Non-HLA type 1 diabetes genes modulate disease risk together with HLA-DQ and islet autoantibodies.

The possible interrelations between human leukocyte antigen (HLA)-DQ, non-HLA single-nucleotide polymorphisms (SNPs) and islet autoantibodies were investigated at clinical onset in 1-34-year-old type 1 diabetes (T1D) patients (n=305) and controls (n=203). Among the non-HLA SNPs reported by the Type 1 Diabetes Genetics Consortium, 24% were supported in this Swedish replication set including that the increased risk of minor PTPN22 allele and high-risk HLA was modified by GAD65 autoantibodies. The association between T1D and the minor AA+AC genotype in ERBB3 gene was stronger among IA-2 autoantibody-positive patients (comparison P=0.047). The association between T1D and the common insulin (AA) genotype was stronger among insulin autoantibody (IAA)-positive patients (comparison P=0.008). In contrast, the association between T1D and unidentified 26471 gene was stronger among IAA-negative (comparison P=0.049) and IA-2 autoantibody-negative (comparison P=0.052) patients. Finally, the association between IL2RA and T1D was stronger among IAA-positive than among IAA-negative patients (comparison P=0.028). These results suggest that the increased risk of T1D by non-HLA genes is often modified by both islet autoantibodies and HLA-DQ. The interactions between non-HLA genes, islet autoantibodies and HLA-DQ should be taken into account in T1D prediction studies as well as in prevention trials aimed at inducing immunological tolerance to islet autoantigens.

The association between the PTPN22 1858C>T variant and type 1 diabetes depends on HLA risk and GAD65 autoantibodies.

The single nucleotide polymorphism 1858C>T in the PTPN22 gene is associated with type 1 diabetes (T1D) in several populations. Earlier reports have suggested that the association may be modified by human leukocyte antigen (HLA), as well as by islet autoantibodies. In a large case-control study of Swedish incident T1D patients and controls, 0-34 years of age, we tested whether the odds ratio (OR) measure of association was dependent on HLA or autoantibodies against the islet autoantigens glutamic acid decarboxylase 65 kDa autoantibodies (GADA), insulin, islet antigen-2, or islet cell. The association between the carrier status of 1858C>T allele in PTPN22 (PTPN22(CT+TT)) and T1D was modified by HLA. In addition, in GADA-positive T1D, the OR was 2.83 (2.00, 3.99), whereas in GADA-negative T1D, the OR was 1.41 (0.98, 2.04) (P for comparison=0.007). The OR of association between PTPN22(CT+TT) and GADA-positive T1D declined with increasing HLA-risk category from 6.12 to 1.54 (P=0.003); no such change was detected in GADA-negative T1D (P=0.722) (P for comparison=0.001). However, the absolute difference in risk between PTPN22(CC) and PTPN22(CT+TT) subjects with high-risk HLA was five times higher than that for subjects with low-risk HLA. We hypothesize that the altered T-cell function because of the PTPN22(1858C>T) polymorphism is exclusively associated with GADA-positive T1D at diagnosis.

Epidemiological features of Wilms' tumor: results of the National Wilms' Tumor Study.

Nearly 2,000 children with Wilm's tumor registered in a national clinical trial during 1969-81 showed high rates of aniridia, hemihypertrophy, cryptorchidism, hypospadias, and other genitourinary anomalies. Patients with bilateral disease, who constituted 5% of the total, had younger ages at diagnosis and an increased incidence of congenital anomalies and renal blastemal rests. Those with multicentric unilateral lesions had more blastemal rests but were otherwise indistinguishable from the unicentric cases. The 20 familial cases had none of the features usually associated with genetic tumors: neither younger ages nor an increase in bilaterality nor associated congenital anomalies. These observations suggest that the fraction of Wilm's tumors that is due to an inherited mutation may be substantially smaller than previously supposed and support the concept that the disease arises from a variety of pathogenetic pathways.

Childhood brain tumors and paternal occupation in the aerospace industry.

