Niels Carlsen

Population-based risk estimates of Wilms tumor in sporadic aniridia

Abstract. Aniridia is a severe eye disease characterized by iris hypoplasia; both sporadic cases and familial cases with an autosomal dominant inheritance exist. Mutations in the PAX6 gene have been shown to be the genetic cause of the disease. Some of the sporadic cases are caused by large chromosomal deletions, some of which also include the Wilms tumor gene (WAGR syndrome), resulting in an increased risk of developing Wilms tumor. Based on the unique registration of both cancer and aniridia cases in Denmark, we have made the most accurate risk estimate to date for Wilms tumor in sporadic aniridia. We have found that patients with sporadic aniridia have a relative risk of 67 (confidence interval: 8.1–241) of developing Wilms tumor. Among patients investigated for mutations, Wilms tumor developed in only two patients out of 5 with the Wilms tumor gene (WT1) deleted. None of the patients with smaller chromosomal deletions or intragenic mutations were found to develop Wilms tumor. Our observations suggest a smaller risk for Wilms tumor than previous estimates, and that tumor development requires deletion of WT1. We report a strategy for the mutational analysis of aniridia cases resulting in the detection of mutations in 68% of sporadic cases and 89% of familial cases. We also report four novel mutations in PAX6, and furthermore, we have discovered a new alternatively spliced form of PAX6.

Comparison of intramuscular therapy with Erwinia asparaginase and asparaginase Medac: pharmacokinetics, pharmacodynamics, formation of antibodies and influence on the coagulation system

Asparaginase comes from different biological sources and the various preparations have different pharmacokinetic properties, and their tendency to induce side‐effects is different. Erwinia asparaginase (ASNase) has a shorter half‐life than the Escherichia coli preparations, and it has been reported to be less immunogenic than the E. coli preparations and to induce fewer coagulation disorders. Children with newly diagnosed acute lymphoblastic leukaemia (ALL) were included in this study. Twenty‐seven patients were treated with Erwinia ASNase (induction therapy 30·000 IU/m2/d i.m. for 10 d, and re‐induction therapy 30·000 IU/m2 twice a week for 2 weeks) and 15 were treated with ASNase Medac (induction therapy 1·000 IU/m2/d i.m. for 10 d, and re‐induction therapy 5·000 IU/m2 i.m. twice a week for 2 weeks). Blood samples were drawn to determine enzyme activity, l‐asparagine, anti‐asparaginase antibodies, and coagulation parameters. After i.m. administration, Erwinia ASNase displayed a protracted absorption phase compared to ASNase Medac. The mean bioavailability after i.m. administration was 27% for Erwinia ASNase and 45% for ASNase Medac respectively. Mean trough enzyme activities during induction therapy were Erwinia ASNase 1748 IU/l and ASNase Medac 272 IU/l, and during re‐induction therapy Erwinia ASNase 83 IU/l and ASNase Medac 147 IU/l. We conclude that in this setting, therapy with ASNase Medac resulted in sufficient treatment during both phases of therapy, whereas treatment with Erwinia ASNase resulted in unnecessarily intense therapy during the induction phase and insufficient treatment during the re‐induction phase. There was no significant difference in the incidence of antibody formation, and therapy with Erwinia ASNase resulted in a more pronounced influence on the coagulation parameters than therapy with ASNase Medac.

Fusion of NUP98 and the SET binding protein 1 (SETBP1) gene in a paediatric acute T cell lymphoblastic leukaemia with t(11;18)(p15;q12)

Three NUP98 chimaeras have previously been reported in T cell acute lymphoblastic leukaemia (T‐ALL): NUP98/ADD3, NUP98/CCDC28A, and NUP98/RAP1GDS1. We report a T‐ALL with t(11;18)(p15;q12) resulting in a novel NUP98 fusion. Fluorescent in situ hybridisation showed NUP98 and SET binding protein 1(SETBP1) fusion signals; other analyses showed that exon 12 of NUP98 was fused in‐frame with exon 5 of SETBP1. Nested polymerase chain reaction did not amplify the reciprocal SETBP1/NUP98, suggesting that NUP98/SETBP1 transcript is pathogenetically important. SETBP1 has previously not been implicated in leukaemias; however, it encodes a protein that specifically interacts with SET, fused to NUP214 in a case of acute undifferentiated leukaemia.

The association of reduced folate carrier 80G>A polymorphism to outcome in childhood acute lymphoblastic leukemia interacts with chromosome 21 copy number

The reduced folate carrier (RFC) is involved in the transport of methotrexate (MTX) across the cell membrane. The RFC gene (SLC19A1) is located on chromosome 21, and we hypothesized that the RFC80 G>A polymorphism would affect outcome and toxicity in childhood leukemia and that this could interact with chromosome 21 copy number in the leukemic clone. A total of 500 children with acute lymphoblastic leukemia treated according to the common Nordic treatment protocols were included, and we found that the RFC AA variant was associated with a 50% better chance of staying in remission compared with GG or GA variants (P = .046). Increased copy numbers of chromosome 21 appear to improve outcome also in children with GA or GG variant. In a subset of 182 children receiving 608 high-dose MTX courses, we observed higher degree of bone marrow toxicity in patients with the RFC AA variant compared with GA/GG variants (platelet 73 vs 99/105 x 10(9)/L, P = .004, hemoglobin 5.6 vs 5.9/6.0 mmol/L, P = .004) and a higher degree of liver toxicity in patients with RFC GG variant (alanine aminotransferase 167 vs 127/124 U/L, P = .05). In conclusion, the RFC 80G>A polymorphism interacts with chromosome 21 copy numbers and affects both efficacy and toxicity of MTX.

