Frank Berthold

Classification and diagnostic prediction of cancers using gene expression profiling and artificial neural networks.

The purpose of this study was to develop a method of classifying cancers to specific diagnostic categories based on their gene expression signatures using artificial neural networks (ANNs). We trained the ANNs using the small, round blue-cell tumors (SRBCTs) as a model. These cancers belong to four distinct diagnostic categories and often present diagnostic dilemmas in clinical practice. The ANNs correctly classified all samples and identified the genes most relevant to the classification. Expression of several of these genes has been reported in SRBCTs, but most have not been associated with these cancers. To test the ability of the trained ANN models to recognize SRBCTs, we analyzed additional blinded samples that were not previously used for the training procedure, and correctly classified them in all cases. This study demonstrates the potential applications of these methods for tumor diagnosis and the identification of candidate targets for therapy.

Revisions of the international criteria for neuroblastoma diagnosis, staging, and response to treatment.

PURPOSE AND METHODS: Based on preliminary experience, there was a need for modifications and clarifications in the International Neuroblastoma Staging System (INSS) and International Neuroblastoma Response Criteria (INRC). In 1988, a proposal was made to establish an internationally accepted staging system for neuroblastoma, as well as consistent criteria for confirming the diagnosis and determining response to therapy (Brodeur GM, et al: J Clin Oncol 6:1874-1881, 1988). A meeting was held to review experience with the INSS and INRC and to revise or clarify the language and intent of the originally proposed criteria. Substantial changes included a redefinition of the midline, restrictions on age and bone marrow involvement for stage 4S, and the recommendation that meta-iodobenzylguanidine (MIBG) scanning be implemented for evaluating the extent of disease. Other modifications and clarifications of the INSS and INRC are presented. In addition, the criteria for the diagnosis of neuroblastoma were modified. ...

International criteria for diagnosis, staging, and response to treatment in patients with neuroblastoma.

Neuroblastoma is one of the most common tumors in childhood. However, it often has been difficult to compare clinical and laboratory studies of this disease due to a lack of uniform criteria for diagnosis, staging, and response. An international group of conferees addressed each of these issues and reached a consensus. Specific criteria for making a diagnosis of neuroblastoma are defined. A new neuroblastoma staging system is proposed that takes into account the most important elements of current but incompatible systems. Finally, criteria for response to treatment are standardized. The criteria proposed herein represent an international consensus of essentially every major pediatric oncology group or organization in the United States, Europe, and Japan. The staging system should be referred to as the International Neuroblastoma Staging System, and the response criteria as the International Neuroblastoma Response Criteria. Implementation of these criteria will greatly facilitate the comparison of clinical and laboratory studies by different groups and countries. Furthermore, these criteria should serve as a foundation on which future modifications or improvements can be based.

Postoperative neoadjuvant chemotherapy before radiotherapy as compared to immediate radiotherapy followed by maintenance chemotherapy in the treatment of medulloblastoma in childhood: results of the german prospective randomized trial hit ’91

