Marrow-ablative chemotherapy followed by tandem autologous hematopoietic cell transplantation in pediatric patients with malignant brain tumors.

To improve survival in young children with malignant brain tumors, irradiation-avoiding or -minimizing marrow-ablative chemotherapy (HDCx) with autologous hematopoietic cell transplantation (AuHCT) has been investigated. We evaluated the outcome of 44 children with malignant brain tumors treated with HDCx and tandem AuHCT at Children's Hospital Los Angeles between June 1999 and July 2012. Forty-four children with malignant brain tumors were studied. Twenty-one had medulloblastoma/primitive neuro-ectodermal tumor, eight atypical teratoid/rhabdoid tumor (ATRT), five high-grade glioma, four malignant germ cell tumor, three ependymoma and three choroid plexus carcinoma. Twenty-nine patients received three tandem transplants and 15 received two tandem transplants, respectively. The 5-year PFS and overall survivals (OS) for all patients were 46.3±8.2% and 51.7±8.5%, respectively. The PFS and OS for 27 newly diagnosed patients were 68.9±9.9% and 73.5±9.3%, respectively, compared with 17 transplanted at relapse 11.8±9.8% (P<0.001) and 15.1±12.3% (P=0.0231), respectively. The 5-year PFS and OS in 13 previously unirradiated patients were 74±13% and 74±13% versus 33.2±9.8% and 40.2±10.6% in 31 irradiated patients (P=0.11 and P=0.239), respectively. One patient died of transplant-related toxicity. HDCx with tandem AuHCT is feasible and safe in children with malignant brain tumors with encouraging irradiation-free survival in newly diagnosed children.

Activity of irinotecan and temozolomide in the presence of O6-methylguanine-DNA methyltransferase inhibition in neuroblastoma pre-clinical models.

BACKGROUND: The combination of temozolomide (TMZ) and irinotecan is a regimen used in neuroblastoma patients with recurrent disease. O(6)-methylguanine-DNA methyltransferase (MGMT) may have a function in resistance to TMZ. Using neuroblastoma pre-clinical models, we determined whether the inhibition of MGMT by O(6)-benzylguanine (O6-BG) could enhance the anti-tumour activity of TMZ and irinotecan. METHODS: The cytotoxicity of TMZ and irinotecan, either alone or in combination, was measured in five neuroblastoma cell lines in the presence or absence of O6-BG with a fluorescence-based cell viability assay (DIMSCAN). Anti-tumour activity was measured in three neuroblastoma xenograft models. RESULTS: MGMT mRNA and protein were expressed in 9 out of 10 examined cell lines. Pretreatment of cells with 25 µM O6-BG decreased MGMT protein expression and enhanced The TMZ cytotoxicity by up to 0.3-1.4 logs in four out of five tested cell lines. TMZ (25 mg kg(-1) per day for 5 days every 3 weeks for four cycles) did not significantly improve mice survival, whereas the same schedule of irinotecan (7.5 mg kg(-1) per day) significantly improved survival (P<0.0001) in all three xenograft models. Combining O6-BG and/or TMZ with irinotecan further enhanced survival. CONCLUSION: Our in vitro and in vivo findings suggest that irinotecan drives the activity of irinotecan and TMZ in recurrent neuroblastoma. Inhibitors of MGMT warrant further investigation for enhancing the activity of regimens that include TMZ.

Central nervous system (CNS) tumors in the first six months of life: the Children's Hospital Los Angeles experience, 1979-2005.

BACKGROUND: The authors report the experience at the Children's Hospital Los Angeles with brain tumors diagnosed before 6 months of age, describing the characteristics of the patients, their tumors, treatment strategies, and prognostic factors. METHODS: Thirty-three children who were identified between 1979 and 2005 were included. Twelve were female (36%). There were 11 gliomas, 9 choroid plexus tumors, 8 medulloblastomas and supratentorial primitive neuroectodermal tumors (PNET), 2 atypical teratoid/rhabdoid tumors (ATRT), and 1 each of ependymoma, craniopharyngioma, and immature teratoma. Locations of primary tumors included 21 supratentorial (64%) and 7 posterior fossa, and 5 tumors involved both compartments. The treatment strategies included 5 patients with biopsy only, 18 less than gross total resections (

Specific cytogenetic abnormalities are associated with a significantly inferior outcome in children and adolescents with mature B-cell non-Hodgkin's lymphoma: results of the FAB/LMB 96 international study.