Treatment of the X‐linked lymphoproliferative, Griscelli and Chédiak–Higashi syndromes by HLH directed therapy

Griscelli syndrome type 2 (GS2), the X‐linked lymphoproliferative (XLP) and the Chédiak–Higashi (CHS) syndromes are diseases that all may develop hemophagocytic syndromes. We wanted to investigate whether the treatment protocols for hemophagocytic lymphohistiocytosis (HLH) can also be used for these syndromes.

Neuroblastoma: epidemiology and pattern of regression. Problems in interpreting results of mass screening.

Neuroblastomas are malignant, embryonic tumors, and most are probably congenital in origin. Neuroblastoma is a heterogeneous neoplastic disease with respect to response to treatment and prognosis. Some of its clinical behavior is ascribed to the ability of the tumor to regress spontaneously. Spontaneous regression of neuroblastoma is occasionally observed in young infants, but occurs extremely rarely in older children. This feature is in accordance with the concept that tumors develop by a series of changes, progressing from a dependent state to an autonomous state. Spontaneous regression was documented in less than 2% of Danish neuroblastoma patients. However, the “true” incidence of regression may be higher. Epidemiological data suggest that, in recent decades, borderline lesions may be included among truly malignant neuroblastomas. The recent birth cohorts compiled in Denmark show that 1 in 7,000 live births will develop neuroblastoma before 15 years of age. If 65% of all childhood neuroblastomas could be detected at or before the age of 6 months, then the expected prevalence by screening would be 1/11,000. One can speculate that 1/18,000 live births might possibly benefit from screening at age 6 months, since 59% of these children had tumors in stages III and IV diagnosed after the age of 6 months. However, the proof that screening can only detect lesions that would have progressed to become malignant tumors depends on observing an appropriate fall in the incidence of the malignant tumors in older children and a decline in mortality. Therefore, it is of some concern that the incidence of neuroblastomas appears to actually increase with screening while mortality has been little effected. It is of equal concern that certain early stage tumors detected by screening may not have evolved into advanced stage tumors if they were left untreated.

Late cardiac effects of anthracycline containing therapy for childhood acute lymphoblastic leukemia

At present about 80% of children with acute lymphoblastic leukemia (ALL) will be cured following treatment with multi‐drug chemotherapy. A major concern for this growing number of survivors is the risk of late effects of treatment. The aim of this study was to determine whether signs of cardiomyopathy were present in patients treated in childhood with cumulative anthracycline doses of less than 300 mg/m2.

Prognostic factors in neuroblastomas treated in denmark from 1943 to 1980: A statistical estimate of prognosis based on 253 cases

Multivariate analysis on an unselected patient population consisting of all 253 children treated for neuroblastoma in Denmark during 1943 to 1980 shows that stage, age, and treatment given are independent prognostic variables. Calendar year of diagnosis, sex of the patient, and site of primary tumor were not significant prognostic factors. Further analysis shows that multimodal treatment with surgery, irradiation, and chemotherapy, especially in patients older than 1 year of age with Stage II disease, has influenced the survival significantly. The fact that age at diagnosis and the administration of chemotherapy have independent prognostic significance can be explained by the theory that all neuroblastomas are virtually congenital; therefore, the difference in age at diagnosis largely reflects the difference in growth rates of the tumor. Thus, according to this theory, age may be a measure of the probability of micrometastases in addition to the clinical extent or stage of the disease, as it represents the duration of the disease. Additional chemotherapy may thus have eradicated these micrometastases in the older children, since the age influence on Stage II disease disappeared when multimodal treatment was given in this study. The implications for treatment policy are discussed in view of this theory.

Long-term cardiac follow-up of children treated with anthracycline doses of 300 mg/m2 or less for acute lymphoblastic leukemia†

The cardiotoxic effect of anthracyclines has been well described for moderate to high cumulative doses (>350 mg/m2). However, the question of whether sub‐clinical signs of cardiomyopathy may develop and progress over time in children receiving doses of <350 mg/m2 is controversial. The aim of the present study was to examine cardiac function with serial echocardiography from diagnosis to last follow‐up, relapse, or death, and to investigate whether suspected risk factors (e.g., age at diagnosis, gender, cumulative dose, and length of follow‐up) had a significant influence on cardiac function.

Polymorphisms in the ABCB1 gene and effect on outcome and toxicity in childhood acute lymphoblastic leukemia.

The membrane transporter P-glycoprotein, encoded by the ABCB1 gene, influences the pharmacokinetics of anti-cancer drugs. We hypothesized that variants of ABCB1 affect outcome and toxicity in childhood acute lymphoblastic leukemia (ALL). We studied 522 Danish children with ALL, 93% of all those eligible. Risk of relapse was increased 2.9-fold for patients with the 1199GA variant versus 1199GG (P=0.001), and reduced 61% and 40%, respectively, for patients with the 3435CT or 3435TT variants versus 3435CC (overall P=0.02). The degree of bone marrow toxicity during doxorubicin, vincristine and prednisolone induction therapy was more prominent in patients with 3435TT variant versus 3435CT/3435CC (P=0.01/P<0.0001). We observed more liver toxicity after high-dose methotrexate in patients with 3435CC variant versus 3435CT/TT (P=0.03). In conclusion, there is a statistically significant association between ABCB1 polymorphisms, efficacy and toxicity in the treatment of ALL, and ABCB1 1199G>A may be a new possible predictive marker for outcome in childhood ALL.