Improved therapy for children with acute lymphoblastic leukemia and unfavorable presenting features: a follow-up report of the Childrens Cancer Group Study CCG-106.

PURPOSE: On past Childrens Cancer Group (CCG) trials, children with acute lymphoblastic leukemia and unfavorable presenting features had obtained an event-free survival (EFS) rate of no better than 50%. Following promising pilot experience, this study was conducted to determine the benefit and morbidity of two intensive experimental regimens, Reg A, based on the Berlin-Frankfurt-Münster (BFM) 1976 regimen, and Reg B, the New York regimen. PATIENTS AND METHODS: Between February 1983 and November 1984, 217 eligible children with acute lymphoblastic leukemia and unfavorable presenting features were entered and randomly assigned to receive Reg A, Reg B, or Reg C, the control regimen. Assignment to Reg C was halted in November 1984 after interim analyses showed an inferior outcome. Subsequently, between November 1984 and March 1987, an additional 328 patients were randomly allocated to receive Reg A or Reg B. RESULTS: The 7-year EFS rate was 63% (+/- 6%, 1 SD) for Reg A, 61% (+/- 6%) for Reg B, and 40% (+/- 6%) for Reg C (P < .006). The difference between Reg A or Reg B and Reg C remained greater than 20 percentage points for EFS at 7 years and 15 percentage points for survival. Relative to Reg C, patients on Reg A accrued 16.3 additional days of hospitalization on average and, on Reg B, 20.2 days. EFS and survival were similar on Reg A and Reg B, but Reg B required more days of parenteral therapy and greater exposure to anthracyclines and alkylating agents. CONCLUSION: Both Reg A and Reg B provided a better outcome than Reg C for children with acute lymphoblastic leukemia and unfavorable presenting features. Outcomes on Reg A and Reg B were similar. Use of the more effective but more toxic regimens resulted in 78 additional hospital days per relapse prevented on Reg A and 101 days on Reg B. The current CCG trial for this population builds on Reg A.

Microsatellite instability assessment in prediction of drug resistance in childhood Burkitt's and large cell diffuse malignant non-Hodgkin lymphoma (MNHL).

BACKGROUND: Genomic instability may, especially with DNA directed treatment, be associated with increased therapeutic response; absence may be associated with drug resistance. In childhood MNHL, drug response is variable. At present the degree of presence of microsatellite variation, i.e., intrinsic DNA instability is not known. AIMS: To determine presence and range of microsatellite variability in common childhood MNHL. METHODS: 1.3.1. Study Populations. Consecutive, unselected (1976-96) cases of childhood Large Cell diffuse, N = 16; (9T,7B), age range 1y5m-16y8m; Burkitt's Lymphoma, n = 13, age range 4y2m-14y. Non-malignant/pre-treatment tissue of 20 cases, 13 LC, 7 Burkitt's MNHL. 1.3.2. Molecular Pathology. Routine DNA extraction, amplifications at loci D3S 1304 and D3S1537 (both closely distal to VHL, tumour suppressor gene); ELN gene D7S1870; IFNA D1S243 (1p36) which show microsatellite variation. Isotopic labelling in amplification, non-denaturing gel electrophoresis, autoradiography. RESULTS: Microsatellite variability was found 3/16 LC and 2/13 Burkitt's MNHL. LC MNHL, 4 abnormal areas: n = 1, 3 abnormal areas: n = 1, 2 abnormal areas n = 1; Burkitt's MNHL, 3 abnormal areas: n = 1, 1 abnormal area n = 1. No variability was found in the normal (constitutional) DNA of any of the 20 patients studied. CONCLUSIONS: Microsatellite variability occurred in 5/29 patients with common types of childhood MNHL, indicating a limited contribution to reduced drug resistance through this mechanism.

High resolution confocal laser scanning microscopy analysis of DNA excision repair capability in small volume marrow samples exposed to DNA directed treatment moieties.

BACKGROUND: Assessment of resistance to drug moieties in tissue culture is complicated by limited sample, clonal selection and alteration of cycling fraction and cycle duration in clonally mixed lesions. DNA damage assessment by single cell gel electrophoresis (SCGE) of excised DNA is limited by non-linear analysis in fluorescent light microscopy. Confocal Laser Scanning Microscopy (CLSM) with high N.A. magnification allows for quantitation of total excised DNA fragment size distribution but is still limited by the large volume required for labour intensive SCGE, precluding multi-exposure clinical testing. AIMS: To optimise sample requirement for SCGE and CLS. METHODS: Standard slide mounted bed gels were punched with multiple coded 6 mm wells and filled with suspensions of cells subjected to drug/concentration variations. After SCGE, 30 microns frozen sections were prepared of each well and mounted in ethidium bromide solution on multi-well hydrophilic slides to allow for short working distance of high resolution CLSM in a Zeiss Axiovert L410 SM. Testing for feasibility, reproducibility and consistency used both cultured standard leukaemic cell lines, normal human control marrow and clinical samples. RESULTS AND CONCLUSION: Multiple well SCGE followed by frozen section, high resolution CLSM allows for rapid analysis of high numbers of multiple drug exposure permutations clinically required.