Data from a case-control study of childhood brain tumors were analyzed to examine the possibility that paternal occupation in the aerospace industry is related to the development of a brain tumor in offspring. Parents of 51 children with brain tumors diagnosed in western Washington State during 1978-81 were interviewed, and their responses were compared to those of parents of 142 children selected at random from this population. Among all children, proportions of case and control fathers who had ever been employed in the aerospace industry were nearly identical [relative risk (RR) = 0.94; 95% confidence interval (CI) = 0.40-2.19]. Employment in the aerospace industry during the period from 1 year prior to birth to the time of diagnosis and any employment in the manufacturing part of the industry were not associated with increased risk. However, stratification by age at diagnosis revealed an increased risk associated with father's ever-employment in the industry (RR = 2.10; 95% CI = 0.79-5.60) for children under 10 years old. A corresponding decreased risk (RR = 0.12; 95% CI = 0.01-1.08) was found for children over 10 years old. Because of the relatively small number of cases with a positive paternal occupational history, interpretations of the difference in the direction of the association according to age at diagnosis must remain tentative ones.

Second malignant neoplasms in survivors of Wilms' tumor: a report from the National Wilms' Tumor Study.

Between October 1969 and December 1982, 2,438 patients were enrolled in the National Wilms' Tumor Study and contributed 14,381 person-years of observation to a follow-up study for the occurrence of second malignant neoplasms (SMNs). Fifteen SMNs were observed, whereas 1.77 would have been expected from U.S. incidence rates for 1973-1977 [relative risk = 8.5; 95% confidence interval (CI) = 4.7, 14.0]. Ten years after the Wilms' tumor diagnosis, the cumulative risk of SMN was 1%. The relative risks compared to standard rates were 12/1.11 = 10.8 (95% CI = 5.6, 18.9) for those who received radiation as part of the initial course of treatment and 3/0.60 = 5.0 (95% CI = 1.0, 14.6) for those who did not, but this difference was not statistically significant. Preliminary data suggest that substantial numbers of SMNs occur as patients are followed greater than 10 years from diagnosis.

U.S. centralized cancer patient data system for uniform communication among cancer centers.

The evolution of the Centralized Cancer Patient Data System, a cooperative venture of the 21 comprehensive cancer centers in the United States, and its structure at the end of 3 years of data collection are described. Functions of the data system are detailed in terms of input and output. It is concluded that the short-run objective of establishing a data system to provide high-quality patient data that ae 21 comprehensive cancer centers in the United States, and its structure at the end of 3 years of data collection are described. Functions of the data system are detailed in terms of input and output. It is concluded that the short-run objective of establishing a data system to provide high-quality patient data that ae 21 comprehensive cancer centers in the United States, and its structure at the end of 3 years of data collection are described. Functions of the data system are detailed in terms of input and output. It is concluded that the short-run objective of establishing a data system to provide high-quality patient data that are comparable among cancer centers has been largely accomplished. Moreover, the very process of setting up the national data system has benefited the participating centers by upgrading their individual cancer registries. For the future, the goal is to realize the research potential of this new cooperative data collection mechanism, as well as the accumulating data themselves. Progress toward the long-term goal is just beginning.

Age distribution of Wilms' tumor: report from the National Wilms' Tumor Study.

Median ages at diagnosis were 36.5 and 42.5 mo for 1523 males and 1678 females with unilateral Wilms' tumor registered between October 1969 and December 1985; they were 23.5 mo and 30.5 mo for 100 males and 141 females with bilateral disease. The median age for multicentric, unilateral cases was intermediate between the bilateral and unicentric medians. Patients with hemihypertrophy in addition to their Wilms' tumor had a typical age distribution, whereas those with aniridia or characteristic genitourinary anomalies were substantially younger. Patients with perilobar nephroblastomatosis had a median age of 35.5 mo and those with intralobar nephroblastomatosis, a median age of 18.5 mo. Most of the bilateral disease occurred in the presence of one or both of these precursor lesions. These findings suggest heterogeneity in the pathogenesis of Wilms' tumor, having implications for genetic counselling, and call into question certain aspects of Knudson and Strong's two-stage mutational model for the origin of Wilms' tumor.

Renal failure in the Denys-Drash and Wilms' tumor-aniridia syndromes.

Nearly 6000 patients enrolled in four clinical trials of the National Wilms' Tumor Study Group during 1969-1995 were followed until death or for a median of 11.0 years of survival for the onset of renal failure (RF). Thirteen of 22 patients with Denys-Drash syndrome and 10 of 46 patients with the Wilms' tumor aniridia syndrome developed RF. The cumulative risks of RF at 20 years from Wilms' tumor diagnosis were 62% and 38%, respectively. Only 21 cases of RF were observed among 5358 patients with unilateral disease who did not have characteristic congenital genitourinary anomalies, and their risk was <1%. Although other explanations cannot be completely excluded, the high rate of RF in patients with the aniridia syndrome challenges the view that nephropathy is associated uniquely with missense mutations in the WT1 gene. It suggests the possibility of a further gradation in the spectrum of phenotypes associated with different WT1 mutations. Patients with Wilms' tumor and aniridia or genitourinary abnormalities should be followed closely throughout life for signs of nephropathy or RF.