Clinical studies showed that advanced stage, high LDH, poor response to reduction therapy and combined bone marrow and central nervous system disease are significantly associated with a decreased event-free survival (EFS) in pediatric mature B-cell non-Hodgkin's lymphoma (B-NHL) treated on FAB/LMB96. Although rearranged MYC/8q24 (R8q24) is characteristic of Burkitt lymphoma (BL), little information is available on other cytogenetic abnormalities and their prognostic importance. We performed an international review of 238 abnormal karyotypes in childhood mature B-NHL treated on FAB/LMB96: 76% BL, 8% Burkitt-like lymphoma, 13% diffuse large B-cell lymphoma (DLBCL). The main BL R8q24-associated chromosomal aberrations were +1q (29%), +7q and del(13q) (14% each). The DLBCL appeared heterogeneous and more complex. Incidence of R8q24 (34%) was higher than reported in adult DLBCL. The prognostic value of cytogenetic abnormalities on EFS was studied by Cox model controlling for the known risk factors: R8q24, +7q and del(13q) were independently associated with a significant inferior EFS (hazard ratio: 6.1 (P=0.030), 2.5 (P=0.015) and 4.0 (P=0.0003), respectively). The adverse prognosis of R8q24 was observed only in DLBCL, whereas del(13q) and +7q had a similar effect in DLBCL and BL. These results emphasize the significant biological heterogeneity and the development of cytogenetic risk-adapted therapy in childhood mature B-NHL.

Feasibility of upfront dose-intensive chemotherapy in children with advanced-stage Hodgkin's lymphoma: preliminary results from the Children's Cancer Group Study CCG-59704.

BACKGROUND: Treatment strategies involving dose intensification have recently demonstrated improvements in cure compared with older trials. However, dose-intensive therapy is associated with increased acute and long-term toxicities, particularly in pediatric patients. The Children's Cancer Group initiated this pilot study to assess the feasibility and toxicity of a moderate dose-intensive regimen, BEACOPP (bleomycin, etoposide, doxorubicin, cyclophosphamide, vincristine, procarbazine and prednisone), in children and adolescents with advanced-stage Hodgkin's lymphoma (HL). PATIENTS AND METHODS: Children with stage IIB or IIIB with bulk disease, or stage IV were eligible. Induction consisted of four cycles of escalated dose BEACOPP. The rapidity of response, defined as >70% reduction in disease burden, was assessed after two and four cycles. Rapid responders then received consolidation therapy as per gender-specific guidelines to reduce the risk of gender-specific long-term toxicities of therapy, i.e. females received four cycles of COPP/ABV (cyclophosphamide, vincristine, procarbazine, prednisone, doxorubicin, bleomycin and vinblastine) without radiation therapy and males received two cycles of ABVD (doxurubicin, bleomycin, vinblastine and dacarbazine) with involved field radiation therapy (IFRT). Slow responders received four cycles of BEACOPP and IFRT. RESULTS: Ninety-nine patients were enrolled. Myelosuppression was frequent. Non-hematological grade 4 toxicities included allergic reaction (two patients), hypotension (one), mucositis (four), infection (three), seizure (one) and elevated transaminases (one). Typhlitis developed in four patients; three recovered and completed dose-modified chemotherapy, while one died of sepsis associated with grade 4 neutropenia. A rapid response was achieved by 45 and 72% of patients after two and four cycles, respectively. There are no disease progressions or secondary malignancies to date. There is only one reported relapse to date. Median follow-up for the cohort is 6 months. CONCLUSIONS: BEACOPP chemotherapy is feasible and generally well tolerated in children with advanced-stage HL. The absence of reported progressive disease and only one relapse to date is encouraging.

Comparison of long-term outcome of children and adolescents with disseminated non-lymphoblastic non-Hodgkin lymphoma treated with COMP or daunomycin-COMP: A report from the Children's Cancer Group.

