Pediatric and Developmental Pathology

Necrotizing enterocolitis (NEC), a disease affecting predominantly premature infants, is a leading cause of morbidity and mortality in neonatal intensive care units. Although several predisposing factors have been identified, such as prematurity, enteral feeding, and infection, its pathogenesis remains elusive. In the past 20 years, we have established several animal models of NEC in rats and found several endogenous mediators, especially platelet-activating factor (PAF), which may play a pivotal role in NEC. Injection of PAF induces intestinal necrosis, and PAF antagonists prevent the bowel injury induced by bacterial endotoxin, hypoxia, or challenge with tumor necrosis factor-α (TNF) plus endotoxin in adult rats. The same is true for lesions induced by hypoxia and enteral feeding in neonatal animals. Human patients with NEC show high levels of PAF and decreased plasma PAF-acetylhydrolase, the enzyme degrading PAF. The initial event in our experimental models of NEC is probably polymorphonuclear leukocyte (PMN) activation and adhesion to venules in the intestine, which initiates a local inflammatory reaction involving proinflammatory mediators including TNF, complement, prostaglandins, and leukotriene C₄. Subsequent norepinephrine release and mesenteric vasoconstriction result in splanchnic ischemia and reperfusion. Bacterial products (e.g., endotoxin) enter the intestinal tissue during local mucosal barrier breakdown, and endotoxin synergizes with PAF to amplify the inflammation. Reactive oxygen species produced by the activated leukocytes and by intestinal epithelial xanthine oxidase may be the final pathway for tissue injury. Protective mechanisms include nitric oxide produced by the constitutive (mainly neuronal) nitric oxide synthase, and indigenous probiotics such as Bifidobacteria infantis. The former maintains intestinal perfusion and the integrity of the mucosal barrier, and the latter keep virulent bacteria in check. The development of tissue injury depends on the balance between injurious and protective mechanisms.

Synovial sarcoma is the most common nonrhabdomyosarcomatous soft tissue sarcoma in children and adolescents and is characterized by a reciprocal t(X;18)(p11;q11) which results in the fusion of the SYT gene on chromosome 18q11 to either of two closely related genes, SSX1 (Xp11.23) or SSX2 (Xp11.21). Detection of this translocation or its resultant gene fusion by molecular methods is helpful in the pathologic diagnosis of synovial sarcoma, especially in poorly differentiated tumors. This study was designed to evaluate the utility of a real-time reverse transcriptase-polymerase chain reaction (RT-PCR) assay to detect and distinguish SYT-SSX1 and SYT-SSX2 fusions in fresh and archival specimens of synovial sarcoma in pediatric patients seen at St. Jude Children's Research Hospital. In addition, the clinico-pathologic features of the tumors with SYT-SSX1 vs. SYT-SSX2 fusions were compared. The 25 patients studied had a median age of 13 years 9 months (range 5 to 19 years). Estimates of survival and event-free survival at 5 years were 78.7 ± 10.5% and 56.2 ± 13.2%, respectively. Seventeen (68%) tumors were monophasic, eight (32%) were biphasic. Seven tumors contained poorly differentiated areas. Positive results for either SYT-SSX1 or SYT-SSX2 were obtained in 21/25 (84%) cases. Three cases did not have a detectable gene fusion and one had no amplifiable RNA. SYT-SSX1 transcripts were found in 18/24 (75%) of the tumors while SYT-SSX2 transcripts were identified in 3/24 (12.5%). All of the poorly differentiated tumors and seven out of eight tumors from patients who developed lung metastases had an SYT-SSX1 fusion transcript. Real-time PCR is useful in detecting and distinguishing SYT-SSX1 from SYT-SSX2 gene fusions in synovial sarcoma. Valuable aspects of this methodology are the applicability to both frozen and formalin-fixed samples, decreased labor costs, and the rapidity of results. In addition, distinguishing SYT-SSX1 from SYT-SSX2 fusions with these methods allow for prospective collection of information that may clarify issues of prognostic relevance.