Increased birth weights of National Wilms' Tumor Study patients suggest a growth factor excess.

An analysis of over 1800 patients with Wilms' tumor revealed significantly higher birth weights than newborns in the general United States population. The highest birth weights were found not only in patients diagnosed with the Beckwith-Wiedemann syndrome (mean, 3.78 kg), as had been expected, but also in those with hemihypertrophy (3.80 kg) or perilobar nephrogenic rests (3.56 kg) in addition to their Wilms' tumor. The birth weights of Wilms' tumor patients with intralobar nephrogenic rests (3.43 on average kg) and of those without associated anomalies (3.45 kg) were slightly but still significantly higher on average than national birthweights (overall mean, 3.35 kg) adjusted for gender, race, and year of birth in each subgroup. Birth weights of children with aniridia and Wilms' tumor (2.99 kg) were lower than the national mean. Among more than 3000 patients with Wilms' tumor, heights and weights at diagnosis were significantly higher for the subgroups of patients with Beckwith-Wiedemann syndrome or hemihypertrophy, and height was lower for those with aniridia or characteristic genitourinary anomalies, when compared to other patients with Wilms' tumor. These data suggest prenatal effects of growth factors on the development of Wilms' tumors, or vice versa, and provide further epidemiological support for heterogeneity in the pathogenesis of Wilms' tumors associated with perilobar nephrogenic rests versus intralobar nephrogenic rests.

Loss of heterozygosity for chromosomes 16q and 1p in Wilms' tumors predicts an adverse outcome.

We have prospectively analyzed Wilms' tumors from 232 patients registered on the National Wilms' Tumor Study for loss of heterozygosity (LOH) on chromosomes 11p, 16q, and 1p. These chromosomal aberrations were found in 70 (33%), 35 (17%), and 21 (12%) of the informative cases, respectively. LOH for two of these regions occurred in only 25 cases, and only one tumor harbored LOH at all three sites. There was no statistically significant association between LOH at any of the three regions and either the stage or histological classification of the tumor. Patients with tumor-specific LOH for chromosome 16q had relapse rates 3.3 times higher (P = 0.01) and mortality rates 12 times higher (P < 0.01) than patients without LOH for chromosome 16q. These differences remained when adjusted for histology or for stage. Patients with LOH for chromosome 1p had relapse and mortality rates three times higher than those for patients without LOH for chromosome 1p, but these results were not statistically significant. In contrast, LOH for chromosome 11p had no effect on measures of outcome. These molecular markers may serve to further stratify Wilms' tumor patients into biologically favorable and unfavorable subgroups, allowing continued use of the clinical trial mechanism in the study of Wilms' tumor.

High telomerase reverse transcriptase (hTERT) messenger RNA level correlates with tumor recurrence in patients with favorable histology Wilms' tumor.

Telomerase is a reverse transcriptase that maintains chromosome ends, compensating for the progressive loss of DNA that occurs during replication. High telomerase enzyme activity is an unfavorable prognostic feature for several types of cancers. We investigated whether telomerase level predicts outcome for patients with the pediatric renal malignancy Wilms' tumor. In a case-cohort study of 78 patients with favorable histology Wilms' tumor, we compared tumor telomerase levels in patients with and without eventual recurrence. Three measures of telomerase were used: (a) telomerase enzyme activity; (b) expression of hTR, the RNA component of telomerase; and (c) mRNA expression of hTERT, the gene that encodes the catalytic component of the enzyme. Of the evaluable samples, 81% had detectable telomerase activity, 97% had detectable hTERT transcript, and 100% had detectable hTR. Weak correlations were observed between telomerase activity and hTR level (r = 0.34, P = 0.02) and between telomerase activity and hTERT mRNA level (r = 0.32, P = 0.04). Of the variables assessed, only hTERT mRNA expression correlated with outcome. The median hTERT mRNA level in tumors with recurrence was higher than that in tumors without recurrence (1.42 versus 0.97 units, P = 0.023, Wilcoxon). Univariate analysis of hTERT mRNA level as a continuous variable suggested that each unit increase in hTERT mRNA level increased the risk of recurrence (RR) by a factor of 1.66 [95% confidence interval (CI), 1.2-2.3; P < 0.005]. Compared with tumors with hTERT mRNA levels of 0-1 units, tumors with hTERT mRNA levels of 1-2 units had a RR of 2.72 (95% CI, 0.91-8.13; P = 0.074), and tumors with hTERT mRNA levels >2 units had a RR of 6.40 (95% CI, 1.49-27.67, P = 0.013). Multivariate analysis of hTERT mRNA level as a predictor of recurrence, adjusted for tumor stage and age at diagnosis, revealed a RR of 1.48 (95% CI, 0.9-2.6; P = 0.16). Measurement of hTERT mRNA level may, therefore, enable clinicians to identify a population of patients at high risk for recurrence and to adjust their therapy accordingly. A larger study will be necessary to determine whether hTERT expression is an independent prognostic indicator. Further biological investigation is warranted to discern whether the link between high hTERT expression and unfavorable prognosis is causative or correlative.