BACKGROUND: Early Children's Cancer Group (CCG) trials indicated that the cyclophosphamide, vincristine, methotrexate, and prednisone (COMP) regimen was superior to the LSA2L2 regimen for non-lymphoblastic (NLB) non-Hodgkin lymphoma (NHL). Studies by other groups suggested that addition of anthracyclines to standard therapies could improve outcome. Therefore, in 1983 CCG initiated study CCG-503, a randomized trial of COMP vs. daunomycin-COMP (D-COMP) in children and adolescents with disseminated NLB NHL. PROCEDURES: Between December 1983 and April 1990, 404 eligible patients were entered. Patients without central nervous system (CNS) or marrow involvement were randomized to receive COMP (N = 139) or D-COMP (N = 145). Randomization was stratified by histology and site of disease. Patients with CNS or marrow involvement (stage IV) were non-randomly treated with D-COMP (N = 120). RESULTS: Ten-year event-free survival in COMP and D-COMP patients was similar: 55 +/- 4.3% (Estimate +/- SE) vs. 57 +/- 4.2% (not significant). Stage I-III patients with large-cell (LC) NHL had worse 10-year event-free survival (EFS) (48 +/- 4.9%) than those with small non-cleaved cell (SNCC) NHL disease (61 +/- 3.5%, P < 0.05 in multivariate analysis), but equivalent survival (65 +/- 4.7% vs. 63 +/- 3.5%) due to significantly higher salvage rates in LC patients, especially those failing more than 12 months from diagnosis. Ten-year EFS in stage IV patients was 39 +/- 5.2%. Addition of daunomycin resulted in higher rates of grade 3/4 hematologic toxicity and stomatitis, as well as late cardiac-related deaths. The incidence of second malignant neoplasms was 1.0% at 10 years. CONCLUSIONS: Addition of daunomycin to standard COMP therapy did not improve outcome in pediatric disseminated NLB NHL. Patients with LC disease had a significantly reduced long-term EFS, but were retrieved at a higher rate than patients with SNCC disease, resulting in equivalent long-term survival.

Palpable breast cancers are inherently different from nonpalpable breast cancers.

BACKGROUND: We examined the clinicopathologic profile of T1 cancers to determine whether palpable cancers are different from nonpalpable cancers. METHODS: A prospective database was reviewed. Palpable T1 cancers were compared with nonpalpable T1 cancers. Initial significance was determined by chi2 analysis. Factors found to be significant were then reanalyzed. controlling for tumor size by logistic or linear regression, as appropriate. RESULTS: Of 1263 T1 cancers treated between 1981 and 2000, 857 (68%) were palpable and 401 (32%) were nonpalpable. Palpability correlated with pathologic tumor size, mitotic grade, nuclear grade, high S-phase, lymphovascular invasion, nodal positivity, and lack of extensive intraductal component, multifocality, and multicentricity. There was no significant difference in estrogen receptor, progesterone receptor or Her-2/neu status, ploidy, or DNA index. Breast cancer-specific survival was worse for patients with palpable cancers. CONCLUSIONS: Palpable cancers are inherently different from nonpalpable cancers, with a less diffuse growth pattern, higher metastatic potential, higher proliferative activity, more nuclear abnormalities, and a worse prognosis.

Age and TP53 mutation frequency in childhood malignant gliomas: results in a multi-institutional cohort.

Malignant astrocytoma is one of the most deadly primary central nervous system tumors. Although significant progress has been made in understanding the molecular pathways that lead to the development of these tumors in adults, comparatively little analysis has been done in childhood astrocytomas, which are less common and have a more favorable prognosis. Our previous studies of an institutional cohort of children with malignant gliomas suggested the existence of distinct molecular pathways of tumorigenesis in younger versus older children, based on the finding of a high frequency of TP53 mutations in tumors from children >3 years of age at diagnosis, compared with those from younger children. In the current study, the association between TP53 mutations and age was examined in greater detail using the multi-institutional group of children enrolled in Children's Cancer Group Study 945, the largest cohort of childhood high-grade gliomas analyzed to date. Seventy-seven tumors with centrally reviewed diagnoses of anaplastic astrocytoma or glioblastoma multiforme had sufficient archival histopathological material for microdissection-based genotyping. Sections were examined histologically, and topographic targets that contained malignant tissue were isolated by microdissection and subjected to PCR-based amplification and sequencing of TP53 exons 5-8. Twenty-six tumors (33.8%) had mutations in those exons. Mutations were observed in 2 of 17 tumors (11.8%) from children <3 years of age at diagnosis versus 24 of 60 tumors (40%) from older children, a difference that was statistically significant (P = 0.04), in agreement with our previous results. Whereas malignant gliomas in older children have a frequency of mutations comparable to tumors that arise in young adults, those from children <3 years old do not. The association between age and frequency of TP53 mutations among pediatric malignant gliomas indicates the probable existence of two distinct pathways of molecular tumorigenesis in younger versus older children.