Apoptotic fraction in childhood ALL assessed by DNA in situ labelling is ploidy independent.

BACKGROUND: Apoptotic cell fraction and presence or degree of aneuploidy may both affect treatment outcome in childhood acute lymphoblastic leukaemia (ALL), which is largely defined by drug resistance. Independence of the variables is at present not established. Until the development of in situ labelling of cells committed to the apoptotic pathway, the fraction of cells in apoptosis could not be determined objectively. AIM: To determine the relationship between apoptotic cell fraction and karyotype in childhood ALL using in situ labelling. METHODS: 1.3.1. Study Groups and Samples. Diagnostic, pretreatment bone marrow trephine and aspirate samples of 24 consecutive, unselected cases of childhood ALL were included in the study: Normal karyotype (n = 11, 5M,6F), high hyperdiploid aneuploidy (DNA index > 1.5, n = 7, 1M,6F), complex karyotypic anomalies (n = 6, 5M,1F). 1.3.2. Apoptotic Cell Labelling. In situ labelling of the 3'-OH ends of the apoptosis specific DNA (Klenow) fragment (Frag-EL, CalBiochem, USA). 1.3.3. Quantitation. Apoptotic cell fraction was established using 10 systematically random fields of > 20 nuclei. Results were tabled per group. After calculations of means, differences between groups were assessed using t-test. RESULTS: Apoptotic cell fraction, ranging from < 1 to 95%, did not differ statistically significant between the three study groups. CONCLUSION: Apoptotic cell fraction in childhood leukaemia is independent of ploidy status and euploid karyotypic anomalies.

PCNA bearing structures are retained in apoptotic phase of childhood ALL cell cycle.

BACKGROUND: Drug resistance to DNA directed therapy may depend on proliferative as well as apoptotic cell fraction. PCNA/Ki67 ratio excess, possibly reflecting DNA excision repair, is of additional interest to drug resistance in MTT testing. The cell cycle phase/antigen expression pattern in childhood acute lymphoblastic leukemia (ALL) is not known. AIMS: To study the relationship between nuclear expression of PCNA, Ki-67 and Frag-EL positivity in childhood ALL. METHODS: 1.3.1. Study Groups. Diagnostic bone marrow trephine biopsies of 32 consecutive unselected cases of childhood ALL were included in the study. 1.3.2. Immunohistochemistry. Commercially available Moab PCNA (PC10, DAKO, USA), Ki-67 (MM1, NovaCastra, UK) were used to label cycling cells in routinely processed 5 microns paraffin sections. 1.3.3. In-Situ Labelling of Apoptotic Cells. The 3'-OH ends of apoptosis specific DNA fragments were labelled in-situ on subsequent 10 microns sections (Frag-EL, CalBiochem, USA). 1.3.4. Quantitation. After blinding and randomisation, 10 systematic random fields of > 20 nuclei and nuclear size bias correction was used to determine positive nuclei fraction. RESULTS: While the sum of apoptotic and proliferative cell fraction (Ki-67 + Frag-EL%) equalled 100% in 5/32 cases, PCNA expression into at least the early phases of apoptosis ([%PCNA-%K-67] > [100-%Frag-EL] was found in 17/32 cases. CONCLUSIONS: PCNA/Ki67 ratio excess may not reflect DNA excision repair activity but rather slow degradation of antigen bearing structures limiting relevance to drug resistance study.

Apoptosis corrected proliferation fraction in childhood ALL is related to karyotype.

