Enhanced Risk Stratification for Children and Young Adults with B-Cell Acute Lymphoblastic Leukemia: A Children's Oncology Group Report.

Current strategies to treat pediatric acute lymphoblastic leukemia rely on risk stratification algorithms using categorical data. We investigated whether using continuous variables assigned different weights would improve risk stratification. We developed and validated a multivariable Cox model for relapse-free survival (RFS) using information from 21199 patients. We constructed risk groups by identifying cutoffs of the COG Prognostic Index (PICOG) that maximized discrimination of the predictive model. Patients with higher PICOG have higher predicted relapse risk. The PICOG reliably discriminates patients with low vs. high relapse risk. For those with moderate relapse risk using current COG risk classification, the PICOG identifies subgroups with varying 5-year RFS. Among current COG standard-risk average patients, PICOG identifies low and intermediate risk groups with 96% and 90% RFS, respectively. Similarly, amongst current COG high-risk patients, PICOG identifies four groups ranging from 96% to 66% RFS, providing additional discrimination for future treatment stratification. When coupled with traditional algorithms, the novel PICOG can more accurately risk stratify patients, identifying groups with better outcomes who may benefit from less intensive therapy, and those who have high relapse risk needing innovative approaches for cure.

Alternative V kappa gene rearrangements in a murine B cell lymphoma. An explantation for idiotypic heterogeneity.

Idiotype variants of 38C13, a murine B cell lymphoma, have been isolated by immunoselection with antiidiotype mAbs. The V region genes for the kappa light chains and mu heavy chains expressed by these tumor cells were sequenced and compared. There was no evidence for V region somatic point mutation in this tumor. However, while the heavy chain genes were all identical, the light chain genes were all different. The light chain genes of each variant were derived from the V kappa-Ox1 gene family and joined to J kappa 4, whereas the light chain gene of the parental tumor was derived from the V kappa 9 family and joined to J kappa 2. Two of the variants used the identical V kappa gene but differed by the inclusion of a variable number of additional nucleotides in the V/J joint. Thus, the idiotypic heterogeneity of this B cell lymphoma arises as a consequence of alternative light chain rearrangements rather than point mutation. This process repetitively uses members of the same V kappa gene family. Two of the variants use the identical V kappa and J kappa gene segments but differ by the presence of extra nucleotides at the V kappa/J kappa joint.

Idiotype variant cell populations in patients with B cell lymphoma.

Using isolated idiotype (Id) protein we generated panels of antibodies in two patients with follicular lymphoma, one of whom had never received prior chemo-or radiotherapy. Flow cytometry and frozen section tissue staining of tumor with these monoclonal antibodies (mAb) revealed multiple subpopulations within each tumor. Individual mAb stained between 7% and 83% of surface Ig+ cells in the tumor samples. These subpopulations were overlapping and no single antibody recognized all the tumor cells. However, combinations of antibodies seemed to capture total tumor in both cases. In some instances, the percentage of tumor stained by a single mAb varied over time, and differed between lymph nodes sampled at the same time. Because a single species of Id protein was used to generate mAb in each case, it appears that the antibodies were directed against idiotopes variably shared by different populations within each tumor, and this was confirmed by crossblocking studies. Tumor cells from one patient were fused to a nonsecreting heteromyeloma line K6H6/B5, and most of the resulting hybrids secreted Id protein. Four mAb were used to screen the Id proteins secreted by these hybrids, and 11 different variants (16 maximal) were found. Southern blot analysis of rearranged Ig genes was done in two hybrids and biopsy material. Identically rearranged light-chain genes were seen but it appeared as though extensive somatic variation had occurred in heavy chain genes. These studies indicate that: striking Id variation can exist at diagnosis in untreated patients, the percentage of tumor represented by an individual variant may change with time and may differ between tumor sampled from different anatomical locations, and somatic variation appears to be responsible for the observed heterogeneity. Although this degree of variation makes anti-Id antibody therapy more difficult, appropriate combinations of mAb should be more efficacious than single antibodies in such cases.

