Secondary Brain Tumors in Children Treated for Acute Lymphoblastic Leukemia at St Jude Children's Research Hospital By Andrew W. Walter, Michael L. Hancock, Ching-Hon Pui, Melissa M. Hudson, Judith S. Ochs, Gaston K. Rivera, Charles B. Pratt, James M. Boyett, and Larry E. Kun Purpose: To evaluate the incidence of and potential risk factors for second malignant neoplasms of the brain following treatment for childhood acute lymphoblastic leukemia (ALL). Patients and Methods: The study population consisted of 1,612 consecutively enrolled protocol patients treated on sequential institutional protocols for newly diagnosed ALL at St Jude Children's Research Hospital (SJCRH) between 1967 and 1988. The median follow-up duration is 15.9 years (range, 5.5 to 29.9 y). Results: The cumulative incidence of brain tumors at 20 years is 1.39% (95% confidence interval [CI], 0.63% to 2.15%). Twenty-two brain tumors (10 high-grade gliomas, one low-grade glioma, and 11 meningiomas) were diagnosed among 21 patients after a median latency of 12.6 years (high-grade gliomas, 9.1 years; meningiomas, 19 years). Tumor type was linked to outcome, with patients who developed high-grade tu-

mors doing poorly and those who developed low-grade tumors doing well. Risk factors for developing any secondary brain tumor included the presence of CNS leukemia at diagnosis, treatment on Total X therapy, and the use of cranial irradiation, which was dosedependent. Age less than 6 years was associated with an increased risk of developing a high-grade glioma. Conclusion: This single-institution study, with a high rate of long-term data capture, demonstrated that brain tumors are a rare, late complication of therapy for ALL. We report many more low-grade tumors than others probably because of exhaustive long-term follow-up evaluation. The importance of limiting cranial radiation is underscored by the dose-dependent tumorigenic effect of radiation therapy seen in this study. J Clin Oncol 16:3761-3767. o 1998 by American Society ofClinical Oncology.

TNCREASINGLY EFFECTIVE THERAPIES for childI hood cancers over the past three decades have revolutionized the practice of pediatric oncology. As cure rates increase, more children are at risk for long-term toxicities. Among the most dreaded late sequelae is the development of a second malignant neoplasm.1- 3 This phenomenon, although recognized for many years, is of growing concern as the number of cancer survivors increases. Survivors of childhood acute lymphoblastic leukemia (ALL) form a large population at risk for developing secondary brain tumors. Brain tumors have been described as one of the most frequently seen types of second malignant neoplasm in the ALL survivor population. 4 We evaluated the incidence of this late sequela in a large cohort of patients for whom detailed follow-up information is available. Because of the drastically different outcomes for patients who developed low-grade tumors (primarily meningiomas) as opposed to high-grade tumors (high-grade gliomas), we analyzed these two subgroups separately as well. Finally, we sought to identify patient characteristics and/or therapyrelated risk factors that might play a role in the development of secondary brain tumors following treatment for childhood ALL.

enrolled onto seven sequential Total Therapy studies (V through XI) at St Jude Children's Research Hospital (SJCRH). 5 Specific treatment regimens have been described in detail previously5 and are summarized.

PATIENTS AND METHODS From 1967 through 1988, 1,612 consecutive patients •'18 years of age who had newly diagnosed ALL (or undifferentiated leukemia) were

Antileukemia Therapy The St Jude Total Therapy Program for childhood ALL spans several previously defined eras. Our study focused on eras 2 through 4. Patients treated during the first era (studies I through IV, 1962 to 1966) were excluded from this analysis because CNS-directed therapy was absent or inadequate during this period, and there were few long-term survivors. The second era (studies V through IX, 1967 to 1979, n = 826) was characterized by the introduction of cranial irradiation (age-adjusted, 24 Gy for most patients) and intrathecal methotrexate for

From the Departments of Hematology-Oncology, Radiation Oncology, Pathology and Laboratory Medicine, and Biostatistics, St Jude Children's Research Hospital; and Departments of Pediatrics, and Radiation Oncology, University of Tennessee, Memphis, College of Medicine, Memphis, TN. Submitted June 25, 1997; accepted July 22, 1998. Supported in partby grantsno. P30 CA-21765, POI CA-20180, and POI CA-23099 from the National Cancer Institute, Bethesda, MD; and by the American Lebanese Syrian Associated Charities (ALSAC), Memphis, TN. Presented in part at the Seventh International Conference on Late Effects in CancerSurvivors, June 15, 1996, Buffalo, NY Address reprint requests to Andrew W Walter MS, MD, Department of Hematology-Oncology, St Jude Children Research Hospital,332 North Lauderdale, Memphis, TN 38105; Email [email protected]. © 1998 by American Society of Clinical Oncology. 0732-183X/98/1612-0002$3.00/0

