, the homogeneous population of large cells expressed CD45, CD38, HLA-DR and CD7 positivity. Other specific T-, B- and NK-cell markers tested negative. Polymerase chain reaction demonstrated Epstein-Barr virus (EBV) and HHV-8 in the malignant effusion. CONCLUSION: Primary effusion from lymphoma with molecular evidence of HHV-8 and EBV coinfection represents a distinct clinical and morphologic entity in AIDS patients. However, immunophenotypic markers of malignant clones can be diverse in different cases.     

A possible new syndrome with growth-hormone secreting pituitary adenoma, colonic polyposis, lipomatosis, lentigines and renal carcinoma in association with familial testicular germ cell malignancy: A case report

Introduction

Cytofluorographic and molecular techniques are effective adjuncts in diagnosing intraocular lymphoma. Primary intraocular lymphoma is an uncommon entity predominantly of B cell origin and rarely with a T cell phenotype. The aim of the present paper is to report a case of a CD8-positive, TCR-α/β-negative intraocular T cell lymphoma and review the literature.

Case presentation

T cell neoplasia was detected based on flow cytometric demonstration of an abnormal T cell population and polymerase chain reactions for immunoglobulin and T-cell receptor rearrangements demonstrating evidence of monoclonality. Flow cytometry revealed a T cell population aberrantly expressing T-cell lineage markers. This T cell population expressed CD2, bright CD3, CD8, bright CD7, CD38, CD69, and variable CD25. T-cell receptor γ gene rearrangement studies demonstrated evidence of T-cell gene rearrangement confirming that the T cells were monoclonal.

Conclusion

We herein report the rare case of a TCR α/β-negative CD8+ intraocular T-cell lymphoma suggestive of gamma/delta origin diagnosed by flow cytometry and polymerase chain reaction.     

Fine needle aspiration of lymphoblastic lymphoma. A multiparameter diagnostic approach.

Sixteen fine needle aspiration (FNA) biopsies of lymphoblastic lymphoma (LBL) that were used to either initially diagnose disease (12) or document relapse (4) were reviewed. Cellular aspirates (2 x 10(7) cells) were readily obtained for immunologic, DNA/RNA flow cytometric and immunoglobulin and/or T-cell receptor gene rearrangement studies. Cytologically, aspirates were characterized by intermediate-sized cells (9.5-18.5 microns) with fine nuclear chromatin, small, inconspicuous nucleoli, irregular nuclear contours and scant basophilic cytoplasm. Frequent mitotic figures were seen (1-14 figures per 1,000 cells). Fourteen cases demonstrated a T-cell phenotype with considerable phenotypic variability. One case demonstrated a precursor B-cell phenotype, and another demonstrated biphenotypic expression with both T-cell and myeloid differentiation. Eleven of 14 cases (79%) were positive for terminal deoxynucleotidyl transferase. Thirteen of 15 cases (87%) manifested diploid DNA content by flow cytometric analysis and were characterized by intermediate proliferative activity (S+G2M 12.7 +/- 8.7% SD) and intermediate mean RNA index (1.3 +/- .5 SD). T beta gene rearrangements were demonstrated in four of five phenotypic T-cell LBL cases analyzed, with concomitant JH gene rearrangements observed in three cases, confirming that bigenotypic rearrangements characterize some T-cell LBLs. We conclude that FNA samples are adequate for accurate characterization of LBL and may obviate the need for surgical biopsy.     

Qualitative and quantitative analysis of peritoneal fluids from women with gynecologic diseases. Comparison of cytology and flow cytometry for the detection of malignancy in lavage and ascitic fluids

A prospective study was undertaken to compare flow cytometric (FCM) analysis to conventional cytologic evaluation for the detection of malignant cells in peritoneal fluids (peritoneal lavages and ascitic fluids) from women with gynecologic diseases. The 94 peritoneal fluids analyzed came from 63 cancer patients (with epithelial ovarian carcinomas) and 31 control patients (with benign gynecologic diseases). The FCM DNA histograms were generated using propidium iodide as a DNA fluorochrome. Samples for cytologic analysis were stained with the standard May-Grünwald-Giemsa or Papanicolaou stains. Of the 94 samples, 90 were evaluable cytologically while 70 were suitable for FCM analysis. The sensitivities were 55% for FCM DNA analysis and 80% for cytologic analysis. FCM DNA analysis had a 30% false-positive rate; cytologic analysis produced no false-positive results. These results indicate that there is no advantage in employing FCM analysis instead of conventional cytologic evaluation for the detection of malignant cells in peritoneal fluids from gynecologic cases.     