BACKGROUND: Tumour doubling time, a parameter in drug sensitivity testing, reflects both cell proliferation and apoptosis. Variable apoptosis fractions may explain the poor correlation of S-fraction and drug response. DNA aneuploidy (reflecting intrinsic DNA instability) may, by increasing apoptosis, affect drug response. AIM: To assess the relationship between apoptosis corrected proliferation fraction and DNA ploidy in childhood acute lymphoblastic leukemia (ALL). METHODS: 1.3.1. Study Groups. Thirty two consecutive, unselected diagnostic cases of childhood ALL were included in the study. 1.3.2. Karyotype. A normal karyotype was found in 15 cases (7M, 8F, age 8 m-12 yrs); high hyperdiploid aneuploidy (DNA index > 1.5) was found in 7 patients (1M, 7F, age 3-12 yrs) whereas complex karyotypic anomalies, but with 2n or near 2n DNA were present in 10 patients (7M, 3F, age 1 y 7 m -16 yrs). 1.3.3. Proliferation Fraction Assessment. Immunocytochemical demonstration of S-phase associated nuclear expression of the Ki-67 antigen (MM1, NovaCastra, UK). 1.3.4. Apoptosis Fraction Assessment. Binding of a horse radish peroxidase labelled DNA probe for the 3'-OH ends of apoptosis derived Klenow fragments (Frag-EL, CalBiochem, USA). 1.3.5. Quantitation. Computer assisted image analysis (Quantimet 570C), of 10 systematically random fields of a minimum of 20 nuclei each. A nuclear size bias correcting counting frame and rule were used to correct for cell proliferation associated nuclear volume increase and for the expected nuclear volume reduction resulting from apoptosis. RESULTS: Corrected for apoptosis, proliferation fraction was highest (mean 57.5%, range 1-100) in poor prognosis, complex karyotype anomalies. Good prognosis, high hyper diploidy showed significantly lower proliferation rates (mean 24.7%, range 12-40) (p < 0.01, t-test). CONCLUSION: Apoptosis corrected cell proliferation rate in childhood ALL is not independent of karyotype abnormality which may partly explain a relation to therapy response and prognosis.

Proliferation and apoptosis does not affect presenting white cell count in childhood ALL.

BACKGROUND: Treatment response/drug resistance in childhood acute lymphoblastic leukaemia (ALL) is related to presenting white cell count. This relationship might be explained by high proliferation fraction or by absence of significant apoptosis, but is presently unknown. AIMS: To study the relationship between proliferation and apoptosis in childhood ALL. METHODS: 1.3.1. Study Groups. Thirty consecutive, unselected cases of childhood ALL (15M,15F). White cell count varied from 1-200 at presentation. 1.3.2. Proliferation/Apoptosis Fraction. Immunocytochemical detection of Ki-67 (MM1, NovaCastra, UK) on 5 microns paraffin slides, Immunocytochemical detection of apoptosis specific DNA (Klenow) fragments by labelling of 3'-OH ends in 10 microns paraffin sections (Frag-EL, CalBiochem, USA). 1.3.3. Quantitation. Image analysis (Quantimet 570C) using nuclear size bias correcting counting frame and rule. Calculation of proliferation (%Ki-67) fraction and of apoptosis corrected proliferation fraction (%Ki-67/100-%apoptosis). RESULTS: Using both linear, logarithmic regression as well as power analysis, no relationship between variables was detected. CONCLUSION: Presenting white cell count is not related to apoptotic or cell proliferative activity or to net tumour growth defined by the balance between these two processes. The relationship to treatment resistance still requires explanation.

Differential kinetics of drug resistance in human leukaemic cells measured by SCGE/CLSM.

BACKGROUND: New analogues of DNA directed chemotherapy moieties are available for comparative efficacy testing in human neoplastic disease. In addition to MTT testing direct assessment of DNA excision repair activity after direct exposure of marrow cells may provide information on relative DNA effects in vitro. AIMS: To assess the ability of SCGE/high resolution CLSM to detect differences in drug resistance between human neoplastic cell lines in the DNA excision repair response to chemotherapy. METHODS: Eight human leukaemia samples (4 childhood, 4 adult) were exposed to 1 hour of single concentrations of daunorubicin, DaunoXome (courtesy NeXstar Pharmaceuticals Inc, USA), cyclophosphamide and 4-hydroperoxycyclophosphamide (4-HC, courtesy Dr. M. Colvin, Duke University, USA), followed by SCGE/high resolution CLSM with quantitation of total excised DNA. Differences between cases/drug moieties/exposures were analysed. RESULTS: Although generally equal effect dose levels for DaunoXome were lower than for standard daunorubicin, patients/individual neoplastic cells differed considerably in optimal dose levels. Conventional cyclophosphamide in comparison to 4-HC showed inconsistent results indicating considerable differences in the level of drug resistance to the conventional product. CONCLUSIONS: Direct testing for drug resistance patterns in DNA directed drug moieties by SCGE/CLSM reveals individual variability of human malignant cell lines warranting comparison with results of MTT testing and in-vivo patient response.

