Second malignant neoplasms in children after treatment of soft tissue sarcoma.

Currently, approximately 67% of children diagnosed with cancer can be expected to survive more than 5 years. Among the most significant late effects of cancer therapy is the development of second malignant neoplasm (SMN). This study was performed to identify the factors associated with the development of second malignant neoplasms after treatment for soft tissue sarcomas in childhood. Retrospectively the charts of 20 children who developed second malignant neoplasms after treatment for primary childhood soft tissue sarcoma were reviewed. Presentation, age at diagnosis, tumor histology, extent of tumor, treatment, family histories (when available), and outcome were recorded. The mean age of the patients (10 boys, 10 girls) was 8.5 years of age (range, 1 to 20 years). Most primary tumors were rhabdomyosarcoma (14/20) and occurred in an extremity (10/20). Ninety percent of the patients (18/20) had a complete response to treatment of the primary cancer. Eleven out of 20 received combined chemotherapy and radiation therapy. The most common secondary malignancy was a bone sarcoma (6/20), followed by brain tumors (n = 3), leukemia (n = 2), and other sarcomas (n = 2). Four of the bone sarcomas developed in the field of radiation treatment. Median follow-up was 16 years (range, 1 to 26 years). The median time to development of a SMN was 11.4 years (range, 1.5 to 21 years). Survival after a second malignancy was only 30%. Two patients developed a third malignant neoplasm. The occurrence of a secondary malignancy represents a serious complication of childhood cancer. Certain tumors are related directly to treatment such as osteosarcoma within irradiated fields and secondary leukemias or lymphomas after certain chemotherapy regimens. Combined radiotherapy and chemotherapy may play an additive role in the development of second malignant neoplasms. Genetic factors may predispose affected patients to the development of both primary and secondary malignancies. Close surveillance of children previously treated for childhood cancers is warranted.

The human homolog of murine Evi-2 lies between two von Recklinghausen neurofibromatosis translocations.

Von Recklinghausen neurofibromatosis (NF1) is one of the most common inherited human disorders. The genetic locus that harbors the mutation(s) responsible for NF1 is near the centromere of chromosome 17, within band q11.2. Translocation breakpoints that have been found in this region in two patients with NF1 provide physical landmarks and suggest an approach to identifying the NF1 gene. As part of our exploration of this region, we have mapped the human homolog of a murine gene (Evi-2) implicated in myeloid tumors to a location between the two translocation breakpoints on chromosome 17. Cosmid-walk clones define a 60-kb region between the two NF1 translocation breakpoints. The probable role of Evi-2 in murine neoplastic disease and the map location of the human homolog suggest a potential role for EVI2 in NF1, but no physical rearrangements of this gene locus are apparent in 87 NF1 patients.

Characterization of a translocation within the von Recklinghausen neurofibromatosis region of chromosome 17.

The genetic defect causing von Recklinghausen neurofibromatosis (NF1) has been mapped to the proximal long arm of chromosome 17 by linkage analysis. Flanking markers have been identified, bracketing NF1 in 17q11.2 and laying the foundation for isolating the disease gene. Recently, a family in which a mother and her two children show both the symptoms of NF1 and the presence of a balanced translocation, t(1;17)(p34.3;q11.2), has been identified. We have examined the possibility that the translocation has occurred in or near the NF1 gene by constructing a somatic cell hybrid line containing the derivative chromosome 1 (1qter-p34.3::17q11-qter). On chromosome 1, the breakpoint occurred between SRC2 and D1S57, which are separated by 14 cM. The translocation breakpoint was localized on chromosome 17 between D17S33 and D17S57, markers that also flank NF1 within a region of 4 cM. These data are consistent with the possibility that the translocation event is the cause of NF1 in this pedigree. Consequently, the isolation of the translocation breakpoint, by approach from either the chromosome 1 or the chromosome 17 side, may facilitate the identification of the NF1 gene.

Two NF1 translocations map within a 600-kilobase segment of 17q11.2.

Balanced translocations, each involving chromosome 17q11.2, have been described in two patients with von Recklinghausen neurofibromatosis (NF1). To better localize the end points of these translocation events, and the NF1 gene (NF1) itself, human cosmids were isolated and mapped in the immediate vicinity of NF1. One cosmid probe, c11-1F10, demonstrated that both translocation breakpoints, and presumably NF1, are contained within a 600-kilobase Nru I fragment.

Precise localization of NF1 to 17q11.2 by balanced translocation.

A female patient is described with von Recklinghausen neurofibromatosis (NF1) in association with a balanced translocation between chromosome 17 and 22 [46,XX,t(17;22)(q11.2;q11.2)]. The breakpoint in chromosome 17 is cytogenetically identical to a previously reported case of NF1 associated with a 1;17 balanced translocation and suggests that the translocation events disrupt the NF1 gene. This precisely maps the NF1 gene to 17q11.2 and provides a physical reference point for strategies to clone the breakpoint and therefore the NF1 gene. A human-mouse somatic cell hybrid was constructed from patient lymphoblasts which retained the derivative chromosome 22 (22pter----22q11.2::17q11.2----17qter) but not the derivative 17q or normal 17. Southern blot analysis with genes and anonymous probes known to be in proximal 17q showed ErbA1, ErbB2, and granulocyte colony-stimulating factor (CSF3) to be present in the hybrid and therefore distal to the breakpoint, while pHHH202 (D17S33) and beta crystallin (CRYB1) were absent in the hybrid and therefore proximal to the breakpoint. The gene cluster including ErbA1 is known to be flanked by the constitutional 15;17 translocation breakpoint in hybrid SP3 and by the acute promyelocytic leukemia (APL) breakpoint, which provides the following gene and breakpoint order: cen-SP3-(D17S33,CRYB1)-NF1-(CSF3,ERBA1, ERBB2)-APL-tel. The flanking breakpoints of SP3 and API are therefore useful for rapidly localizing new markers to the neurofibromatosis critical region, while the breakpoints of the two translocation patients provide unique opportunities for reverse genetic strategies to clone the NF1 gene.