Purpose: The German Society of Pediatric Hematology and Oncology (GPOH) conducted a randomized, prospective, multicenter trial (HIT ’91) in order to improve the survival of children with medulloblastoma by using postoperative neoadjuvant chemotherapy before radiation therapy as opposed to maintenance chemotherapy after immediate postoperative radiotherapy. Methods and Materials: Between 1991 and 1997, 158 patients were enrolled and 137 patients randomized. Seventy-two patients were allocated to receive neoadjuvant chemotherapy before radiotherapy (arm I, investigational). Chemotherapy consisted of ifosfamide, etoposide, intravenous high-dose methotrexate, cisplatin, and cytarabine given in two cycles. In arm II (standard arm), 65 patients were assigned to receive immediate postoperative radiotherapy, with concomitant vincristine followed by 8 cycles of maintenance chemotherapy consisting of cisplatin, CCNU, and vincristine (“Philadelphia protocol”). All patients received radiotherapy to the craniospinal axis (35.2 Gy total dose, 1.6 Gy fractionated dose / 5 times per week followed by a boost to posterior fossa with 20 Gy, 2.0 Gy fractionated dose). Results: During chemotherapy Grade III/IV infections were predominant in arm I (40%). Peripheral neuropathy and ototoxicity were prevailing in arm II (37% and 34%, respectively). Dose modification was necessary in particular in arm II (63%). During radiotherapy acute toxicity was mild in the majority of patients and equally distributed in both arms. Myelosuppression led to a mean prolongation of treatment time of 11.5 days in arm I and 7.5 days in arm II, and interruptions in 35% of patients in arm I. Quality control of radiotherapy revealed correct treatment in more than 88% for dose prescription, more than 88% for coverage of target volume, and 98% for field matching. At a median follow-up of 30 months (range 1.4–62 months), the Kaplan-Meier estimates for relapse-free survival at 3 years for all randomized patients were 0.70 ± 0.08; for patients with residual disease: 0.72 ±0.06; without residual disease: 0.68 ± 0.09; M0: 0.72 ± 0.04; M1: 0.65 ± 0.12; and M2/3: 0.30 ± 0.15. For all randomized patients without M2/3 disease: 0.65± 0.05 (arm I) and 0.78 ± 0.06 (arm II) (p < 0.03); patients between 3 and 5.9 years: 0.60 ± 0.13 and 0.64 ± 0.14, respectively, but patients between 6 and 18 years: 0.62 ± 0.09 and 0.84 ± 0.08, respectively (p < 0.03). In a univariate analysis the only negative prognostic factors were M2/3 disease (p < 0.002) and an age of less than 8 years (p < 0.03). Conclusions: Maintenance chemotherapy would seem to be more effective in low-risk medulloblastoma, especially in patients older than 6 years of age. Neoadjuvant chemotherapy was accompanied by increased myelotoxicity of the subsequent radiotherapy, causing a higher rate of interruptions and an extended overall treatment time. Delayed and/or protracted radiotherapy may therefore have a negative impact on outcome. M2/3 disease was associated with a poor survival in both arms, suggesting the need for a more intensive treatment. Young age and M2/3 stage were negative prognostic factors in medulloblastoma, but residual or M1 disease was not, suggesting a new stratification system for risk subgroups. High quality of radiotherapy may be a major contributing factor for the overall outcome.

Myeloablative megatherapy with autologous stem-cell rescue versus oral maintenance chemotherapy as consolidation treatment in patients with high-risk neuroblastoma: a randomised controlled trial

BACKGROUND Myeloablative megatherapy is commonly used to improve the poor outlook of children with high-risk neuroblastoma, yet its role is poorly defined. We aimed to assess whether megatherapy with autologous stem-cell transplantation could increase event-free survival and overall survival compared with maintenance chemotherapy. METHODS 295 patients with high-risk neuroblastoma (ie, patients with stage 4 disease aged older than 1 year or those with MYCN-amplified tumours and stage 1, 2, 3, or 4S disease or stage 4 disease and <1 year old) were randomly assigned to myeloablative megatherapy (melphalan, etoposide, and carboplatin) with autologous stem-cell transplantation (n=149) or to oral maintenance chemotherapy with cyclophosphamide (n=146). The primary endpoint was event-free survival. Secondary endpoints were overall survival and the number of treatment-related deaths. Analyses were done by intent to treat, as treated, and treated as randomised. FINDINGS Intention-to-treat analysis showed that patients allocated megatherapy had increased 3-year event-free survival compared with those allocated maintenance therapy (47% [95% CI 38-55] vs 31% [95% CI 23-39]; hazard ratio 1.404 [95% CI 1.048-1.881], p=0.0221), but did not have significantly increased 3-year overall survival (62% [95% CI 54-70] vs 53% [95% CI 45-62]; 1.329 [0.958-1.843], p=0.0875). Improved 3-year event-free survival and 3-year overall survival were also recorded for patients given megatherapy in the as-treated group (n=212) and in the treated-as-randomised group (n=145). Two patients died from therapy-related complications during induction treatment. No patients given maintenance therapy died from acute treatment-related toxic effects. Five patients given megatherapy died from acute complications related to megatherapy. INTERPRETATION Myeloablative chemotherapy with autologous stem-cell transplantation improves the outcome for children with high-risk neuroblastoma despite the raised risk of treatment-associated death.