Outcomes of treatment of children and adolescents with recurrent non-Hodgkin's lymphoma and Hodgkin's disease with dexamethasone, etoposide, cisplatin, cytarabine, and l-asparaginase, maintenance chemotherapy, and transplantation: Children's Cancer Group Study CCG-5912.

PURPOSE: To determine the toxicity and response rate in children treated with dexamethasone, etoposide, cisplatin, high-dose cytarabine, and L-asparaginase (DECAL) for recurrent non-Hodgkin's lymphoma (NHL) and Hodgkin's disease (HD). PATIENTS AND METHODS: Ninety-seven children with recurrent NHL (n = 68) or HD (n = 29) were enrolled. Treatment consisted of two cycles of DECAL, then bone marrow transplantation or up to four cycles of ifosfamide, mesna, and etoposide alternating with DECAL maintenance therapy. RESULTS: After two cycles of DECAL induction therapy, complete response (CR) or partial response (PR) was reported in 19 (65.5%; 10 CRs and nine PRs) of 29 patients with HD and 29 (41.6%; 23 CRs and six PRs) of 68 patients with NHL. When only 24 patients with HD and 58 patients with NHL who were assessable for response were considered, the response rates were 79.2% (19 of 24 patients) and 50.0% (29 of 58 patients), respectively. Five-year event-free survival was 26% +/- 9% and 23% +/- 5% in patients with HD and NHL, respectively. Five-year survival was 31% +/- 14% and 30% +/- 6%, respectively. Although median time to treatment failure was significantly longer in patients with HD (EFS, P =.002; survival, P =.011), this difference did not translate into a higher long-term survival. Grade 3 or 4 toxic effects were observed during induction in 70 (72%) of 97 patients and during maintenance in 45 (70%) of 64 courses of DECAL therapy. Pancytopenia and systemic infections in particular were frequently observed. Other toxic effects were uncommon. Although not a formal part of the therapy or the study design, 42 patients who responded to therapy who underwent bone marrow transplant did not show any benefit from this approach. CONCLUSION: DECAL is an effective and tolerable salvage regimen for treating patients with recurrent NHL and HD.

Breast cancer after augmentation mammoplasty.

BACKGROUND: It is thought that implants interfere with breast cancer diagnosis and that cancers in women who have had breast augmentation carry a worse prognosis. METHODS: A prospective breast cancer database was reviewed, comparing augmented and nonaugmented patients for details of histology, palpability, tumor size, nodal status, mammographic status, receptor status, nuclear grade, stage, and outcome. RESULTS: Ninety-nine cancers in augmented women and 2857 cancers in nonaugmented women were identified. Among these women, mammography was normal in 43% of those who had had augmentation and in 5% of those who had not. Augmented women were more likely to have palpable cancers (83% vs. 59%) and nodal involvement (48% vs. 36%), and less likely to have ductal carcinoma in situ (DCIS) (18% vs. 28%). When comparing only women younger than 50, the differences in invasiveness and nodal status lost significance. Cancers diagnosed in the 1990s were more likely to be nonpalpable and noninvasive than those diagnosed in the 1980s. This trend was more pronounced in the augmented population. CONCLUSIONS: Augmented patients were more likely to have palpable cancers, although the overall stage and outcome were similar to those of nonaugmented women. Although there have been significant improvements in our ability to diagnose early breast cancer over the past two decades, mammography continues to be suboptimal in augmented women.

A novel snake venom disintegrin that inhibits human ovarian cancer dissemination and angiogenesis in an orthotopic nude mouse model.