Diagnostic accuracy of the immunocytochemical study of body fluids

The diagnostic accuracy of the immunocytochemical characterization of body fluids was evaluated in 100 specimens (35 pleural, 40 peritoneal, 7 pericardial and 18 cerebrospinal [CSF] fluids) in comparison with routine morphologic examination. The immunochemical markers used for all specimens were common-leukocyte antigen, epithelial membrane antigen, epithelial keratin and desmin. Additional immunocytochemical studies for neurofilaments, glial fibrillary acidic protein, vimentin and melanoma-associated antigen were performed on the CSF specimens. The study confirmed the accuracy of the immunocytochemical characterization of cells in body fluids using a panel of immunocytochemical stains. These methods are recommended as an adjunct to improve the accuracy of the cytologic diagnosis of body fluids, especially in cases with diagnostically difficult morphologic features.     

Contribution of flow cytometric immunophenotyping to the evaluation of tissues with suspected lymphoma?

BACKGROUND: A critical analysis of the contribution of flow cytometric immunophenotyping (FCI) to the evaluation of lymph nodes and extranodal tissues with suspected lymphoma by a large, retrospective approach has not been reported previously and represents the purpose of this study. METHODS: A total of 278 lymph nodes and 95 extranodal tissue specimens submitted over a 2-year period with complete histologic, FCI, and immunohistochemical (IH) data formed the basis of the study. RESULTS: The FCI data contributed significantly to or was consistent with the final tissue diagnosis in the majority (94%) of the tissue samples. There is no well-described utility of flow cytometry markers for Hodgkin's lymphoma (HL) due to the usual scarcity of tumor cells in the final cell suspensions obtained from these tumors. However, the FCI data excluded non-Hodgkin's lymphoma (NHL) and suggested the possible usefulness of CD15 and CD30 by FCI in HL. In addition, immunophenotypic data by FCI in combination with touch imprint cytomorphology was useful in excluding a diagnosis of NHL in cases of nonhematopoietic malignancies and was particularly useful in defining the following hematopoietic tumors and malignancies: thymoma, T-cell lymphoblastic lymphoma, leukemia cutis, and plasma cell dyscrasia. Thus, IH was not essential for the diagnosis in these latter cases and was performed in only two cases (one thymoma and one plasma cell dyscrasia). Of interest, FCI supported the diagnosis in 3 cases of Ewing's sarcoma/primitive neuroectodermal tumor by detection of CD56 on the surface of the malignant cell. Only 11% of NHL were "negative" by FCI (i.e., an aberrant T-cell or monoclonal B-cell population was not identified). Reasons for these discrepancies included partial tissue involvement by the NHL with sampling differences, T-cell rich or lymphohistiocytic-rich variants with a small population of monoclonal B cells, marked tumoral sclerosis, poor tumor preservation, and T-cell NHL without an aberrant immunophenotype. Only 60% of CD30+ anaplastic large cell lymphomas (ALCL) were CD30+ by FCI. CONCLUSIONS: FCI data should always be correlated with light microscopy if no FCI abnormalities are detected; IH may need to be performed in selected cases. It is less necessary to perform microscopic examination of tissues when the FCI data are positive and indisputable. However, in selected cases in which FCI data is diagnostic, microscopic observations may provide additional information due to sampling.     