In our previous studies, we described the development of the secretory (mucosal) immune system (SIS) in human fetuses in the second trimester of pregnancy. In the present study, we examined the presence and distribution of components of this system in human embryos and early fetuses in the first trimester. An immunohistochemical study was performed on 17 embryos and 9 fetuses (4 to 12 wk of development) using antibodies against secretory component (SC), joining (J) chain, immunoglobulins (IgA, IgM, IgG), subsets of T and B lymphocytes, and macrophages. Cells positive for SC, J chain, and IgG were found in epithelial tissues from wk 4 of pregnancy. In the internal organs, such as the myocardium and endocardium, capillary endothelium, epithelium of the kidney tubules and some others, only J chain and immunoglobulins were seen. IgA was weakly reactive in tissues where SC and/or J chain were presented. IgM was very weak or absent. Among the cellular components of the SIS, only macrophages were seen in 4-wk-old embryos. CD3⁺ and CD20⁺ lymphocytes were found at wk 7 to 8. IgA- and IgM-positive lymphocytes appeared at the end of wk 9. The SIS is widespread in embryonic and early fetal periods and begins to function before the appearance of the common immune system in the developing organism. The first functional components of the SIS, such as IgG and IgA observed in this study, are most probably of maternal origin.

Pediatric small round cell tumors still pose tremendous diagnostic problems. In difficult cases, the ability to detect tumor-specific gene fusion transcripts for several of these neoplasms, including Ewing sarcoma/peripheral primitive neuroectodermal tumor (ES/PNET), synovial sarcoma (SS), alveolar rhabdomyosarcoma (ARMS), and desmoplastic small round cell tumor (DSRCT) using reverse transcriptase–polymerase chain reaction (RT-PCR), can be extremely helpful. Few studies to date, however, have systematically examined several different tumor types for the presence of multiple different fusion transcripts in order to determine the specificity and sensitivity of the RT-PCR method, and no study has addressed this issue for formalin-fixed material. The objectives of this study were to address the specificity, sensitivity, and practicality of such an assay applied strictly to formalin-fixed tissue blocks. Our results demonstrate that, for these tumors, the overall sensitivity for detecting each fusion transcript is similar to that reported in the literature for RT-PCR on fresh or formalin-fixed tissues. The specificity of the assay is very high, being essentially 100% for each primer pair when interpreting the results from visual inspection of agarose gels. However, when these same agarose gels were examined using Southern blotting, a small number of tumors also yielded reproducibly detectable weak signals for unexpected fusion products, in addition to a strong signal for the expected fusion product. Fluorescence in situ hybridization (FISH) studies in one such case indicated that a rearrangement that would account for the unexpected fusion was not present, while another case was equivocal. The overall specificity for each primer pair used in this assay ranged from 94 to 100%. Therefore, RT-PCR using formalin-fixed paraffin-embedded tissue sections can be used to detect chimeric transcripts as a reliable, highly sensitive, and highly specific diagnostic assay. However, we strongly suggest that the final interpretation of the results from this assay be viewed in light of the other features of the case, including clinical history, histology, and immunohistochemistry, by the diagnostic pathologist. Additional studies such as FISH may be useful in clarifying the nature of equivocal or unexpected results.

The surgical report is an important means of documenting normal and abnormal findings, and for distilling such information into a meaningful clinico-pathologic correlation. An audit of the quality of placental reports from four laboratories was performed using an arbitrary numerical scoring scheme that examined the gross, histologic, and commentary components of each report. The mean scores from the four laboratories were not statistically different from each other. Three (2%) and 48 (33%) of the 147 singleton placentas scored less than 50 and 75%, respectively, on this scoring scheme. None and 14 (41%) of the placentas from 34 multiple pregnancies scored less than 50 and 75%, respectively. Different aspects of the gross and histologic examination were reported variably by the laboratories. Commentaries on gross or histologic abnormalities, and in relation to clinical indications, were inconsistently reported. The standards of placental surgical reporting can be improved. The use of templates and checklists for reporting of placentas may be considered.