OVCAR-5 is a human epithelial carcinoma cell line of the ovary, established from the ascitic fluid of a patient with progressive ovarian adenocarcinoma without prior cytotoxic treatment. The unique growth pattern of ovarian carcinoma makes it an ideal model for examining the anticancer activity of contortrostatin (CN), a homodimeric disintegrin from southern copperhead venom. FACS analysis revealed that OVCAR-5 is integrin alphavbeta3 negative, but alphavbeta5 positive. CN effectively blocks the adhesion of OVCAR-5 cells to several extracellular matrix proteins and inhibits tumor cell invasion through an artificial basement membrane. In a xenograft nude mouse model with intraperitoneal introduction of OVCAR-5 cells, intraperitoneal injection of CN was used for therapy. Tumor dissemination in CN-treated versus control groups was studied by gross examination, and antiangiogenic potential was examined by factor VIII immunohistochemistry and image analysis. CN not only significantly inhibited ovarian cancer dissemination in the nude mouse model, but it also dramatically prevented the recruitment of blood vessels to tumors at secondary sites.

The delta valve: how does its clinical performance compare with two other pressure differential valves without antisiphon control?

OBJECTIVE: The Delta valve is a pressure differential valve with a siphon control device. The valve mechanism is normally closed, but is designed to open in response to positive ventricular pressure, thereby avoiding overdrainage of cerebrospinal fluid (CSF). As a result, the incidence of subdural fluid collections as well as postural symptoms is purportedly reduced. In addition, the valve might reduce the number of obstructions as there would be no negative pressure sucking tissue and debris into the shunt system. In order to assess whether use of the Delta valve reduced the number of shunt-related problems as compared with two other pressure differential valves without an antisiphon component, we performed a retrospective review of all children undergoing CSF diversion procedures at our institution. METHODS: We reviewed the charts of 1, 193 patients. Cases included 2,325 ventriculoperitoneal (V-P) shunt insertions or revisions from January 1, 1985, to December 31, 1994, performed at our institution. The Delta valve and two pressure differential valves without antisiphon function were exclusively inserted during the following time periods: Holter-Hausner (H-H): January 1, 1985, to August, 1987; Heyer-Schulte (H-S): August, 1987, to June, 1991, and Delta: June, 1991, to December 31, 1994. RESULTS: Of the cases reviewed, 475 patients underwent insertion of a V-P shunt at the Childrens Hospital of Los Angeles and had a total of 686 shunt operations. Median follow-up was 3 years and ranged up to 10 years. Kaplan-Meier analysis documented that 67% of H-H, 71% of H-S and 70% of the Delta valves were functioning at 1-year follow-up. At 2-year follow-up, 66% of H-H, 64% of H-S and 65% of the Delta valves were functioning. The difference was not statistically significant. The occurrence rate for symptomatic subdural fluid collections was 0.7% (1/130) for H-H, 2.2% (3/139) for H-S and 1.0% (2/206) for the Delta valve (p = 0.52). The combined breakage/obstruction rate for the series was 7.7% (10/130) for H-H, 2.9% (4/139) for H-S and 4.9% (10/206) for the Delta valve (p = 0.19). No Delta valves malfunctioned secondary to fibrous capsule affecting the antisiphon device. CONCLUSIONS: In conclusion, it appears that performance of the Delta valve was not significantly different from the H-H and H-S valves, two valves without an antisiphon device. There was no significant difference in the occurrence of symptomatic subdural fluid collections based upon valve type, or in the combined valve breakage/obstruction rates based upon valve type.

Large-cell lymphoma arising in the mediastinum in children and adolescents is associated with an excellent outcome: a Children's Cancer Group report.