Journal of Clinical Oncology, Vol 16, No 12 (December), 1998: pp 3761-3767

Downloaded from ascopubs.org by 37.44.207.97 on January 28, 2017 from 037.044.207.097 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

3761

WALTER ET AL

3762 the majority of patients. Era 3 (study X, 1979 to 1983, n = 428) incorporated epipodophyllotoxins and modified CNS-directed treatment to increase the use of intrathecal methotrexate and, in lower risk cases, randomized patients to receive low-dose or no cranial radiation (18 Gy). In the fourth era (study XI, 1984 to 1988, n = 358), intensified early chemotherapy was followed by the use of alternating pairs of non-cross-resistant drugs. During this period, triple intrathecal chemotherapy (methotrexate, hydrocortisone, and cytarabine) replaced singleagent methotrexate. CNS irradiation was reserved for patients with higher risk ALL (18 Gy) or those with CNS leukemia at diagnosis (24 Gy). Relapse therapy for patients treated during era 2 relied on the same agents used during initial therapy. Subsequently, patients with relapsed or refractory disease were enrolled onto clinical trials designed to evaluate different strategies of retrieval therapy, including allogeneic bone marrow transplantation. Patients with an isolated CNS relapse were routinely irradiated with craniospinal irradiation and given systemic chemotherapy, with or without intrathecal chemotherapy.

Chart Review Antileukemia therapy delivered was confirmed by direct chart review of any patient who presented with CNS leukemia, failed to respond to induction therapy, had excessive toxicity, had a relapse of any type, or in whom the planned duration of therapy was significantly altered for any reason. CNS-directed therapy, including irradiation and intrathecal chemotherapy, was estimated for patients whose therapy did not deviate from their initial Total Therapy protocol (n = 968 and 961, respectively). In addition, these patients must have been free of CNS leukemia at diagnosis and must have completed the initial protocol without experiencing a major adverse event. These estimates took into account a patient's predicted time on protocol and planned therapy during that time. Any additional therapy given for relapse was tabulated as described earlier.

Follow- Up Procedures After they completed therapy, all patients treated during the study period were examined at least annually by their primary attending oncologist or, after 1984, in our After Completion of Therapy Clinic to monitor remission status and late effects of therapy. Patients aged 18 years and those who remained in remission at least 10 years after diagnosis were discharged from the institution and monitored by their local physician. The status of these patients is monitored by mail questionnaires distributed annually by the hospital's tumor registrar. Secondary malignancies are routinely documented in the SJCRH data base for childhood ALL. Extensive information was available for children cared for at SJCRH for their brain tumors. For patients cared for outside the institution, records were requested from the treating physician. For deaths that occurred outside the institution, death certificates were routinely requested and the reported cause was verified through telephone conversations with the local physician and/or family.3

StatisticalAnalysis The distributions of categorical presenting features (age, CNS disease at diagnosis, sex, and race) were compared between patients with and without secondary brain tumors by Fisher's exact test. The cumulative incidence functions of secondary brain tumors in patient subgroups 7 were estimated 6 and were compared by Gray's test. Patients' times at risk were calculated as time from diagnosis of ALL to development of a

first brain tumor, development of any other second malignant neoplasms, or last follow-up date for surviving patients without second neoplasms. Death or the development of any other type of second malignancy was considered a competing event.