Wilms' tumor and paternal occupation.

A case-control study was conducted to examine the relationship between Wilms' tumor and paternal occupational exposures. The case group consisted of 200 children diagnosed as having Wilms' tumor who were registered at selected National Wilms' Tumor Study institutions during the period June 1, 1984, to May 31, 1986. Disease-free controls were matched to each case using a random digit dialing procedure. The parents of cases and controls completed a self-administered questionnaire. There was no consistent pattern of increased risk for paternal occupational exposure to hydrocarbons or lead found in this study. However, certain paternal occupations were found to have an elevated odds ratio (OR) of Wilms' tumor, including vehicle mechanics, auto body repairmen, and welders. Offspring of fathers who were auto mechanics had a 4- to 7-fold increased risk of Wilms' tumor for all 3 time periods. The largest increased odds ratio for auto mechanics was in the preconception period [OR = 7.58; 95% confidence interval (CI) = 0.90-63.9]. Welders had a 4- to 8-fold increased odds ratio, with the strongest association during pregnancy (OR = 8.22; CI = 0.95-71.3). Although chance cannot be excluded as a possible explanation, association of Wilms' tumor with these occupations has been reported in previous studies. Further study is needed to provide data on the specific occupational exposures involved.

Toxic deaths in the Second National Wilms' Tumor Study.

There were 122 deaths among 803 children registered, randomized, and followed in the second National Wilms' Tumor Study; 17 occurred in children apparently free of disease and were attributable to causes other than tumor progression. Seven deaths were attributed to infection during periods of drug-induced leukopenia; four were due to liver failure; and one each was attributable to radiation pneumonopathy, intestinal obstruction, renal failure, myocardial disease, and encephalopathy. The cause of one death was unexplained. Of particular concern were four (of 47) infants under one year of age with group I or II disease who had toxic deaths. Subsequent to these experiences the doses of all chemotherapeutic agents were reduced by 50% for infants under one year of age. No deaths from toxicity were observed thereafter in infants. An analysis of the therapeutic effect of this dose reduction showed three of 47 relapsed on full dose and five of 54 on half dose. The difference is not statistically significant. This report is a further demonstration of the potentially serious vulnerability of infants to standard doses of anticancer drugs even when they are calculated on a per kilogram basis.

Chemotherapy-related toxicity in infants treated according to the Second National Wilms' Tumor Study.

Babies under 12 months of age have been included in the National Wilms' Tumor Study (NWTS) series of clinical trials. Undue chemotherapy-related toxicity was encountered early during the course of the second NWTS. The prescribed doses of actinomycin D (AMD), vincristine (VCR), and Adriamycin ([ADR] doxorubicin; Adria Laboratories, Columbus, OH) were therefore halved. The frequency of severe hematologic toxic episodes was reduced (30 of 64 or 47% for babies receiving full doses [FD], and six of 48 or 13% for those given reduced doses [RD]). Similar reductions in pulmonary and hepatic effects were noted, and treatment-related deaths were reduced from 6% to 0 for the FD and RD samples, respectively. These frequencies among RD babies were similar to those encountered in 530 older children administered FD. Reduction of dose did not compromise therapeutic effectiveness.

Treatment outcomes in patients less than 2 years of age with small, stage I, favorable-histology Wilms' tumors: a report from the National Wilms' Tumor Study.