PURPOSE: Large-cell lymphoma (LCL) arising in the mediastinum (LCL-M) is a heterogeneous group of non-Hodgkin's lymphoma (NHL) that includes B-cell lymphomas as well as T-cell lymphomas, including anaplastic LCL. LCL-M is well recognized in young adults but is less well characterized and infrequent in children and adolescents. METHODS: A retrospective review of Children's Cancer Group therapeutic studies for nonlymphoblastic lymphomas (CCG-551, CCG-503, CCG-552, and CCG-5911) identified 20 patients with LCL-M, representing 7.2% of all LCLs classified by central pathology review. RESULTS: The patients ranged in age from 4 to 19 years (median, 12.5 years; mean, 12 years); 55% of the patients were male. Although a variety of chemotherapy regimens were used, response was excellent, with all 20 patients (100%) achieving a complete response. Four patients (20%) experienced relapse locally or in distant sites including brain and kidney. One patient died of sepsis during therapy. For the 20 patients with LCL-M, the product-limit estimated 5-year event-free survival (EFS) and 5-year overall survival (OS) rates are 75% +/- 10% and 85% +/- 8%, respectively. For disseminated LCLs (192 cases), the EFS and OS rates were 50% +/- 4% and 63% +/- 4%, respectively, which differ significantly from the those of the LCL-M cases (EFS, P =.025; OS, P =.034). The 5-year EFS and OS rates for patients with localized LCL (67 cases) were 92 +/- 3% and 97 +/- 2%, respectively. CONCLUSION: LCL-M is a heterogeneous group of NHLs that makes up approximately 7.2% of LCL in children and adolescents. Response to therapy and OS in this young age group seems excellent and superior to that of disseminated LCLs but inferior to that of other localized LCL. Future studies of LCL-M will evaluate short intense chemotherapy administered without radiation therapy.

Correlation of neurosurgical subspecialization with outcomes in children with malignant brain tumors.

OBJECTIVE: This study was performed to evaluate the association between the type of neurosurgeon (general or pediatric) and either the extent of tumor removal or the frequency of complications in children undergoing malignant brain tumor resections. METHODS: Data were analyzed from three recent Children's Cancer Group studies: two on medulloblastomas/primitive neuroectodermal tumors and one on malignant gliomas. Neurosurgeons were classified as general neurosurgeons, as designated pediatric neurosurgeons in their institutions, or as members of the American Society of Pediatric Neurosurgeons (ASPN), which requires pediatric neurosurgical experience and practice standards. RESULTS: Data forms from 732 children were analyzed; 485 were from children with medulloblastomas/primitive neuroectodermal tumors, and 247 were from children with malignant gliomas. Operations were performed by 269 neurosurgeons, including 213 general neurosurgeons, 29 designated pediatric neurosurgeons, and 27 ASPN members. The mean number of operations per surgeon was 1.8, 4.9, and 7.6 for general neurosurgeons, designated pediatric neurosurgeons, and ASPN members, respectively. There was a significant relationship between the extent of tumor resection or the amount of residual tumor and the type of neurosurgeon. Designated pediatric neurosurgeons and ASPN members were more likely to remove more than 90% of the tumor and to leave less than 1.5 cc of residual tumor than were general neurosurgeons (P<0.05). In these studies, the probability of extensive tumor removal correlated with the number of operations the neurosurgeon performed (P<0.01). Neurological complications occurred in the following proportion of cases: general neurosurgeons, 23%; designated pediatric neurosurgeons, 32%; and ASPN members, 18%. CONCLUSION: Pediatric neurosurgeons are more likely than general neurosurgeons to extensively remove malignant pediatric brain tumors. In these tumors, extent of removal has been demonstrated to influence survival.

Brachytherapy for primary and recurrent nasopharyngeal carcinoma: 20 years' experience at Long Beach Memorial.

PURPOSE: We evaluated treatment outcomes of patients with mostly locally advanced primary and recurrent cancer of the nasopharynx managed with interstitial and intraluminal brachytherapy. METHODS AND MATERIALS: This is a retrospective analysis of 56 patients with cancer arising from the nasopharynx treated with interstitial and intracavitary afterloading brachytherapy from 1978 to 1997. Patients were divided into three treatment groups: 15 patients with primary cancer (Group 1), 34 patients with recurrent or persistent disease (Group 2), and 7 patients with cancer in the nasopharynx who had history of previous definitive radiation therapy to the nasopharynx for head and neck cancer (Group 3). Fifty-three percent of patients in Group 1 had 1992 AJCC Stage IV disease, and 49% of patients in Groups 2 and 3 had extensive disease (defined as T3, T4, or parapharyngeal extension). Group 1 received megavoltage radiation to 50-60 Gy followed by a boost to the primary site and neck (in cases of persistent neck disease) with a combination of interstitial and intracavitary brachytherapy (mean dose 33-37 Gy). Five patients received chemotherapy, and 6 patients received hyperthermia. Groups 2 and 3 patients were treated with brachytherapy implants (mean dose 50-58 Gy) without external beam radiation. Twenty-five patients received chemotherapy either before or during radiation, and 21 patients received hyperthermia. RESULTS: The overall survival at 2, 5, and 10 years for patients in Group 1 was 79%, 61%, and 61%, respectively, and for patients in Groups 2 and 3 combined was 48%, 30%, and 20%, respectively. Cause-specific survival at 2, 5, and 10 years was 87%, 74%, and 74%, respectively, for patients in Group 1; and 82%, 60%, and 60%, respectively, for patients in Groups 2 and 3. Local control at 2, 5, and 10 years was 93%, 93%, and 77%, respectively, for patients in Group 1; and 81%, 59%, and 49%, respectively, for patients in Groups 2 and 3. Control in the neck at 2, 5, and 10 years was achieved in 93%, 93%, and 93% of patients, respectively, in Group 1; and 88%, 81%, and 81%, respectively, for patients in Groups 2 and 3. Disease-free survival was 87%, 74%, and 62%, respectively, for patients in Group 1, and 56%, 41%, and 34%, respectively, for patients in Groups 2 and 3. There were 4 peri-operative deaths. One death (2%) was attributable to the development of late complications. Forty-five percent of patients experienced some form of late complications. CONCLUSION: Interstitial afterloading brachytherapy can provide effective treatment for nasopharyngeal cancers, especially for locally persistent/recurrent and locally extensive lesions.