Immunoglobulin light chain variable region gene sequences for human antibodies to Haemophilus influenzae type b capsular polysaccharide are dominated by a limited number of V kappa and V lambda segments and VJ combinations.

The immune repertoire to Haemophilus influenzae type b capsular polysaccharide (Hib PS) appears to be dominated by certain light chain variable region genes (IgVL). In order to examine the molecular basis underlying light chain bias, IgVL genes have been cloned from a panel of heterohybridomas secreting human anti-Hib PS (antibody) (anti-Hib PS Ab). One hybridoma, representative of the predominant serum clonotype of anti-Hib PS Ab in older children and adults following immunization or Hib infection, uses a V kappa II segment identical to the germline gene A2, and a JK3 segment. A second kappa hybridoma uses a member of the V kappa I family and a JK4 segment. Four lambda antibodies, all cross-reactive with the structurally related antigen Escherichia coli K100 PS, use V lambda VII segments which are 96-98% homologous to one another, and may originate from a single germline gene. Two additional lambda antibodies, not K100-cross-reactive, are encoded by members of the V lambda II family. All lambda antibodies use highly homologous J lambda 2 or J lambda 3 segments. The VJ joints of all lambda antibodies and the V kappa II-encoded antibody are notable for the presence of an arginine codon, suggesting an important role in antigen binding. Although more complex than heavy chain variable region gene usage, a significant portion of serum anti-Hib PS Ab is likely to be encoded by a limited number of V kappa and V lambda segments and VJ combinations, which may be selectively expressed during development, or following antigen exposure.

Restricted immunoglobulin VH usage and VDJ combinations in the human response to Haemophilus influenzae type b capsular polysaccharide. Nucleotide sequences of monospecific anti-Haemophilus antibodies and polyspecific antibodies cross-reacting with self antigens.

To examine the human antibody repertoire generated against a biologically significant antigen we have obtained sequences of heavy chain variable region genes (IgVH) from 15 monoclonal antibodies specific for the capsular polysaccharide of Haemophilus influenzae type b (Hib PS). All VH segments are members of the VH3 family and 9 of 15 are members of the smaller VH3b subfamily. Restriction is evident by the shared use of certain VDJ joints in independent hybridomas from different subjects. Two hybridomas generated from the same subject demonstrate identical heavy chain variable region gene sequences but differ in isotype and rearrange alternative light chain variable region genes (IgVL), suggesting that in a normal immune response, a single pre-B cell clone may use different light chain rearrangements and give rise to progeny capable of reacting with antigen. Using a polymerase chain reaction assay optimized to detect base pair differences among VH genes we demonstrate that at least a portion of expressed anti-Hib PS VH genes have undergone somatic mutation. Anti-Hib PS heavy chain genes are homologous to VH segments encoding autoantibodies and two hybridomas secrete anti-Hib PS antibody that cross-reacts with self antigens (double-stranded DNA and single-stranded DNA). Comparison of VH regions of self-reactive and monospecific anti-Hib PS Ab demonstrates no consistent structural feature correlating with fine antigen specificity. These data demonstrate significant restriction in VH usage and VDJ recombination in the anti-Hib PS response and confirm that autoantibodies may be elicited during normal immune responses.

Familial bone marrow monosomy 7. Evidence that the predisposing locus is not on the long arm of chromosome 7.

Loss of expression of a tumor-suppressing gene is an attractive model to explain the cytogenetic and epidemiologic features of cases of myelodysplasia and acute myelogenous leukemia (AML) associated with bone marrow monosomy 7 or partial deletion of the long arm (7q-). We used probes from within the breakpoint region on 7q-chromosomes (7q22-34) that detect restriction fragment length polymorphisms (RFLPs) to investigate three families in which two siblings developed myelodysplasia with monosomy 7. In the first family, probes from the proximal part of this region identified DNA derived from the same maternal chromosome in both leukemias. The RFLPs in these siblings diverged at the more distal J3.11 marker due to a mitotic recombination in one patient, a result that suggested a critical region on 7q proximal to probe J3.11. Detailed RFLP mapping of the implicated region was then performed in two additional unrelated pairs of affected siblings. In these families, DNA derived from different parental chromosome 7s was retained in the leukemic bone marrows of the siblings. We conclude that the familial predisposition to myelodysplasia is not located within a consistently deleted segment on the long arm of chromosome 7. These data provide evidence implicating multiple genetic events in the pathogenesis of myelodysplasia seen in association with bone marrow monosomy 7 or 7q-.