PURPOSE: Retrospective analyses were performed to determine the effect of tumor weight and therapy modifications on outcome in patients less than 2 years of age with stage I favorable-histology Wilms' tumors. PATIENTS AND METHODS: The 4-year relapse-free and overall survival percentages for patients randomized to different treatment regimens in National Wilms' Tumor Studies (NWTS)-1, -2, and -3 were calculated and compared. RESULTS: The 4-year relapse-free survival percentages of patients whose specimen weight was less than 550 g were found to be 89.1% on NWTS-1, 96.0% on NWTS-2, and 93.2% on NWTS-3. There was no evidence that the relapse-free survival of these patients had improved over time (P value for trend = .99). The 4-year relapse-free survival percentage for similar age and stage patients whose specimen weight was 550 g or greater was significantly poorer than that of patients with smaller tumors (P = .02). CONCLUSION: Changes in the NWTS treatment regimens over a period of more than 20 years have not improved on the excellent prognosis of patients who are less than 2 years of age at diagnosis and who have a stage I, favorable-histology Wilms' tumor with specimen weight less than 550 g. These data could be used as the basis for a future trial in which a subgroup of such patients is treated with nephrectomy only.

Prognostic factors for children with recurrent Wilms' tumor: results from the Second and Third National Wilms' Tumor Study.

The characteristics of 367 stage I-IV National Wilms' Tumor Study (NWTS) children who relapsed after initial treatment for unilateral disease in the second and third NWTS trials (NWTS-2 and -3) were analyzed to identify features predictive of survival. Although modifications in initial therapy resulted in a lower rate of first relapse in these two studies compared with NWTS-1, all previously identified prognostic factors after relapse remained statistically significant predictors of survival. Tumor histology, length of initial remission, initial therapy with two v three drugs, and site of relapse each were independently predictive of postrelapse survival. The 3-year postrelapse survival for all 367 patients was 30% +/- 3%; however, subgroups classified by these prognostic factors were identified that had 3-year postrelapse survival rates of greater than 40%. These subgroups included patients who had tumors of favorable histology (FH) that recurred (1) only in the lungs, (2) in the abdomen when radiotherapy (RT) was not initially given, or (3) that were originally stage I, (4) that were originally treated with only two drugs, or (5) that recurred 12 months or more after diagnosis. These results were achieved with the use of standard treatments, ie, surgery, RT, and chemotherapy using dactinomycin (AMD), vincristine (VCR), and Adriamycin [( ADR] doxorubicin; Adria Laboratories, Columbus, OH). It is suggested that patients in these groups might be managed with aggressive use of conventional therapies until newer chemotherapeutic agents and drug combinations are shown to be more effective. Patients with adverse prognostic features at relapse have a poor survival expectancy with standard measures. Salvage attempts for these patients are better based on experimental approaches.

Characteristics and outcome of children with Beckwith-Wiedemann syndrome and Wilms' tumor: a report from the National Wilms Tumor Study Group.

PURPOSE: Children with Beckwith-Wiedemann syndrome (BWS) are at increased risk for developing Wilms' tumor (WT). We reviewed the National Wilms Tumor Study Group (NWTSG) records to assess clinical characteristics and outcome of patients with WT and BWS. METHODS: In the NWTSG, treating clinicians were asked to report, for each enrolled patient, whether the patient had BWS. Between 1980 and 1995, 4,669 patients were treated on two consecutive NWTSG protocols (NWTS 3 and NWTS 4). We retrospectively reviewed the clinical characteristics and treatment outcomes of BWS patients compared with patients with WT without BWS. RESULTS: Fifty-three children enrolled onto NWTS 3 and 4 were reported to have BWS. BWS patients were more likely to present with lower-stage tumors (P =.0001), with more than half (27 of 53) presenting with stage I disease. The overall treatment outcomes for the BWS patients were nearly identical to those without BWS, with overall survival at 4 years from diagnosis at 89% and 90%, respectively. Overall, 21% of the patients with BWS had bilateral disease, either at diagnosis (nine of 53) or as metachronous contralateral recurrence (two of 53). BWS patients enrolled onto NWTS 4 had smaller tumors than those enrolled onto NWTS 3 (P =.02), a trend not seen in the non-BWS patients. CONCLUSION: Like children without BWS, children with BWS and WT have an excellent prognosis with modern treatment regimens. There is a high risk of bilateral disease, and increasingly smaller tumors are being detected. This suggests that a national trial assessing the role of ultrasound screening followed by nephron-sparing surgery for some patients may be appropriate.

Severe hepatic toxicity after treatment with vincristine and dactinomycin using single-dose or divided-dose schedules: a report from the National Wilms' Tumor Study.