Outcome of CNS disease at diagnosis in disseminated small noncleaved-cell lymphoma and B-cell leukemia: a Children's Cancer Group study.

PURPOSE: To examine the impact of initial CNS involvement on outcome and patterns of failure in patients with disseminated small noncleaved-cell lymphoma and B-cell leukemia who were treated in four successive Children's Cancer Group trials. PATIENTS AND METHODS: Of 462 patients with disseminated disease, 49 (10.6%) had CNS disease at diagnosis (CNS+). CNS disease included meningeal disease or CNS parenchymal masses with or without cranial neuropathies (CSF+/Mass; CNPs) in 36 patients and isolated CNPs in 13. Of the CNS+ patients, 28 had M2 (5% to 25% blasts) or M3 (> 25% blasts) bone marrow involvement. All patients received protocol-based systemic and intrathecal chemotherapy. Thirty-six patients also received CNS irradiation. RESULTS: Relapses occurred in 21 (43%) of 49 patients, predominantly in the CNS (71%) and bone marrow (52%). The 3-year event-free survival +/- SE for all patients with CNS+ disease was 45% +/- 7%. Patients with CSF+/Mass had a nominally higher treatment failure rate compared with patients with CNS- after adjusting for marrow status and lactate dehydrogenase (LDH) diagnosis, with a relative failure rate (RFR) of 1.52 (95% confidence interval [CI], 0.88 to 2.6; P =.15). In comparison, the RFRs for patients with M2 or M3 marrow and for those with LDH levels greater than 500 IU/L after adjusting for CNS disease were 1.4 (95% CI, 0.96 to 2.0; P =.029) and 2.2 (95% CI, 1.5 to 3.0; P <.001), respectively. The RFR for patients with isolated CNPs was 0.87 (95% CI, 0.36 to 2.1; P =.76). CONCLUSION: We conclude that, with the treatments used during the period covered by these studies, the presence of CSF+/Mass CNS disease at diagnosis was associated with a nominally worse outcome independent of initial bone marrow status and LDH level, but the effect was not statistically significant.

A strategic view of randomized trial design in low-incidence paediatric cancer.

We use a statistical model to examine the relationship between alpha level, sample size, trial duration, patient accrual rate and therapeutic innovation rate on the increase in treatment efficacy achieved after a series of two-treatment randomized phase III trials. In a setting where the trials include most of the patients in the target population for inference, as in some paediatric cancers, we show that the traditional criteria by which one determines trial size are difficult to justify and apply. In particular, using as a measure of evidence type I error levels larger than the typical 5 per cent for judging treatment differences, and performing smaller trials than one would usually consider feasible, yields on average, over a 25-year research course, larger gains in cure rate. Judicious choice of type I error rate and trial size keeps the chance of worsening treatment efficacy at a low level, even while increasing the chance of making large improvements in cure rate. We propose that a more appropriate view of trial design in low-incidence cancer settings is in the overall context of the research setting and long-term goals rather than in the narrow context of the current single trial. From this viewpoint, insistence on large trials and stringent evidence for accepting new treatments can be counter-productive, in that likely gains in efficacy of treatment will be smaller over the long term.

