Giulio J. D'Angio

Bone sarcomas linked to radiotherapy and chemotherapy in children

We estimated the risk of subsequent bone cancer among 9170 patients who had survived two or more years after the diagnosis of a cancer in childhood. As compared with the general population, the patients had a relative risk of 133 (95 percent confidence interval, 98 to 176) and a mean (+/- SE) 20-year cumulative risk of 2.8 +/- 0.7 percent. Detailed data on treatment were obtained on 64 patients in whom bone cancer developed after childhood cancer. As compared with 209 matched controls who had survived cancer in childhood but who did not have bone cancer later, patients who had had radiation therapy had a 2.7-fold risk (95 percent confidence interval, 1.0 to 7.7) and a sharp dose-response gradient reaching a 40-fold risk after doses to the bone of more than 6000 rad. The relative dose-response effect among patients who had been treated for retinoblastoma resembled that among patients with all other types of initial tumors, although the cumulative risk of bone cancer in the retinoblastoma group was higher. Similar numbers of patients were treated with orthovoltage and megavoltage; the patterns of risk among categories of doses did not differ according to the type of voltage. After adjustment for radiation therapy, treatment with alkylating agents was also linked to bone cancer (relative risk, 4.7; 95 percent confidence interval, 1.0 to 22.3), with the risk increasing as cumulative drug exposure rose. We conclude that both radiotherapy and chemotherapy with alkylating agents for childhood cancer increase the subsequent risk of bone cancer.

Prognostic Factors in Neuroblastoma

Known prognostic factors in neuroblastoma were analyzed in 124 children to determine which were independent and which were most useful in predicting outcome. The following factors were analyzed: age, sex, stage of disease, serum neuron‐specific enolase (NSE), serum ferritin, E‐rosette inhibition, urinary catecholamines, and histologic type according to the criteria of Shimada. Estimates of survival were calculated using the method of Kaplan and Meier. The overall survival for 124 patients was 60% at 2 years. There were significant differences in survival by pathology, age, NSE, ferritin, vanilmandelic acid (VMA): homovanillic acid (HVA) ratio, and stage. There was a strong association among NSE, age, stage, and ferritin. Using the recursive partitioning approach, it was possible to subdivide patients into three groups (based on diagnostic values of ferritin, age, and stage) with a good, intermediate, and poor prognosis and estimated 2‐year survival of 100%, 62%, and 19%, respectively. Further analysis could not be done because of small numbers in the subgroups, but the results suggest that combinations of age, stage, serum ferritin, and histologic type may be able to define two populations: favorable with 80% + 2‐year survival and unfavorable with less than 20%.

Clear cell sarcoma of the kidney : A review of 351 cases from the National Wilms tumor Study Group pathology center

We reviewed 351 cases of clear cell sarcoma of the kidney (CCSK), including 182 cases entered on National Wilms Tumor Study Group (NWTSG) trials 1-4 for which clinical follow-up information was available. Tumors were restaged using NWTS 5 criteria. Mean age at diagnosis in the NWTS group was 36 months with a range of 2 months to 14 years. The male to female ratio was 2:1. Typical gross features included large size (mean diameter 11.3 cm), a mucoid texture, foci of necrosis, and prominent cyst formation. Nine major histologic patterns were identified (classic, myxoid, sclerosing, cellular, epithelioid, palisading, spindle, storiform, and anaplastic); virtually all tumors contained multiple patterns that blended with one another. Immunohistochemical stains were performed on 45 cases; only vimentin was consistently immunoreactive. Consistently negative results with other antibodies helped exclude other tumors in the differential diagnosis; all CCSKs were cytokeratin-negative, including epithelioid tumors that mimicked Wilms tumor, and MIC2-negative, including cellular tumors that mimicked primitive neuroectodermal tumor. The p53 gene product was rarely overexpressed in non-anaplastic CCSKs, but strikingly overexpressed in two of three anaplastic CCSKs. Overall survival was 69%. Multivariate analysis revealed four independent prognostic factors for survival: treatment with doxorubicin, stage, age at diagnosis, and tumor necrosis. Of note, stage 1 patients had a remarkable 98% survival rate. No other histologic or clinical variable independently correlated with survival.

Surgery-related factors and local recurrence of Wilms tumor in National Wilms Tumor Study 4.