Expansions of clonal and oligoclonal T cells in B-cell chronic lymphocytic leukemia are primarily restricted to the CD3(+)CD8(+) T-cell population

B-cell chronic lymphocytic leukemia (B-CLL) is characterized by the accumulation of mature-appearing clonal B cells exhibiting coexpression of CD5 and CD23. In addition to the accumulation of neoplastic B cells, numerous T-cell abnormalities also occur in B-CLL patients. In this study, the presence, and distribution within the T-cell subsets, of clonal/oligoclonal T cells was studied. Multicolor flow cytometric techniques were employed using combinations of anti-CD3, anti-CD4, and anti-CD8 antibodies coupled with antibodies specific for V(alpha) and V(beta) T-cell receptor (TCR) epitopes. Molecular studies of TCR gene sequences were done to confirm the presence of clonal/oligoclonal T-cell populations. In the flow cytometric studies, examination of V(alpha)/V(beta)expression found evidence of clonal/oligoclonal expansion in 9 of 19 patients studied. In eight of the nine patients, the expansions were restricted to the CD3(+)CD8(+) cell population. Molecular analyses were performed in 16 patients, 12 of whom showed a clonal or oligoclonal pattern. Of the four patients who were negative in the molecular analyses, all demonstrated flow cytometric evidence of clonal/oligoclonal expansions. Thus, when the flow cytometric and molecular analyses were considered together, all 16 patients for whom parallel analyses were done showed evidence of clonal/oligoclonal expansions. These results confirm previous work demonstrating that the majority of B-CLL patients harbor clonal/oligoclonal expansions within the T-cell population. Additionally, based on the relative numbers of cells expressing specific V(alpha) or V(beta)epitopes, these results show that these expansions occur primarily within the CD3(+)CD8(+) T-cell population.     

Flow cytometric analysis and cytopathology of body cavity fluids

A total of 75 samples of body cavity fluids from 71 patients were analyzed by both flow cytometry (FCM), to detect cells with an abnormal DNA content (aneuploidy), and by conventional cytopathology. Samples included 27 pleural fluids, 35 peritoneal fluids, 11 peritoneal washings and 2 pericardial fluids. For cytologic examination, the samples were prepared using standard techniques. Samples for FCM analysis were centrifuged and exposed to a hypotonic solution containing detergent and propidium iodide, a DNA intercalating fluorescent stain. Aneuploidy as well as cytologic malignancy were found in 17 samples. Forty-seven samples had normal DNA histograms by FCM and were also cytologically negative. Four samples suspicious by cytology but normal by FCM were from patients with renal-cell carcinoma (two samples from the same patient), endometrial adenocarcinoma without metastasis and chronic lymphocytic leukemia. Three samples abnormal by FCM but negative by cytology were from patients with ovarian cystadenoma, cirrhosis and uterine leiomyoma. FCM showed aneuploidy in four cytologically negative samples from patients with histologically proven malignancy (lymphoma, colonic adenocarcinoma, cervical squamous cell carcinoma, and endometrial adenosquamous carcinoma). Based on these results, FCM analysis combined with conventional cytopathology yielded 100% sensitivity, 100% predictive value of a negative result and 94% specificity. This rapid and quantitative FCM analysis of body cavity fluids can be a very useful adjunct to conventional diagnostic cytopathology.     

Pleural effusion cytology in a case of cytophagic histiocytic panniculitis (subcutaneous panniculitic T-cell lymphoma). A case report

BACKGROUND: Cytophagic histiocytic panniculitis (CHP) presents with subcutaneous panniculitis associated with hemophagocytic syndrome. Many cases of CHP are now being classified as a natural disease progression of subcutaneous panniculitic T-cell lymphoma (SPTL). There have been no cytologic reports dealing with pleural aspirates in cases of CHP or SPTL. CASE: A pleural aspirate obtained from a 19-year-old female revealed lymphoma cells and hemophagocytic histiocytes. A skin biopsy specimen showed the presence of CD8-positive lymphoma cells in fat lobules associated with cytologically benign histiocytes with erythrophagocytosis and lymphophagocytosis. CONCLUSION: Hemophagocytic histiocytes were seen in the pleural effusion from a patient with SPTL.     

