No abstract

Background. An early observation suggests that children older than 6 years of age at diagnosis of neuroblastoma constitute a favorable prognostic group. Methods. Kaplan‐Meier plots of survival of all such patients diagnosed at the Children's Hospital of Pittsburgh 1975–1992 were compared with curves of concurrently treated patients with Stage IV disease who were 1‐6 years of age at diagnosis (“younger patients”). Known prognostic features, including stage and primary site of disease, pattern of metastases, histopathology, MYCN gene amplification, and urinary catecholamine metabolite ratios, were reviewed. Results. Of 17 children diagnosed after the age of 6 years (“older patients”), 13 patients had Evans' Stage IV disease and 4 had Stage III disease. The median survival was 3.24 years (range, 0.63–15.04 years) for the entire cohort and 3.07 years for those children with Stage IV disease. This compared with a median survival of 1.05 years in 34 concurrent younger patients (P < 0.01). In most cases, disease in these older patients was characterized by a short‐lived complete or partial remission followed by aggressive recurrent disease that was partially and only transiently chemo‐ or radiosensitive. Only 3 patients (2 with Stage IV disease) are in continuous complete remission at 3, 510/12, and 141/2 years from diagnosis. Although poor prognostic factors were common, including the presence of bony metastases (12/17), biopsy material from pretreatment tumor specimens demonstrated a single MYCN gene copy number in all patients and favorable histology in 15 of 16 samples. Conclusion. Older children with neuroblastoma have a more indolent course than do younger patients, a finding that appears to be related to favorable histology and the absence of MYCN amplification. Examination of larger numbers of such patients from cooperative groups should lead to a better understanding of what appears to be a subset of pediatric patients with neuroblastoma who may benefit from specifically tailored treatment protoc.

Purified splenic and thymic lymphocytes from the ACI and F344 strains of inbred rats were disrupted by controlled hypotonic treatment, and their plasma membranes were prepared by sucrose density gradient centrifugation. The plasma membrane preparations were highly purified as judged by the structural appearance of the smooth membrane vesicles, by the 10- to 15-fold enrichment of 5'-nucleotidase, which cytochemically localized exclusively in the plasma membranes of intact lymphocytes, by the high cholesterol to phospholipid molar ratio (0.7-1.0), and by the very low specific activities of the enzymes associated predominantly with mitochondria, lysosomes, and endoplasmic reticulum. The protein and the lipid contents of the membranes were 48-55 and 37-48%, respectively. The total lipid content of plasma membranes was characteristically higher in thymic than splenic lymphocytes from both ACI and F344 strains. The specific activity of 5'-nucleotidase was similar in splenic lymphocyte membranes of the ACI strain, and in both the thymic and splenic lymphocyte membranes of the F344 strain. In contrast, the thymic lymphocyte membranes in the ACI strain showed half as much 5'-nucleotidase specific activity. Cytochemical results indicated that the 5'-nucleotidase is located on the outside surface of the lymphocyte plasma membranes.

The acceptance of the fetal allograft in the face of a completely competent maternal immune system remains a perplexing and critical problem in transplantation immunology . Varieties of potential mechanisms by which the fetus could avoid rejection have been proposed by Medawar (1) and by Billingham (2), and these proposals set the stage for much of the work in this field. The mechanisms proposed to prevent immune destruction of the fetus by the mother have postulated roles for blocking antibodies, immunosuppressive antibody-antigen complexes, nonimmunoglobulin immunosuppressive molecules, suppressor cells, and selective maternal immunoincompetence (reviewed in references 3-8). Many of these arguments turn on the question of what type of transplantation antigens encoded by the MHC are present in the placenta. Class I antigens have been demonstrated in the placenta of the mouse (9-13), human (14-21), and rat (22-24) using alloantisera and mAbs. Other studies in the human showed that the placenta also carries unique class I, or class I-like, antigens, the TLX antigens (25, 26), and class I molecules that carry only broadly crossreactive public antigenic determinants (27). Class II antigens are absent from the placenta in both the mouse (9, 28) and the human (14)(15)(16)(17)(18)(19) 29) .Studies in the rat demonstrated that during pregnancy the maternal strain in certain mating combinations made an antibody response to the paternal component of the fetal histocompatibility antigens without any prior sensitization (22, 30, 31) . The most potent response occurred when the u X a (female X male) haplotypes were mated. The alloantibody (22) and mAb (30) responses were directed against a placental class I molecule, the pregnancy-associated (Pa)' antigen, that has a broadly shared public antigenic determinant but not the allele-specific determinant of a classical class I transplantation antigen . The Pa antigen isolated from lymphocytes has a heavy chain of 46 kD and is associated with the ,B2-microglobulin molecule. It appears to be controlled by a diallelic system : the a, d, f, b, and m haplotypes carry the Pa antigen, whereas the n, c, l, u, g, k, and h haplotypes do not . Using alloantisera, both the Pa antigen and the allele-specific RT 1 .Aa antigen were shown to be in the basal, but not in the This work was supported by grants HD-08662 and HD-09980 from the National Institutes of Health.' Abbreviation used in this paper: Pa, pregnancy-associated.J. Exp. MED.