A novel intron element operates posttranscriptionally To regulate human N-myc expression.

Precisely regulated expression of oncogenes and tumor suppressor genes is essential for normal development, and deregulated expression can lead to cancer. The human N-myc gene normally is expressed in only a subset of fetal epithelial tissues, and its expression is extinguished in all adult tissues except transiently in pre-B lymphocytes. The N-myc gene is overexpressed due to genomic amplification in the childhood tumor neuroblastoma. In previous work to investigate mechanisms of regulation of human N-myc gene expression, we observed that N-myc promoter-chloramphemicol acelyltransferase reporter constructs containing sequences 5' to exon 1 were active in all cell types examined, regardless of whether endogenous N-myc RNA was detected. In contrast, inclusion of the first exon and a portion of the first intron allowed expression only in those cell types with detectable endogenous N-myc transcripts. We investigated further the mechanisms by which this tissue-specific control of N-myc expression is achieved. Using nuclear run-on analyses, we determined that the N-myc gene is actively transcribed in all cell types examined, indicating a posttranscriptional mode of regulation. Using a series of N-myc intron 1 deletion constructs, we localized a 116-bp element (tissue-specific element [TSE]) within the first intron that directs tissue-specific N-myc expression. The TSE can function independently to regulate expression of a heterologous promoter-reporter minigene in a cell-specific pattern that mirrors the expression pattern of the endogenous N-myc gene. Surprisingly, the TSE can function in both sense and antisense orientations to regulate gene expression. Our data indicate that the human N-myc TSE functions through a posttranscriptional mechanism to regulate N-myc expression.

Genetics of ancestry-specific risk for relapse in acute lymphoblastic leukemia.

The causes of individual relapses in children with acute lymphoblastic leukemia (ALL) remain incompletely understood. We evaluated the contribution of germline genetic factors to relapse in 2225 children treated on Children's Oncology Group trial AALL0232. We identified 302 germline single-nucleotide polymorphisms (SNPs) associated with relapse after adjusting for treatment and ancestry and 715 additional SNPs associated with relapse in an ancestry-specific manner. We tested for replication of these relapse-associated SNPs in external data sets of antileukemic drug pharmacokinetics and pharmacodynamics and an independent clinical cohort. 224 SNPs were associated with rapid drug clearance or drug resistance, and 32 were replicated in the independent cohort. The adverse risk associated with black and Hispanic ancestries was attenuated by addition of the 4 SNPs most strongly associated with relapse in these populations (for blacks: model without SNPs hazard ratio (HR)=2.32, P=2.27 × 10-4, model with SNPs HR=1.07, P=0.79; for Hispanics: model without SNPs HR=1.7, P=8.23 × 10-5, model with SNPs HR=1.31, P=0.065). Relapse SNPs associated with asparaginase resistance or allergy were overrepresented among SNPs associated with relapse in the more asparaginase intensive treatment arm (20/54 in Capizzi-methorexate arm vs 8/54 in high-dose methotrexate arm, P=0.015). Inherited genetic variation contributes to race-specific and treatment-specific relapse risk.

Genome-Wide Study Links PNPLA3 Variant With Elevated Hepatic Transaminase After Acute Lymphoblastic Leukemia Therapy.