OBJECTIVE To assess the prognostic factors for local recurrence in Wilms tumor. SUMMARY BACKGROUND DATA Current therapy for Wilms tumor has evolved through four studies of the National Wilms Tumor Study Group. As adverse prognostic factors were identified, treatment of children with Wilms tumor has been tailored based on these factors. Two-year relapse-free survival of children in the fourth study (NWTS-4) exceeded 91%. Factors once of prognostic import for local recurrence may lose their significance as more effective therapeutic regimens are devised. METHODS Children evaluated were drawn from the records of NWTS-4. A total of 2482 randomized or followed patients were identified. Local recurrence, defined as recurrence in the original tumor bed, retroperitoneum, or within the abdominal cavity or pelvis, occurred in 100 children. Using a nested case-control study design, 182 matched controls were selected. Factors were analyzed for their association with local failure. Relative risks and 95% confidence intervals were calculated, taking into account the matching. RESULTS The largest relative risks for local recurrence were observed in patients with stage III disease, those with unfavorable histology (especially diffuse anaplasia), and those reported to have tumor spillage during surgery. Multiple regression analysis adjusting for the combined effects of histology, lymph node involvement, and age revealed that tumor spillage remained significant. The relative risk of local recurrence from spill was largest in children with stage II disease. The absence of lymph node biopsy was also associated with an increased relative risk of recurrence, which was largest in children with stage I disease. The survival of children after local recurrence is poor, with an average survival rate at 2 years after relapse of 43%. Survival was dependent on initial stage: those who received more therapy before relapse had a worse prognosis. CONCLUSIONS This study has demonstrated that surgical rupture of the tumor must be prevented by the surgeon, because spills produce an increased risk of local relapse. Both local and diffuse spills produce this risk. Stage II children with local spill appear to require more aggressive therapy than that used in NWTS-4. The continued critical importance of lymph node sampling in conjunction with nephrectomy for Wilms tumor is also established. Absence of lymph node biopsy may result in understaging and inadequate treatment of the child and may produce an increased risk of local recurrence.

31P nuclear magnetic resonance spectroscopic investigation of human neuroblastoma in situ.

Neuroblastoma is a unique tumor of childhood that has a wide range of malignant expression. The prognoses range from excellent, with minimal treatment required, for patients with localized tumors o...

Intracranial ependymomas in children

Between 1970 and 1988, 51 children with intracranial ependymal tumors (33-infratentorial, 18-supratentorial received initial treatment at the University of Pennsylvania. Therapy consisted of total or near total tumor resection in 15 patients and partial resection or biopsy in 36. Postoperative irradiation alone was given to 18, chemotherapy to 4, and a combination of these two modalities to 26. Patients have been followed for a median period of 7.75 years. The 5-year actuarial survival and progression-free survival (PFS) rates are 46% and 30%, respectively. Of the 30 patients who have progressed, 29 did so locally and one died before the site of failure could be determined. Six patients also had disease outside the primary site at relapse; three of them had received craniospinal irradiation. Local control was significantly better for patients whose tumor dose exceeded 4500 cGy (32% vs. 0%, p = .01) and for Caucasian patients (34% vs. 15%, p =.05). Survival was better for patients who were over 4 years of age at diagnosis (55% vs. 30%, p = .04), for patients who received local radiation doses above 4500 cGy (51% vs. 18%, p = .01), and for Caucasian patients (43% vs. 14%, p = .01). Extent of resection, histology, location, the use of cranial or craniospinal irradiation, and the use of chemotherapy did not significantly impact on survival. We conclude that the inability to control local disease remains the single most important factor leading to treatment failure. Older age, higher local radiation dose, and Caucasian race appear to be the only favorable prognostic factors.

A review of 17 IV-S neuroblastoma patients at the children's hospital of Philadelphia

The records of 207 neuroblastoma patients seen at the Childrens Hospital of Philadelphia between 1944 and 1977 were reviewed to study some of the features associated with the unusually good prognosis found in patients with Stage IV‐S neuroblastoma. Initially, 22 patients appeared to fit the criteria of small primary tumor and distant disease in liver, skin, and/or marrow without evidence of bone metastases; 5 patients were subsequently rejected as being incorrectly staged. The remaining 17 patients had abdominal primary tumors and hepatic disease; in 12 of the 17, an enlarged liver was the presenting sign. Six patients had skin lesions, 4 had disease in the marrow on routine smear, and additional sites of spread were pancreas and bowel serosa. The treatment given was not systematic, and it was not possible to correlate any specific form of therapy with a satisfactory outcome. Eleven of 17 patients survived; 6 of 11 survivors had spontaneous regression of all or part of their disease, 5 of 6 who died received irradiation, chemotherapy, or both. Death usually occurred in the first month as a complication of the local disease; 1 patient succumbed to radiation nephritis. This study establishes that the special pattern of widespread neuroblastoma termed Stage IV‐S does exist, and that it is associated with a good prognosis. Careful consideration should be given before selecting treatment for the Stage IV‐S child because spontaneous regression is likely to occur in most of them. In patients with rapidly enlarging livers, renal or pulmonary complications may develop because of liver bulk or coagulopathies. Treatment should be directed to the liver in these cases because distant metastases seldom supervene. Low‐dose irradiation, mild chemotherapy, and possibly surgical release of intraabdominal pressure using a silastic patch have all been effective. Unfortunately, patients occasionally succumb to local disease in spite of these and more aggressive measures.