RESULTS The presenting features and patterns of failure of the 1,612 patients treated for newly diagnosed ALL were similar across the three consecutive treatment eras. 5 The median follow-up duration for all patients is 15.9 years (range, 5.5 to 29.9 y). Of note, 97.1% of surviving patients have been contacted in the 2 years before the date of this analysis, April 1998. Incidence and Histology of Brain Tumors A total of 21 children have developed brain tumors, for a cumulative incidence of 1.39% (95% confidence interval [CI], 0.63% to 2.15%) at 20 years (Fig 1). Individual patient characteristics are listed in Table 1. Pathologic diagnoses consisted of glioblastoma multiforme, n = 4; anaplastic astrocytoma, n = 2; other high-grade glioma, n = 4; low-grade oligodendroglioma, n = 1; and meningioma, n = 11. The median latency from initial diagnosis to the diagnosis of a brain tumor was 12.6 years (range, 5.9 to 29). There was a noticeably longer latency to the development of meningioma (median, 19 years) compared with high-grade gliomas (median, 9.1 years) (Fig 1). Tumor type (low-grade v high-grade) was strongly linked to outcome. Patients with low-grade tumors, which includes all 10 patients with meningioma and the patient with a low-grade glioma, are alive at a median of 2.5 years following diagnosis of the brain tumor (range, 0.5 to 10 years). In contrast, of 10 patients with high-grade gliomas, eight have died (median survival following diagnosis of brain tumor, 7 months; range, 0.1 to 25 months). Two patients remain alive, both with high-grade gliomas, 5 months and 7.8 years following diagnosis of their brain tumors. Risk Factors Patient characteristics. Selected patient characteristics and summaries of CNS treatment for the 21 patients with brain tumors and the remainder of the cohort are compared in Table 2. The presence of CNS leukemia at diagnosis was associated with an increased risk of subsequent brain tumor (Fig 2). The cumulative incidence at 20 years of developing a secondary brain tumor of any type was 4.17% (95% CI, 0% to 8.97%) for patients who presented with CNS leukemia, compared with 1.25% (95% CI, 0.51% to 1.99%) for other patients (P = .002). The presence of CNS leukemia also increased the risk of developing a high-grade tumor (cumulative incidence at 20 years, 2.78% [95% CI, 0% to

Downloaded from ascopubs.org by 37.44.207.97 on January 28, 2017 from 037.044.207.097 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

BRAIN TUMORS AFTER THERAPY FOR ALL

3763

0.1 0.09

-

0.08 CD

Q

'_

Xl

0.07 0.06

---

Any Brain Tumor Meningiomas

-- --

High-Grade Gliomas

.) 0.05

'5 X 0.04 E0.03 0 0.02

-AL_

I ,,,

0.01 0 0

5

10

15

i~,~

~,

30

35

25

20

Years Post Diagnosis

Fig 1. Cumulative incidence rate of any secondary brain tumor and by od ALL mrad histologic subtype for 1,612 patients treated for childhcoad ALL

6.72%] v 0.6% [95% C, 0.16% to 1.04 I%] for all other patients; P = .014). There is confoundin,g between CNS leukemia at diagnosis and the dose of clranial irradiation subsequently delivered; that is, patients wit h CNS leukemia were treated with cranial irradiation at a higher dose. However, CNS leukemia at diagnosis does remain a significant factor, even when the analysis is adj usted for cranial irradiation dose, as detailed later. Young age at diagnosis of ALL (5 Sex Male Female Race White Nonwhite CNS leukemia at diagnosis Yes No Cranial radiation dose, Gy 0 10-21 > 21-30 > 30 Intrathecal administrations 0 1-10 > 10

AllOther Patients

Patients With Brain Tumors

In 1,591)

(n21)

Cumulative(n10) Incidence

No.

%

No.

%

930 661

58 42

16 5

890 701

56 44

1,426 165

P

No.

%

77 23

1.98 0.53

.104

9 1

90 10

1.08 0.15

.045

10 11

48 52

1.35 1.45

.474

6 4

60 40

0.76 0.63

.803

90 10

20 1

95 5

1.42 1.24

.489

9 1

90 10

0.67 1.24

.967

69 1,522

4 96

3 18

14 86

4.17 1.25

.002

2 8

20 80

2.78 0.60

.014

361 384 788 58

23 24 50 4

0 2 5 4

0 10 71 19

0 1.03 1.65 3.23

.015

0 2 6 2

0 20 60 20

0 1.03 0.76 3.23

.043

118 827 646

7 52 41

1 12 8

5 57 38

0.84 1.6 0.96

.234

0 5 5

0 50 50

0 0.74 0.81

.625

0.1

0.1 0.09

0.07

>,

005

.5

0.04

E

0.03

-

CNS at DX

---

No CNS at DX

(%)P

The patients were then divided into four groups based on cranial radiation dose: no irradiation, 10 to -21 Gy, 21 to 30 Gy, and more than 30 Gy. There was a statistically significant increased risk of developing a secondary brain tumor with increasing irradiation dose (Fig 4) regardless of whether all brain tumors or only high-grade tumors were considered in the analysis. The cumulative incidence at 20 years for all brain tumors was 1.03% (95% CI, 0% to 2.55%), 1.65% (95% CI, 0.69% to 2.61%), and 3.23% (95% CI, 0% to 7.69%) for the four dose levels, respectively. The cumulative incidence at 20 years for developing a high-grade tumor specifically was 0%, 1.03% (95% CI, 0% to 2.55%), 0.76%