Remission induction therapy for acute lymphoblastic leukemia (ALL) includes medications that may cause hepatotoxicity, including asparaginase. We used a genome-wide association study to identify loci associated with elevated alanine transaminase (ALT) levels after induction therapy in children with ALL enrolled on St. Jude Children's Research Hospital (SJCRH) protocols. Germline DNA was genotyped using arrays and exome sequencing. Adjusting for age, body mass index, ancestry, asparaginase preparation, and dosage, the PNPLA3 rs738409 (C>G) I148M variant, previously associated with fatty liver disease risk, had the strongest genetic association with ALT (P = 2.5 × 10-8 ). The PNPLA3 rs738409 variant explained 3.8% of the variability in ALT, and partly explained race-related differences in ALT. The PNPLA3 rs738409 association was replicated in an independent cohort of 2,285 patients treated on Children's Oncology Group protocol AALL0232 (P = 0.024). This is an example of a pharmacogenetic variant overlapping with a disease risk variant.

Autoregulation of the human N-myc oncogene is disrupted in amplified but not single-copy neuroblastoma cell lines.

Amplification of the N-myc gene is a significant adverse prognostic factor in neuroblastoma, a common childhood tumor. In non-transformed cells, myc expression is controlled through an autoregulatory circuit, through which elevated Myc protein levels lead to down-regulation of myc transcription. The precise mechanism of myc gene autoregulation is unknown. Loss of c-myc autoregulation has been documented in transformed cells from a number of different lineages, but N-myc autoregulation has not yet been investigated. In neuroblastoma, the increased N-Myc protein produced by amplified tumors would be expected to silence N-myc transcription if the autoregulatory loop were intact. To determine whether N-myc autoregulation is operative in human neuroblastoma, and to localize cis-acting elements which mediate N-myc autosuppression, we transfected a series of N-myc 5' promoter constructs into a panel of human neuroblastoma cell lines carrying one or multiple copies of N-myc. The transfected promoter was equally active in single-copy and amplified lines. Significant promoter activity in the presence of abundant Myc protein in amplified neuroblastoma lines indicates that autoregulation is disabled in this subset of tumors. To investigate whether single-copy lines produce insufficient N-Myc protein to trigger autosuppression yet retain an intact autoregulatory circuit, we transfected neuroblastoma lines with 5' promoter constructs in the presence of a c- or N-myc expression vector. Overexpression of c- or N-Myc resulted in diminution of activity of both the transfected promoter and the endogenous N-myc gene in single-copy, but not amplified lines. Using a series of 5' promoter-deletion minigenes, we localized a cis-acting element required for autoregulation close to the transcription start sites. While the precise mechanism of autosuppression remains unknown, we demonstrated that Myc is incapable of silencing the adenovirus major late promoter (AdMLP) in neuroblastoma cells, indicating that Myc suppression of its own promoter and the AdMLP involve distinct components. These studies provide the first systematic investigation of autoregulation in neuroblastoma, and indicate that single-copy neuroblastoma lines produce insufficient N-Myc protein to activate downstream effector(s) of autosuppression; the autoregulatory circuit is otherwise intact. Amplified lines, in contrast, have lost autoregulation.

Diversity of the apoptotic response to chemotherapy in childhood leukemia.

Apoptosis is the primary mechanism through which most chemotherapeutic agents induce tumor cell death. The purpose of this study was to determine the extent to which blasts from children with leukemia undergo a uniform apoptotic death pathway in vivo. The expression of pro- and anti-apoptotic proteins p53, p21, MDM-2, BCL-2, BCL-X(L), BCL-X(S), and BAX, and caspase-3 activity was determined in circulating blasts collected from the peripheral blood of children with leukemia prior to, and at serial time points following chemotherapy. Culturing blasts ex vivo for 12 h assessed spontaneous apoptosis and the increment induced by chemotherapy. Baseline apoptosis varied between 3% and 29%. Twenty-four hours following chemotherapy the increase in the percentage of cells undergoing apoptosis ranged from <1% to 38%. Eleven of 20 patients who received initial treatment with a p53-dependent drug showed an increase in p53 expression. In these patients, the levels of p53 target genes were also increased. A uniform pattern of BCL-2 family protein expression was not observed and only a minority of samples showed a change that would favor apoptosis. We conclude that that the initial apoptotic response to chemotherapy in children with leukemia is variable involving both p53-dependent and p53-independent pathways.