Updated results of a pilot study of low dose craniospinal irradiation plus chemotherapy for children under five with cerebellar primitive neuroectodermal tumors (medulloblastoma)

PURPOSE Children under 5 years old with medulloblastoma (MB) have a poor prognosis. They are more susceptible to the deleterious effects of craniospinal irradiation (CSART) and have a higher relapse rate when treated with low-dose CSART alone. We, thus, embarked on a prospective trial testing the usefulness of very low dose CSART and adjuvant chemotherapy. This is an update of a previous report on these patients. METHODS AND MATERIALS Between January 1988 and March 1990, 10 patients with medulloblastoma were treated using 18 Gy radiation therapy (RT) to the craniospinal axis, a posterior fossa (PF) boost to 50.4-55.8 Gy and chemotherapy consisting of vincristine (VCR) weekly during RT. This was followed by VCR, cis-diamminedichloroplatinum (CDDP), and lomustine (CCNU) for eight, 6-week cycles. Patients between 18 and 60 months of age without evidence of tumor dissemination were eligible for study. Follow-up was available until September 1994 with a median follow-up for living patients of 6.3 years from diagnosis. RESULTS Actuarial survival at over 6 years is 70 +/- 20%. Three of the 10 patients relapsed and died. In one patient, the relapse developed in the spine and brain outside the posterior fossa, in the second, concurrently in the posterior fossa, brain and spine, and the third, only in the spine. One surviving child developed a brain stem infarct 4.8 years after diagnosis and has since almost fully recovered. A mean intelligence quotient (IQ) score of 103 in six patients surviving at least 1 year is unchanged from the baseline group score of 107. Five children tested at baseline and 2 years following treatment had IQ scores of 101 and 102, respectively. Six children tested at baseline and at 3 years had IQ scores of 106 and 96, respectively. Excluding the child tested shortly after his brain stem infarct, baseline and 3 year IQ scores were 103 and 97, respectively. Five of the seven long-term survivors grew at rates significantly below their expected velocities during the follow-up period, while the others grew normally. Three patients have received growth hormone, and none have required thyroid replacement. CONCLUSIONS These data suggest that medulloblastoma patients can be cured with chemotherapy and reduced doses of craniospinal irradiation. The low doses of CSART given by us in conjunction with cis-platin-based chemotherapy produce minimal neurocognitive damage. Growth velocities in very young children so treated are, however, dramatically reduced. Better means of improving the therapeutic ratio are still needed.

THE ROLE OF RADIATION THERAPY IN THE MANAGEMENT OF CHILDHOOD CRANIOPHARYNGIOMA

Between 1965 and 1986, 31 children were treated for craniopharyngioma at the Childrens Hospital of Philadelphia. Total removal was attempted in all patients. Some patients received radiation therapy following subtotal removal. Of the patients whose first resection was subtotal, five received radiation and seven did not. Four of the 5 patients who were radiated (80%) are stable (median 89 months, range 42-155 months) and one recurred at 42 months, failed salvage with total removal, and subsequently died of disease. Of the seven who were not irradiated, all had recurrences (median 12 months, range 3-192 months) and one died of disease. Nineteen patients initially had total removal and none received adjuvant radiation. One patient died postoperatively. Of the 18 remaining patients, 6 had recurrences (median 24 months, range 7-100 months) and 12 (66%) are stable (median 42 months, range 9-133 months). One of these stable patients died of endocrine complications 24 months after initial surgery. Fourteen of the 31 patients recurred. Two died with recurrence and one required no further treatment. Eleven had second resections following initial surgical removal. Seven of these 11 went on to receive radiation and four did not. All seven who were radiated are stable (median 33 months, range 1-228 months); whereas 1 of the 4 who were not radiated recurred again at 18 months. This patient had a third resection followed by radiation therapy and is now stable at 20 months. After initial surgery (and before radiation, when given) 26 of 31 patients had panhypopituitarism, 4 had partial deficits, and 1 was normal. Severe diencephalic syndrome, loss of visual acuity, and intellectual deficits were no more frequent in patients treated with total removal, subtotal removal, and in patients who received radiation. We conclude that radiation has an important role following subtotal removal and for salvage treatment after initial surgery. Aggressive attempt at total removal does result in prolonged progression-free survival in some patients. Extensive resections may result in significant mortality and endocrine morbidity. This review suggests that subtotal removal and radiation results in outcomes at least as favorable as treatment with total removal alone.

Treatment of advanced neuroblastoma with supralethal chemotherapy, radiation, and allogeneic or autologous marrow reconstitution.

Ten children with recurrent metastatic (stage IV) neuroblastoma received local radiation therapy, supralethal chemotherapy, and total-body irradiation. Rescue with infusions of either allogeneic (four patients) or autologous (six patients) bone marrow followed. The drugs given to the first two patients were individualized combinations based on previous tumor responses. Both patients died with recurrent tumor three and nine months posttransplant. The eight remaining patients were treated more uniformly with local irradiation, VM-26, doxorubicin, melphalan (L-phenylalanine mustard), and 1,000-rad total-body irradiation in three fractions. Two of these patients had cardiac dysfunction and received no doxorubicin. Three children died in the immediate posttransplant period with disseminated fungal infections. A fourth relapsed and died nine months posttransplant. As of December 1, 1983, two children who received allogeneic marrow grafts have survived in complete remission for 54 and 36 months, and two children who received autologous marrow grafts have survived in complete remission for 35 and 22 months. These results suggest that relapsed metastatic neuroblastoma can be controlled by supralethal combinations of chemotherapy and irradiation coupled with bone-marrow rescue.