Treatment outcome and prognostic factors for infants with acute lymphoblastic leukemia treated on two consecutive trials of the Children's Cancer Group.

PURPOSE: Infants represent a very poor risk group for acute lymphoblastic leukemia (ALL). We report treatment outcome for such patients treated with intensive therapy on consecutive Children's Cancer Group (CCG) protocols. PATIENTS AND METHODS: Between 1984 and 1993, infants with newly diagnosed ALL were enrolled onto CCG-107 (n = 99) and CCG-1883 (n = 135) protocols. Postconsolidation therapy was more intensive on CCG-1883. On both studies, prophylactic treatment of the CNS included both high-dose systemic chemotherapy and intrathecal therapy, in contrast to whole-brain radiotherapy, which was used in earlier studies. RESULTS: Most patients (>95%) achieved remission with induction therapy. The most frequent event was a marrow relapse (46 patients on CCG-107 and 66 patients on CCG- 1883). Four-year event-free survival was 33% (SE = 4.7%) on CCG-107 and 39% (SE = 4.2%) on CCG- 1883. Both studies represent an improvement compared with a 22% (SE = 5.1%) event-free survival for historical controls. Four-year cumulative probabilities of any marrow relapse or an isolated CNS relapse were, respectively, 49% (SE = 5%) and 9% (SE = 3%) on CCG-107 and 50% (SE = 5%) and 3% (SE = 2%) on CCG-1883, compared with 63% (SE = 6%) and 5% (SE = 3%) for the historical controls. Independent adverse prognostic factors were age less than 3 months, WBC count of more than 50,000/microL, CD10 negativity, slow response to induction therapy, and presence of the translocation t(4;11). CONCLUSION: Outcome for infants on CCG-107 and CCG- 1883 improved, compared with historical controls. Marrow relapse remains the primary mode of failure. Isolated CNS relapse rates are low, indicating that intrathecal chemotherapy combined with very-high-dose systemic therapy provides adequate protection of the CNS. The overall unsatisfactory outcome observed for the infant ALL population warrants the future use of novel alternative therapies.

A partially grouped logrank test.

We propose a modified version of the standard logrank test for survival data in which the first contribution to the statistic is based on grouping of data before a pre-selected grouping time. The grouping is accomplished by artificially constructing the first table based on the product limit estimates of the proportions surviving at the grouping time. Remaining contributions to the statistic are identical to that of the standard logrank statistic. The approach has the advantage of being uninfluenced by non-proportional hazards differences prior to the grouping time while being almost as efficient as the usual logrank test for proportional-hazards alternatives. The statistic is particularly useful for interim monitoring in situations where early difference between treatments are unimportant and/or lagged treatment effects are anticipated. We report simulation studies to verify and investigate the test's properties.

Interpreting HIV serodiagnostic test results in the 1990s: social risks of HIV vaccine studies in uninfected volunteers. NIAID AIDS Vaccine Clinical Trials Group.

OBJECTIVE: To evaluate the influence of a human immunodeficiency virus (HIV) vaccine given to uninfected volunteers on the interpretation of serodiagnostic HIV test results. DESIGN: Retrospective cohort study. SETTING: 5 AIDS Vaccine Evaluation Units funded by the National Institute of Allergy and Infectious Diseases. PARTICIPANTS: The first 266 healthy adult volunteers (aged 18 to 60 years) who did not have HIV infection and whose history suggested that they were at low risk for acquiring HIV infection. MEASUREMENTS: HIV antibody was measured by enzyme-linked immunosorbent assay (ELISA) and Western blot test, the results of which were interpreted on the basis of four different published criteria. RESULTS: At some time during the first 12 months of the vaccine studies, 68% of volunteers were positive for HIV antibodies by ELISA. Depending on criteria used to interpret Western blot test results, 0% to 44% of volunteers had positive results that might have caused them to be incorrectly labeled as HIV infected. CONCLUSIONS: Significant social risks to volunteers participating in HIV vaccine studies were identified. Persons interpreting HIV serodiagnostic test results must consider that an HIV vaccine can cause a positive result in persons who are not infected.