Lymphoepithelioid cell lymphoma (Lennert's lymphoma). Report of a case with fine needle aspiration cytology

BACKGROUND: Lymphoepithelioid cell lymphoma (LCL) is a rare morphologic variant of peripheral T-cell lymphoma. Although their histopathologic and immunohistochemical findings are well known, the cytopathologic features have not been well documented. This report describes the fine needle aspiration cytology (FNAC) findings of a case of LCL. CASE: A 75-year-old woman presented with cervical, supraclavicular, axillary and mediastinal lymphadenopathy. FNAC of a cervical lymph node was performed. The smears contained a polymorphous infiltrate formed by abundant histiocytes disposed singly or in clusters, small and medium-sized to large atypical lymphoid cells and reactive cells, including eosinophils and plasma cells. Isolated capillary-sized vessels also were observed. Histopathologic and immunohistochemical examination confirmed the diagnosis of Lennert's lymphoma. CONCLUSION: Although histopathologic and immunohistochemical studies were required for a definitive diagnosis, the findings of FNAC in this case appeared distinctive and suggested the possibility of LCL.     

A method to average immunofluorescent histograms

Single parameter gated flow cytometric fluorescent histograms were obtained on normal donor blood mononuclear cells using several commonly available lymphocyte surface markers. A computer method was developed to average single parameter flow cytometric immunofluorescent histograms. The averaged histograms provide a means of pattern recognition for normal lymphocytes and will aid in the clinical evaluation of lymphocytosis. Averaged histograms may also serve as standards for more advanced analysis.     

Cytologic and immunocytologic studies of peripheral T-cell lymphomas

Lymph node aspirates from 18 peripheral T-cell lymphomas (PTLs) were analyzed. Cytologic and immunocytologic studies were performed on Cytospin preparations using the alkaline phosphatase-antialkaline phosphatase method with a panel of monoclonal antibodies (CD3, CD4, CD8, CD19 and CD30). The cytologic diagnosis was confirmed by histologic investigation. Nine lymph node aspirates from patients with Lennert's lymphoma, angioimmunoblastic (AILD)-type PTL and pleomorphic small-cell-type PTL were composed predominantly of small-to-intermediate-sized lymphocytes. An admixture of plasma cells, eosinophils, neutrophils, lymphocytes with an irregular nucleus, granula in the cytoplasm or abundant cytoplasm was also seen. Nine lymph node aspirates from patients with T-immunoblastic lymphoma, pleomorphic large-cell-type PTL and large-cell anaplastic (Ki-1+) lymphoma showed marked cytologic heterogeneity. Immunocytologic investigation of the aspirates using the antibodies CD3, CD4, CD8, CD19 and CD30 was helpful for the differentiation of PTLs from reactive lymphadenopathy and other malignant lymphomas. A strong predominance of CD3+ cells was found in only seven cases. The aspirates expressed a helper/inducer phenotype in 11 cases and a suppressor/cytotoxic phenotype in 4 cases. A T-cell phenotype not corresponding to the normal T-cell phenotype was found in nine cases. In 15 of the 18 cases, the number of CD19+ cells was found to be less than 15%. The large cells of the large-cell anaplastic (Ki-1+) lymphoma expressed the antigens CD30 and CD45 and were negative for CD15. These findings indicate that immunocytologic studies can be used in improving the cytologic diagnosis of PTLs.     

Primary mediastinal large B-cell lymphoma in HIV: report of two cases

Primary mediastinal large B cell lymphoma (PMLBCL) is a subtype of diffuse large B cell lymphoma arising in the mediastinum with distinctive clinical and morphological features. Though diffuse large B cell lymphoma is one of the most common non-Hodgkin lymphoma associated with AIDS, there are no data available regarding the association of HIV and PMLBCL. We report here two cases of PMLBCL arising in AIDS patients. In both cases, PMLBCL presented in a setting of low CD4 T-cell count as rapidly enlarging mediastinal mass. The morphologic and immunophenotypic findings are characteristic of PMLBCL. One of the two patients, a 25-year-old woman who had localized disease and evidence of Epstein–Barr virus in lymphoma cells, did not respond to chemotherapy and died of disease progression 5 months after diagnosis. The second patient, a 38-year-old male with disseminated disease, responded to therapy and is disease-free after 9 months of follow-up.