0.09

U ·uo

20-Year

Incidence

(%)

during era 4 are certainly at continued risk for developing a meningioma. There appeared to be no relationship between the development of a brain tumor and the pattern or history of relapse. Specifically, the presence of CNS relapse did not identify a group at increased risk. Cranial irradiation. There were no instances of brain tumors among the 361 patients who did not receive cranial irradiation as part of their therapy for leukemia. In contrast, for the 1,251 patients who underwent cranial irradiation initially or at relapse, the 20-year cumulative incidence of brain tumors was 1.6% (95% CI, 0.8% to 2.48%; P = .058).

0.08

Patient Wih Hig-Grade Tumors

20-Year

0.08 a) 0

0.06

0.07

-

5

---

6 or over

or under

0.06 -a

'a

0.05 0.04

E 0.03

U

0.02 0.01

0.02 0.01

0 5

10

15

20

25

30

35

Years Post Diagnosis

Fig 2. Cumulative incidence rate of secondary brain tumors by CNS status at diagnosis. Dx, diagnosis.

I-

0 0

5

10

15

20

25

I

I

30

35

Years Post Diagnosis Fig 3. Cumulative incidence rote of secondary high-grade gliomas by age (5 years or younger, 6 years or older) at diagnosis.

Downloaded from ascopubs.org by 37.44.207.97 on January 28, 2017 from 037.044.207.097 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

BRAIN TUMORS AFTER THERAPY FOR ALL

3765

0.2

evaluation (> 90%), not only for the patients who developed secondary brain tumors, but for the remainder of the cohort as well; we were also careful to document relapse therapy as

0.18 0.16 0 C -0

.0

E E

No RT

,--

0.14

-

21Gy 21-30Gy

0.12

well.

I

i

~

The estimated 20-year cumulative incidence was 1.39% for any brain tumor in this series and 0.7% for high-grade tumors. Several other studies have shown an increased risk second malignancies in survivors of childhood ALL. A review of 9,720 patients treated on protocols of the Childrens Cancer Study Group (CCG) between 1972 and 1988

35

identified 43 second malignancies, 24 of which were brain

----> 30Gy

0.1 0.08 0.06

0

-

0.04 0.02 o!

0 0

5

.-......... ..,-r'___ iT-T 15

10 Yea

' . - ------..---

20

25

-

-of

30

arsPost Diagnosis

Fig 4. Cumulative incidence roate of secondary brain tumors by dose received. radiation (RT)

total

(95% CI, 0.14% to 1.3%) and 3.23% (95% CI, 0% to 7.79%) respectively. CNS leukemia. The pre: sence of CNS leukemia at diagnosis was associated with aln increased risk of developing a secondary brain tumor, as described earlier. However, patients who presented with C'NS leukemia at diagnosis were treated with higher doses o)fcranial irradiation (P < .001) which makes it difficult to prove that CNS leukemia is an independent risk factor. Tvwo separate, stratified analyses were performed adjusting ffor each factor. CNS disease at diagnosis remained a signify ficant prognostic factor for the development of a secondary brain tumor after adjustment for radiation dose (P = .007) an Ad,likewise, irradiation dose was significant after adjustment for CNS disease at diagnosis (P = .038). If only high-gradle tumors are considered, irradiation dose adjusted for CNS ddisease at diagnosis is significant at the P = .082 level and thhe presence of CNS leukemia at diagnosis adjusted for irrad iation dose is significant at the P = .038 level. Intrathecalchemotherapy. The use of intrathecal chemotherapy did not increase the risk of developing a secondary brain tumor (P = .96). Thelre was only one brain tumor in the 119 patients who did no areceuaive inrahecal chemotherapy, which resulted in a 2( )-yearcumulative incidence of 0.84% (95% CI, 0% to 2.5 compared 52%), with a 20-year cumulative incidence of 1., 4% (95% CI, 0reated6% to 2.12%) among the 1,493 patients wl chemotherapy. DISC USSION In this single-institution study, we determined the incidence of both high-grade (nmalignant) and low-grade brain tumors in a cohort of prosjpectively treated children with ALL for whom long-term fo Ilow-up data are available. This study is unique in the higl h rate of successful follow-up