Lethal multiple pterygium syndrome (LMPS) is an uncommon fetal-onset disorder of unknown etiology. The pathogenesis of LMPS has been suggested to be early-onset fetal akinesia, fragile collagen, or generalized edema. Information on the neuromuscular pathology of LMPS in the literature is generally scanty. We present the findings from a review of 14 fetuses with features of LMPS from the archives of the Hammersmith Hospital Perinatal Pathology Department. Autopsy reports, photographs, fetograms, and histological sections were examined, and additional special stains and immunostaining were performed on muscle sections. In five cases, there was evidence of autosomal recessive inheritance. One case was later shown to be due to glycogen storage disease type IV. The skeletal muscle bulk was reduced in all fetuses and the remaining muscle showed a range of histological appearances including vacuolar degeneration, dystrophy, a generalized or patchy myotubular appearance, and generalized hypotrophy. In one, the histological appearance was essentially normal. Two cases had abnormalities in the brain. Large motor neurons were present in the anterior spinal horns of all fetuses in whom the spinal cord could be examined. There was no evidence of cartilaginous joint fusion. We conclude that LMPS is the phenotype resulting from fetal akinesia commencing in the first or early second trimester. In the majority of cases, the precise underlying cause will not be identified, however, occasionally a metabolic or neurodevelopmental disorder or a specific primary myopathy may be demonstrated, providing adequate autopsy investigations are undertaken.

The diagnosis of molar pregnancy is a continuing diagnostic problem for many practicing histopathologists who are required to examine specimens of products of conception, particularly since changes in gynecological management in recent years have resulted in uterine evacuation at earlier gestations. The aim of this review is to provide practical, up-to-date, diagnostically useful information regarding the histological diagnosis of molar disease in early pregnancy. Pathophysiological issues relevant to molar pregnancies, such as genetic abnormalities, will be briefly summarized, but nonhistopathological aspects of molar disease will not be covered in detail in this review.

Bone marrow transplantation (BMT) has been shown to reverse or stabilize some manifestations of mucopolysaccharidosis I (Hurler syndrome). Idiopathic hyperammonemia (IHA) is a rare complication of solid organ and BMT that is characterized by elevated serum ammonia, normal liver enzymes, and abrupt onset of neurologic deterioration. We present the case of a 14-month-old male patient with Hurler syndrome who developed fatal IHA (ammonia = 2297 μmol/L) 31 days after a cord blood transplant. A complete autopsy was performed, with examination of both frozen and formalin-fixed paraffin-embedded (FFPE) tissues using a variety of special stains and electron microscopy. Hyperammonemia was documented by analysis of antemortem serum and postmortem cerebrospinal and vitreous fluid. Other causes of hyperammonemia, including Reye syndrome, were excluded. Histologic changes included centrilobular microvesicular steatosis of the liver and storage product present in multiple organs. The highly water-soluble mucopolysaccharide (MPS) storage product was best identified by colloidal iron staining of FFPE and unfixed air-dried fresh frozen liver sections. Alcian blue stains failed to convincingly demonstrate MPS in any of the liver sections. This is the first published report, to our knowledge, of IHA in a posttransplant patient younger than 18 years old or following transplantation for Hurler syndrome. Demonstration of the hepatic centrilobular microvesicular steatosis characteristic of IHA was complicated by the diffuse storage of MPS within the liver. MPS storage can be best detected in the liver using colloidal iron staining. Oil-red-O staining may be useful to document microvesicular steatosis in cases with a clinical history of hyperammonemia following solid organ or BMT. Determining if certain subsets of children are at increased risk for IHA requires further study.

Renal cell carcinomas in children are extremely rare and are usually associated with specific chromosomal rearrangements, different from those seen in adult patients. We present the case of a 9-year-old girl with a renal cell carcinoma with t(X;17) diagnosed at our institution. We also review the pertinent literature, with an emphasis on the genetic and molecular aspects associated with this rare neoplasm.

Cystic fibrosis (CF), an epithelial cell transport disorder caused by mutations of the cystic fibrosis transmembrane conductance regulator (CFTR) gene, is not generally associated with malformations of the central nervous system (CNS). This report describes two African-American children who died at less than 2 years of age with known Chiari I malformations and were found, unexpectedly at autopsy, to have the classic pancreatic and respiratory changes of CF. Both patients had suffered from failure to thrive that had been attributed to their CNS malformations. One child also had recurrent pneumonia and died with Pseudomonas sepsis. Mutational analysis for > 70 common CFTR mutations identified a single delta F508 mutation in one patient and a single 3120 + 1G to A mutation in the other. Their second CFTR mutations were not identified. The association of CF with Chiari I malformation is not likely to be purely coincidental, as the probability of such an occurrence in African-Americans is greater than one in 7,500,000 patients. It is possible that the CFTR gene may play a previously unrecognized role in CNS development. Alternatively, this CNS abnormality may be acquired due to the metabolic and electrolyte imbalances that characteristically occur in CF.