Hybridoma fusion cell lines contain an aberrant kappa transcript.

The V region sequence of a non-productive kappa transcript from two myeloma fusion partners has been determined. This transcript has an aberrant VJ recombination site resulting in a translation stop site at position 105. It is variably expressed in hybridomas made from all fusion partners derived from the original MOPC-21 tumor. The amount of this transcript may greatly exceed levels of the productive light chain mRNA.

A shuttle vector system for the investigation of immunoglobulin gene hypermutation: absence of enhanced mutability in intermediate B cell lines.

Somatic hypermutation focused to the rearranged V(D)J segment of the immunoglobulin (Ig) loci contributes substantially to antibody gene diversification. However, neither the precise B cell subset subject to hypermutation nor the molecular mechanism(s) involved is known. One model proposes that Ig segments may be uniquely susceptible to DNA nicking and subsequent error-prone repair during a specific B cell developmental stage. We describe an SV40-based shuttle vector system for testing such a model. Plasmids containing two distinct Ig segments juxtaposed to the supF marker gene have been passaged through cell lines representing intermediate stages of B cell development, rescued and screened for marker gene mutations. To date we have not demonstrated enhanced supF mutation in any cell line examined, irrespective of the adjacent Ig segment. Thus, these cell lines exhibit normal DNA repair mechanisms and no evidence of increased endonuclease activity on the Ig segments tested. The feasibility of this system will allow similar experiments using other Ig target sequences exposed to a broader range of B cells.

Introduction to recombinant-DNA technology.

Recombinant-DNA technology is now commonly used in virtually every aspect of the biological sciences. The purpose of this brief exposition is to provide an outline of the approaches used to identify genes, to isolate the gene of interest, to amplify the gene if necessary, and to clone genes. A short introduction to the principles of separating very large genes is provided, along with a description of an approach to propagating and cloning these large genes.

Mouse x human heterohybridomas as fusion partners with human B cell tumors.

Surface idiotype (Id) of B cell malignancies is an excellent tumor-specific marker. We have, however, recently described heterogeneity of tumor Id in some cases. We therefore sought a way to isolate, reliably and efficiently, different species of idiotype from a potentially heterogeneous population. In this report we demonstrate our success using a series of mouse X human heterohybridomas as fusion partners with human B cell tumors. Three lines (K6H6/B5, K6H9/G12, SBC/H20) demonstrated excellent fusion efficiency with 75%-85% of wells plated containing hybrids. Two cell lines, K6H9/G12 and SBC/H20 had a tendency to secrete a single Ig chain (heavy or light chain), whereas the K6H6/B5 cell line secreted whole immunoglobulin (Ig) in greater than 80% of the hybrids. This line secreted significant amounts of Ig (2.73 micrograms/ml/10(6) cells) and was relatively stable in culture. Since this line has such a high fusion efficiency the products of normal B cells admixed with tumor may be recovered, allowing the opportunity of isolating host anti-tumor antibodies. In order to prove that hybrids were derived from the tumor, Southern blot analysis of rearranged DNA was performed in selected cases. Fusions with this line provide the potential for recovering many different species of idiotype in a mixed population. This will facilitate the production of mouse monoclonal anti-idiotype antibodies against many variants and against different idiotopes.

Spectrum of Salmonella-associated arthritis.

Arthritis is an uncommon extraintestinal manifestation of Salmonella infection. Three patients with Salmonella-associated arthritis with varying manifestations were seen at Children's Hospital Medical Center in an 11-month period: (1) a 12-year-old girl developed suppurative arthritis due to Salmonella typhimurium that required surgical drainage and prolonged parenteral antibiotic therapy; (2) a 12-year-old girl had migratory polyarthritis following gastrointestinal infection with S typhimurium; the acute synovitis subsided after a six-month period following anti-inflammatory medications; (3) a 14-year-old girl developed conjunctivitis, urethritis, and polyarthritis (Reiter's syndrome) in association with Salmonella gastroenteritis. These patients illustrate the distinct types of arthritis associated with Salmonella gastroenteritis. These patients illustrate that distinct types of arthritis associated with Salmonella, and the association of this organism with both suppurative joint disease and reactive arthritis is reemphasized.