Customized Oligonucleotide Microarray Gene Expression–Based Classification of Neuroblastoma Patients Outperforms Current Clinical Risk Stratification

PURPOSE To develop a gene expression-based classifier for neuroblastoma patients that reliably predicts courses of the disease. PATIENTS AND METHODS Two hundred fifty-one neuroblastoma specimens were analyzed using a customized oligonucleotide microarray comprising 10,163 probes for transcripts with differential expression in clinical subgroups of the disease. Subsequently, the prediction analysis for microarrays (PAM) was applied to a first set of patients with maximally divergent clinical courses (n = 77). The classification accuracy was estimated by a complete 10-times-repeated 10-fold cross validation, and a 144-gene predictor was constructed from this set. This classifiers predictive power was evaluated in an independent second set (n = 174) by comparing results of the gene expression-based classification with those of risk stratification systems of current trials from Germany, Japan, and the United States. RESULTS The first set of patients was accurately predicted by PAM (cross-validated accuracy, 99%). Within the second set, the PAM classifier significantly separated cohorts with distinct courses (3-year event-free survival [EFS] 0.86 +/- 0.03 [favorable; n = 115] v 0.52 +/- 0.07 [unfavorable; n = 59] and 3-year overall survival 0.99 +/- 0.01 v 0.84 +/- 0.05; both P < .0001) and separated risk groups of current neuroblastoma trials into subgroups with divergent outcome (NB2004: low-risk 3-year EFS 0.86 +/- 0.04 v 0.25 +/- 0.15, P < .0001; intermediate-risk 1.00 v 0.57 +/- 0.19, P = .018; high-risk 0.81 +/- 0.10 v 0.56 +/- 0.08, P = .06). In a multivariate Cox regression model, the PAM predictor classified patients of the second set more accurately than risk stratification of current trials from Germany, Japan, and the United States (P < .001; hazard ratio, 4.756 [95% CI, 2.544 to 8.893]). CONCLUSION Integration of gene expression-based class prediction of neuroblastoma patients may improve risk estimation of current neuroblastoma trials.

Neuroblastoma: biology and molecular and chromosomal pathology

Neuroblastoma is the most frequently occurring solid tumour in children, with an incidence of 1.3 cases per 100000 children aged 0-14 years. Despite many advances during the past three decades, neuroblastoma has remained an enigmatic challenge to clinical and basic scientists. 20 years ago, the MYCN gene was found to be amplified in neuroblastomas, and research since then has focused on the search for other genetic markers. It has emerged that neuroblastoma cells, like cells of many other tumour types, often suffer from extensive, non-random genetic damage at multiple genetic loci. Elucidation of the exact molecular make-up of neuroblastomas will enable researchers to analyse how much specific markers, alone or in combination, can help to stratify disease in prospective studies; at present, stratification is based on age, stage, MYCN, and Shimada pathology. Neuroblastoma may be one of the first examples of the use of genetic tumour markers as a tool for defining tumour behaviour and to aid clinical staging.

Improved Treatment Results in High-Risk Pediatric Acute Myeloid Leukemia Patients After Intensification With High-Dose Cytarabine and Mitoxantrone: Results of Study Acute Myeloid Leukemia–Berlin-Frankfurt-Münster 93

PURPOSE To improve outcome in high-risk patients, high-dose cytarabine and mitoxantrone (HAM) was introduced into the treatment of children with acute myelogenous leukemia (AML) in study AML-BFM 93. Patients were randomized to HAM as either the second or third therapy block, for the purpose of evaluation of efficacy and toxicity. PATIENTS AND METHODS A total of 471 children with de novo AML were entered onto the trial; 161 were at standard risk and 310 were at high risk. After the randomized induction (daunorubicin v idarubicin), further therapy, with the exception of HAM, was identical in the two risk groups and also comparable to that in study Acute Myeloid Leukemia-Berlin-Frankfurt-Münster (AML-BFM) 87. RESULTS Overall, 387 (82%) of 471 patients achieved complete remission, and 5-year survival, event-free survival (EFS), and disease-free survival rates were 60%, 51%, and 62%, respectively. Idarubicin induction resulted in a significantly better blast cell reduction in the bone marrow on day 15. Estimated survival and probability of EFS were superior in study AML-BFM 93 compared with study AML-BFM 87 (P =.01, log-rank test). This improvement, however, was restricted to the 310 high-risk patients (remission rate and probability of 5-year EFS in study AML-BFM 93 v study AML-BFM 87: 78% v 68%, P =.007; and 44% v 31%, P =.01, log-rank test). Probability of 5-year EFS among standard-risk patients in study AML-BFM 93 was similar to that in study AML-BFM 87 (65% v 63%, P = not significant). Whether HAM was placed as the second or third therapy block was of minor importance. However, patients who received the less intensive daunorubicin treatment during induction benefited from early HAM. CONCLUSION Improved treatment results in children with high-risk AML in study AML-BFM 93 must be attributed mainly to the introduction of HAM.