Physical mapping of the von Recklinghausen neurofibromatosis region on chromosome 17.

The von Recklinghausen neurofibromatosis (NF1) locus has been linked to chromosome 17, and recent linkage analyses place the gene on the proximal long arm. NF1 probably resides in 17q11.2, since two unrelated NF1 patients have been identified who possess constitutional reciprocal translocations involving 17q11.2 with chromosomes 1 and 22. We have used a somatic-cell hybrid from the t(17;22) individual, along with other hybrid cell lines, to order probes around the NF1 locus. An additional probe, 17L1, has been isolated from a NotI linking library made from flow-sorted chromosome 17 material and has been mapped to a region immediately proximal to the translocation breakpoint. While neither NF1 translocation breakpoint has yet been identified by pulse-field gel analysis, an overlap between two probes, EW206 and EW207, has been detected. Furthermore, we have identified the breakpoint in a non-NF1 translocation, SP-3, on the proximal side of the NF1 locus. This breakpoint has been helpful in creating a 1,000-kb pulsed-field map, which includes the closely linked NF1 probes HHH202 and TH17.19. The combined somatic-cell hybrid and pulsed-field gel analysis we report here favors the probe order D17Z1-HHH202-TH17.19-CRYB1-17L1-NF1- (EW206, EW207, EW203, L581, L946)-(ERBB2, ERBA1). The agreement in probe ordering between linkage analysis and physical mapping is excellent, and the availability of translocation breakpoints in NF1 should now greatly assist the cloning of this locus.

Molecular detection of microscopic and submicroscopic deletions associated with Miller-Dieker syndrome.

Miller-Dieker syndrome (MDS), a disorder manifesting the severe brain malformation lissencephaly ("smooth brain"), is caused, in the majority of cases, by a chromosomal microdeletion of the distal short arm of chromosome 17. Using human chromosome 17-specific DNA probes, we have begun a molecular dissection of the critical region for MDS. To localize cloned DNA sequences to the MDS critical region, a human-rodent somatic cell hybrid panel was constructed which includes hybrids containing the abnormal chromosome 17 from three MDS patients with deletions of various sizes. Three genes (myosin heavy chain 2, tumor antigen p53, and RNA polymerase II) previously mapped to 17p were excluded from the MDS deletion region and therefore are unlikely to play a role in its pathogenesis. In contrast, three highly polymorphic anonymous probes, YNZ22.1 (D17S5), YNH37.3 (D17S28), and 144-D6 (D17S34), were deleted in each of four patients with visible deletions, including one with a ring chromosome 17 that is deleted for a portion of the single telomeric prometaphase subband p13.3. In two MDS patients with normal chromosomes, a combination of somatic cell hybrid, RFLP, and densitometric studies demonstrated deletion for YNZ22.1 and YNH37.3 in the paternally derived 17's of both patients, one of whom is also deleted for 144-D6. The results indicate that MDS can be caused by submicroscopic deletion and raises the possibility that all MDS patients will prove to have deletions at a molecular level. The two probes lie within a critical region of less than 3,000 kb and constitute potential starting points in the isolation of genes implicated in the severe brain maldevelopment in MDS.

Prenatal diagnosis of deletion 17p13 associated with DiGeorge anomaly.

A fetus, subsequently shown to have the deletion 17p13, was detected at 30 weeks' gestation because of multiple anomalies and polyhydramnios on ultrasonography. The fetus died and was born at 34 weeks of gestation. Pathologic examination showed intrauterine growth retardation, double outlet right ventricle (a conotruncal cardiac defect), and thymic hypoplasia suggesting partial DiGeorge anomaly. To our knowledge, DiGeorge anomaly has not been reported previously in conjunction with del(17p) nor in the Miller-Dieker syndrome. Since this deletion is the largest deletion of distal 17p observed so far, one explanation for this association may be the presence of a gene on proximal 17p for neural crest development.

Regional mapping panel for human chromosome 17: application to neurofibromatosis type 1.

A somatic cell hybrid mapping panel was constructed to localize cloned DNA sequences to any of 15 potentially different regions of human chromosome 17. Relatively high-resolution mapping is possible for 50% of the chromosome length in which 12 breakpoints are distributed over approximately 45 megabases, with an average spacing estimated at 1 breakpoint every 2-7 megabases. This high-resolution capability includes the pericentromeric region of 17 to which von Recklinghausen neurofibromatosis (NF1) has recently been mapped. Using 20 cloned genes and anonymous probes, we have tested the expected order and location of panel breakpoints and confirmed, refined, or corrected the regional assignment of several cloned genes and anonymous probes. Four markers with varying degrees of linkage to NF1 have been physically localized and ordered by the panel: the loosely linked markers myosin heavy chain 2 (25 cM) to p12----13.105 and nerve growth factor receptor (14 cM) to q21.1----q23; the more closely linked pABL10-41 (D17S71, 5 cM) to p11.2; and the tightly linked pHHH202 (D17S33) to q11.2-q12. Thus, physical mapping of linked markers confirms a pericentromeric location of NF1 and, along with other data, suggests the most likely localization is proximal 17q.