To evaluate the effect of dactinomycin (AMD) dose and schedule on the frequency of severe hepatic toxicity in unirradiated National Wilms' Tumor Study-4 (NWTS-4) patients, we reviewed the records of 154 children randomized to single-dose AMD and 176 children randomized to divided-dose AMD administration. All the children also received vincristine in identical dose schedules for the first 10 weeks. The frequency of severe hepatic toxicity encountered in the early weeks of therapy was 14.3% (five of 35) among patients treated with 60 micrograms/kg of AMD, 3.7% (four of 108) among patients given 45 micrograms/kg, and 2.8% (five of 176) among patients treated with 15 micrograms/kg per dose times five doses (P = .025). The data suggest an increased frequency of severe hepatic toxicity with the higher, single-dose schedule of administration. However, the frequency of severe hepatic toxicity among the patients in the two remaining groups is markedly higher than the 0.4% observed among similar unirradiated patients in NWTS-3. The relationship of this toxicity to factors such as anesthetic agents, blood transfusions, intercurrent viral infection, or other presently unrecognized causes can be further evaluated only with a detailed investigation such as a case-control study.

Second malignant neoplasms following treatment for Wilm's tumor: a report from the National Wilms' Tumor Study Group.

PURPOSE: The study was undertaken to determine the incidence of second malignant neoplasms (SMNs) in patients treated for Wilms' tumor and demonstrate how the incidence varied with the initial treatment protocol. PATIENTS AND METHODS: Between October 1969 and December 1991, 5,278 assessable patients were enrolled onto the National Wilms' Tumor Study (NWTS) and by the end of 1993 had contributed 39,461 person-years to a follow-up study. Expected numbers of second cancers were calculated by applying national incidence rates to person-years classified by age, sex, and calendar year. RESULTS: Forty-three SMNs were observed, whereas only 5.1 were expected (standardized incidence ratio [SIR], 8.4; 95% confidence interval [CI], 6.1 to 11.4). Fifteen years after the Wilms' tumor diagnosis, the cumulative incidence of a SMN was 1.6% and increasing steadily. Abdominal irradiation received as part of the initial therapy increased the risk of a SMN (SIR, 1.43/10 Gy; 95% CI, 1.13 to 1.81). Doxorubicin potentiated the radiation effect. Among 234 patients who received doxorubicin and greater than 35 Gy of abdominal radiation, eight SMNs were observed, whereas only 0.22 were expected (SIR, 36; 95% CI, 16 to 72). Treatment for relapse further increased the SMN risk by a factor of 4 to 5. CONCLUSION: These results demonstrate the importance of current efforts to limit the use of intensive chemotherapy and radiation therapy, which are now applied only to patients with the most aggressive disease. Continuing close surveillance of the great majority of Wilms' tumor patients who become long-term survivors is essential for early diagnosis of SMNs and other late sequelae of therapy.

Treatment of children with stages II to IV anaplastic Wilms' tumor: a report from the National Wilms' Tumor Study Group.

PURPOSE: To evaluate the effect of the combination of vincristine, dactinomycin, and doxorubicin with (regimen J) or without (regimen DD-RT) cyclophosphamide on the relapse-free survival of children with stages II to IV Wilms' tumor and focal or diffuse anaplasia. PATIENTS AND METHODS: We reviewed the clinical courses of all randomized patients from National Wilms' Tumor Study (NWTS)-3 and NWTS-4 with stages II to IV anaplastic Wilms' tumor, and determined the 4-year relapse-free survival rate separately for those with focal or diffuse anaplasia. Anaplasia was evaluated using newly developed topographic definitions for focal and diffuse anaplasia. RESULTS: The 4-year relapse-free survival rate for five children with focal anaplasia who received regimen DD-RT was 80.0%, compared with 100.0% for eight children who received regimen J (P = .68). The 4-year relapse-free survival rate for 29 children with diffuse anaplasia treated with regimen DD-RT was 27.2%, compared with 54.8% for 30 children treated with regimen J (P = .02). CONCLUSION: We conclude that children with focal anaplasia have an excellent prognosis when treated with vincristine, doxorubicin, and dactinomycin. The addition of cyclophosphamide to the three-drug treatment regimen improved the 4-year relapse-free survival rate of children with stage II to IV diffuse anaplasia. This result suggests that further intensification of the treatment regimen for children with diffuse anaplasia may result in an additional improvement in prognosis.