Treatment of occult bacteremia: a prospective randomized clinical trial.

Antibiotic therapy for children without foci of infection and at risk for bacteremia is controversial. A prospective randomized clinical trial was conducted using expectant antibiotic therapy in children at risk for bacteremia. A total of 96 children (aged 6 to 24 months) with temperature of more than 40 degrees C, no identifiable source of infection, and a leukocyte count greater than or equal to 15,000/microL and/or sedimentation rate greater than or equal to 30 were enrolled. The following tests were performed on all children: blood culture, chest roentgenogram, urinalysis, and urine culture. A lumbar puncture was performed if a child was 12 months or less. Patients were randomized to receive either no antibiotic therapy or Bicillin C-R, 50,000 U/kg intramuscularly, followed by penicillin V, 100 mg/kg/d, orally four times a day for three days. Patients were examined at 24 and 72 hours. Fifty patients were treated expectantly and 46 received no antimicrobial therapy. Ten of the 96 patients were bacteremic (nine had Streptococcus pneumoniae, one had Haemophilus influenzae). Four of the five children treated for bacteremia showed improvement at the first follow-up visit (afebrile and no obvious focus of infection). The five untreated patients showed no improvement; four patients developed focal infections (two had meningitis, two had otitis media) (P less than or equal to .05, Fisher exact test). No complications of expectant therapy were detected. Thus, expectant antibiotic therapy for children who have no obvious source of infection and who meet these criteria associated with occult bacteremia is warranted.

The repertoire of human antibody to the Haemophilus influenzae type b capsular polysaccharide.

Human antibody to the Haemophilus influenzae capsular polysaccharide (Hib CP) is restricted in diversity in the individual and the population with a limited number of variable region genes encoding antibody. Antibody to the Hib CP shows restricted isoelectric focusing gel patterns and light chain usage with frequent restriction to use of only kappa light chains. Shared cross-reactive idiotypes are expressed on antibody. The heavy chain of antibody to the Hib CP is predominantly encoded by two members of the VH3 family--LSG 6.1/M85-like and VH26/30P1-like. In VH the CDR1, based on complete identity in LSG 6.1/M85-like antibodies, CDR2, based on the suggestion of mutation in this region, and CDR3, based on conserved CDR3 usage in unrelated individuals, may be important for antigen binding. Six or more different VL gene families encode antibody. The predominant antibody of the majority of individuals uses the A2-V kappa II gene in germline or near germline configuration, which encodes an idiotype designated HibId-1. Antibody can also be encoded by V kappa I, non-A2 V kappa II, V kappa III, V kappa IV, V lambda II, and V lambda VII genes. Although different VL genes can be used, unrelated individuals appear to use the same V kappa III (A27), V lambda II (V lambda 2.1 and V lambda VII (4A) genes. The VL diversity accounts for differences in fine binding specificity, with A2-V kappa II genes not encoding E. coli K100 CP cross-reactive antibodies and V lambda VII genes and some of the non-A2 V kappa genes encoding cross-reactive antibodies. The arginine in CDR3 of both antibody kappa and lambda light chains and the asparagine in CDR2 of VL sequences and in CDR1 of LSG6.1-M85 VH sequences of antibody appear to be important residues for antigen binding. A relatively limited degree of somatic mutation has occurred in the non-A2 VL genes, V lambda VII, and the VH genes. Further studies comparing the polymorphism of germline V genes to antibody-encoding V genes are needed to clarify this issue. Research comparing this repertoire to repertoires directed to other bacterial CP and to self antigens and defining structure-antigen binding relationships is in progress.

Gene expression profiling. Methods and clinical applications in oncology.

The advent of microarray technology undoubtedly will have great impact on the medical field during the next decade. This article discusses different genomic technologies, statistical methods for data analysis, and clinical applications of microarrays. Emphasis is devoted to integration of microarrays into the field of pediatric oncology.