Logistic regression analysis of small round cell neoplasms: A cytologic study

Abstract

The diagnosis of small round cell neoplasms (SRCN) Key Words: small round cell neoplasms; FNA; logistic regression analysis Small round cell neoplasms (SRCN) are a group of morphologically similar neoplasms of widely differing histogenesis. These neoplasms are composed of sheets of uniform small round cells which can defy classification by light microscopy even when such examination is supplemented by immunohistochemistry and ultramicroscopy. 1 Differential diagnosis of these malignancies is influenced by the age of the patient, their site of occurrence, and slight degrees of differentiation exemplified by minor architectural and cytoplasmic features. The problem of differential diagnosis is compounded when these neoplasms are examined by cytologic methods, most particularly fine-needle aspiration (FNA) cytology. Numerous case reports and series have described the morphologic findings in small round cell neoplasms of the pediatric age group category, 2-31 but few reports 31, 32 have investigated the diagnostic value of such findings or directly compared the incidence of these cytologic features in the various neoplasms.To better define the differential diagnostic value of the cytomorphologic features described within the literature, we performed a logistic regression analysis of these features for the diagnosis of the more common SRCN, exclusive of small-cell carcinoma. Herein we report the results of that study. Materials and MethodsThe files of the Department of Pathology at Duke University and the consultation files of one of the authors (L.J.L.) were searched for all small round cell neoplasms exclusive of small-cell carcinoma but including Ewing's sarcoma, peripheral neuroectodermal tumor, nephroblastoma (Wilms' tumor), neuroblastoma, lymphoma, rhabdoid tumor, rhabdomyosarcoma, and hepatoblastoma. Sixty-two cases were identified and were distributed as follows: Ewing's sarcoma (19 cases), rhabdomyosarcoma (6 cases), neuroblastoma (10 cases), nephroblastoma (14 cases), lymphoma (10 cases), and hepatoblastoma (3 cases). The three cases of hepatoblastoma were not analyzed in the present study. The remaining cases were analyzed for the 28 variables shown in Table I. An additional four variables comprising disassociated cells, central nuclei, cytoplasmic inclusions, and tangles were not analyzed. Disassociated cells and central nuclei were not included in the analysis because only two and three cases, respectively, were negative for these findings. Cytoplasmic inclusions and tangles were not analyzed because only one

Concordance of histopathologic and cytologic grading in musculoskeletal sarcomas

“…In contrast with adult soft tissue sarcomas, subtyping of pediatric sarcomas by FNA appears highly accurate with immunohistochemical and molecular diagnostic techniques supplying further prognostic and diagnostic information. 19,25,26 In summary, our findings support the opinion 19 that FNA can accurately subtype and grade musculoskeletal sarcoma in most cases. Unfortunately, cytologic analysis appears to undergrade a significant minority of cases reducing the overall utility of the technique when a low-grade sarcoma is diagnosed cytologically.…”

Cytological evaluation of head and neck tumors in children—A pattern analysis

“…The commonest tumor in this group in the head and neck area was a PNET/ES. Amongst the features studied, nuclear grooves, molding, 13,14 light and dark cell pattern, and cytoplasmic vacuoles [15][16][17] have been described earlier as diagnostic of PNET/ES. We found the combination of Ewingoid or pale nucleus, cytoplasmic vacuoles, and nuclear molding as reliable clues to the diagnosis of this tumor.…”

Round cell tumors: Classification and immunohistochemistry

“…[3] Fine-needle aspiration cytology (FNAC) plays an important role in the diagnosis of these tumors. [2,4,5]…”