Localized infant neuroblastomas often show spontaneous regression: results of the prospective trials NB95-S and NB97.

PURPOSE The excellent prognosis of localized neuroblastoma in infants, the overdiagnosis observed in neuroblastoma screening studies, and several case reports of regression of localized neuroblastoma prompted us to initiate a prospective cooperative trial on observation of localized neuroblastoma without cytotoxic treatment. PATIENTS AND METHODS For infants with localized neuroblastoma without MYCN amplification, chemotherapy was scheduled only in cases with threatening symptoms; otherwise, the tumor was either resected or observed by ultrasound and magnetic resonance imaging (MRI). RESULTS Of 340 eligible participants, 190 underwent resection, 57 were treated with chemotherapy, and 93 were observed with gross residual tumor. Of those 93 patients with unresected tumors, spontaneous regression was seen in 44, local progression in 28, progression to stage 4S in seven, and progression to stage 4 in four. Time to regression was quite variable, with first signs of regression noted 1 to 18 months after diagnosis and in 15 of 44 patients even after the first year of life. So far, complete regression was observed in 17 of 44 patients 4 to 20 months after diagnosis. Known clinical risk factors were not able to differentiate between patients with regression and regional or metastatic progression. Overall survival (OS; 3-year OS, 0.99 +/- 0.01) and metastases-free survival (rate at 3 years, 0.94 +/- 0.03) for patients with unresected tumors was excellent and was not different from patients treated with surgery or chemotherapy. CONCLUSION Spontaneous regression is regularly seen in infants with localized neuroblastoma and is not limited to the first year of life. A wait-and-see strategy is justified in those patients.

Somatic mutations of WNT/wingless signaling pathway components in primitive neuroectodermal tumors

Primitive neuroectodermal tumors (PNETs) represent the most frequent malignant brain tumors in childhood. The majority of these neoplasms occur in the cerebellum and are classified as medulloblastomas (MB). Most PNETs develop sporadically; however, their incidence is highly elevated in patients carrying germline APC gene mutations. The APC gene encodes a central component of the WNT/wingless developmental signaling pathway. It regulates the levels of cytoplasmic β‐catenin protein that plays a central role in neural development and cell proliferation. We analyzed 87 sporadic PNETs and 10 PNET cell lines for mutations of the APC gene and β‐catenin (CTNNB1) gene using single strand conformational polymorphism (SSCP) and sequencing analysis. We examined the mutation cluster region of APC (codons 1255–1641) for germline variants and somatic mutations. The medulloblastoma cell line MHH‐MED‐2 carried a Glu1317Gln missense germline variant and a sporadic MB sample showed a somatic Pro1319Leu substitution. Mutational analysis of exon 3 of CTNNB1 uncovered 4 PNETs (4.8%) with somatic missense mutations. These mutations caused amino acid substitutions in 3 of 80 medulloblastomas (Ser33Phe, Ser33Cys and Ser37Cys) and 1 of 4 supratentorial PNETs (Gly34Val). All mutations affected GSK‐3β phosphorylation sites of the degradation targeting box of β‐catenin and resulted in nuclear β‐catenin protein accumulation. Deletions of CTNNB1 were not detected by PCR amplification with primers spanning exons 1–5. Our data indicate that inappropriate activation of the WNT/wingless signaling pathway by mutations of its components may contribute to the pathogenesis of a subset of PNETs.