Demonstration of differences in drug resistance by direct testing of DNA excision repair activity following standard and liposomal daunorubicin exposure in normal paediatric marrow using high resolution CLSM.

BACKGROUND: High resolution Confocal Laser Scanning Microscopy (CLSM) may be applied to testing of drug resistance in vitro in clinical setting. Rapid analysis of DNA damage by precise quantitation of excised DNA in bone marrow samples exposed to potential treatment moieties directly after isolation but the relative sensitivity of the integrated method is as yet untested. AIMS: To test the clinical applicability of SCGE/high resolution CLSM for differences in drug resistance in marrow cells. METHODS: Cells from normal bone marrow samples were exposed for identical periods and at 4 concentrations to either 1 hour of standard Daunorubicin (.5, 1, 1.5, 2 micrograms/ml) or 8 hours DaunoXome (courtesy of NeXstar Inc, USA) (.05, .1, .15, .2 microgram/ml). After 2 and 6 hours recovery, cells were harvested for SCGE, randomization, analysis of tail length, total excised DNA and fragment size distribution using high resolution CLSM. RESULTS: Tail length and fragment size distribution was not, but total excised DNA was significantly increased after 0.1 microgram/ml Liposomal Daunorubicin (DaunoXome) compared to 1.0 microgram/ml Daunorubicin. CONCLUSION: SCGE/high resolution CLSM effectively demonstrated differences in Daunorubicin resistance of human marrow cells to alternative formulations. The method has potential for use in clinical testing of neoplastic cell drug resistance.

Restricted genetic heterogeneity in families of patients with acute lymphocytic leukemia.

HLA antigens of the A and B loci were determined on the lymphocytes of 30 patients with acute lymphocytic leukemia (A.L.L.), as well as all of their mothers and 26 of the fathers. Seven of the 26 parents shared a common haplotype. This incidence of 269 per 1,000 contrasts with an expected incidence of 90.7 per 1,000, calculated from haplotype frequencies in a North American population (X2=7.61, P smaller.than 0.01) and a frequency of one in 27 in parents of patients with renal failure in the local population (X2=3.91, P smaller than 0.05). There was no statistical difference between the latter group and the North American controls (X2=0.39, P greater than 0.10). This suggests that the genetic background of a large proportion of patients with A.L.L. has restricted heterogeneity, presumably leading to the increased expression of leukemia associated recessive genes.

Successful reinduction therapy with amsacrine and cyclocytidine in acute nonlymphoblastic leukemia in children. A report from the Childrens Cancer Study Group.

Amsacrine (AMSA) and cyclocytidine were studied as retrieval therapy in 122 pediatric patients with acute nonlymphoblastic leukemia (ANLL). Patients either failed to achieve sustained initial remissions or were in relapse. Induction therapy consisted of intravenous (IV) AMSA (75 mg/m2) from days 1 to 5 and subcutaneous cyclocytidine (600 mg/m2) from days 1 to 7. Maintenance therapy consisted of IV etoposide (VP-16) (100 mg/m2) for 5 days and IV AMSA (100 mg/m2) on day 1. Of 122 patients, 109 were evaluable. There were 13 early deaths. Ninety-six patients received adequate therapy defined as completion of two courses of therapy. Of these 96 patients, 52 achieved complete remission. Fifteen of 33 patients who failed initial induction achieved complete remission. Eighteen of 39 patients who were resistant to anthracyclines had complete responses. There was no direct evidence of AMSA-induced cardiotoxicity. Remission duration was 28 days to 3 or more years (median, 98 days). AMSA and cyclocytidine were effective retrieval therapy for patients who were in relapse or unresponsive to frontline therapy. Duration of remission was short (median, 98 days).

Intensive therapy for children with acute lymphoblastic leukaemia and unfavourable presenting features. Early conclusions of study CCG-106 by the Childrens Cancer Study Group.

229 children with acute lymphoblastic leukaemia (ALL) and with clinical and laboratory features associated with a high risk of treatment failure entered a randomised study of three treatment regimens. Before 1981, such patients had a 3-year event-free survival (EFS) of 47%. Two intensive therapies, the Berlin-Frankfurt-Munster (BFM) 76/79 regimen and the New York (NY) regimen were compared with a control regimen that had achieved the best outcome in previous Trials. Data on 214 cases (93.4%) were analysed. The 3-year EFS was 78% for the BFM and NY regimens and 49% for the control regimen, a significant difference. The differences persisted after stratification by age at onset, sex, white blood cell count at diagnosis, and marrow blast morphology. Control patients were 2.7 times more likely to fail induction, to die, or to relapse than were patients on the intensive regimens.