tumors (14 high-grade gliomas, three primitive neuroectodermal tumors, two meningiomas, two low-grade gliomas, and one case each of brainstem glioma, ependymoma, and medulloblastoma). 4 The 15-year estimated cumulative incidence of any second malignancy in the CCG experience was 2.53% (95% CI, 1.74% to 3.38%), which represents a sevenfold increase compared with the age-adjusted incidence in the general population. For brain tumors, the risk was substantially higher, and exceeded the expected rate by more than 20-fold. Two European studies have also found an increased risk of brain tumor as a second malignancy in patients treated for childhood leukemia. 9 ,10 An important difference between other retrospective series and this study is the underrepresentation of low-grade tumors reported by others. The proportion of low-grade tumors in the CCG series was 17%. The two European studies document only one low-grade tumor between them. In the SJCRH series, 52% of patients with secondary brain tumors had low-grade tumors. It is likely that low-grade tumors (including low-grade gliomas and meningiomas) are missed in late follow-up evaluation in a multiinstitutional setting. Follow-up study at SJCRH is comprehensive, with more than 95% of offtherapy patients still being monitored. This high rate of monitoring increases the chance that benign or less aggressive tumors are captured in the late-effects data base. Because high-grade tumors have the most profound impact on patient health (in contrast to low-grade tumors), they were analyzed separately in this study. The median latency from the initial diagnosis of ALL to the diagnosis of brain tumor was 9.1 years for high-grade gliomas and 19 years for meningiomas (12.6 years overall). These intervals are comparable to those reported previously for treatment-related malignant gliomas and meningiomas in survivors of various types of cancer.'' 5 Latency has been noted to be inversely proportional to irradiation dose in prior studies, with shorter latencies often seen following higher irradiation doses. The shortest latencies are often seen in patients who develop high-grade gliomas, as noted in this study as well 8' 5 ' 8(Fig I).

Downloaded from ascopubs.org by 37.44.207.97 on January 28, 2017 from 037.044.207.097 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

WALTER ET AL

3766 The outcome in our 10 patients with high-grade gliomas has been dismal. Eight have died (median survival time, 7 months), one is alive with disease, and one is alive with no evidence of disease. This experience is typical for highgrade gliomas as a second neoplasm.' 8 Primary high-grade gliomas in children are associated with a slightly better outlook, with a median survival time of approximately 2 years. 1920 All 10 patients in our series with secondary meningiomas are alive. We have not observed a clinically malignant course in those patients, as has been reported by others for radiation-related meningiomas. '4,15,21,22 The cumulative incidence of both low-grade and highgrade brain neoplasms was related to the presence of CNS leukemia at diagnosis. This observation is almost certainly related to the more intensive CNS-directed chemotherapy and irradiation given to these patients. The confounding of CNS disease at diagnosis with more intensive therapy and the relatively small numbers of events makes any attempt to establish independent prognostic value for the factors significant by univariate analysis difficult. Nevertheless, after stratified analyses, both CNS leukemia at diagnosis and cranial irradiation dose remain significant predictors of an increased risk of developing a brain tumor. Neither race nor sex was a risk factor in this series. Young age at diagnosis of ALL was associated with an increased risk of developing a high-grade brain tumor, but did not increase the risk of a secondary brain tumor overall. Brain tumors occurred only among previously irradiated patients in our series, as in the CCG study.4 Increasing doses of cranial irradiation for ALL were associated with an increased risk of secondary brain tumors in general and high-grade tumors specifically in this study. Patients treated during era 3 (Total X) had a higher risk of developing a secondary brain tumor. We looked for major differences in therapy between eras 3 and 4. There were no substantial differences in irradiation schedules between the treatment eras. Epipodophyllotoxins (teniposide) were introduced during era 3, but scheduled doses were, on average, lower than in era 4. Finally, cyclophosphamide was used in both eras 3 and 4 but at a lower dose and in fewer patients in era 3. Hence, the increased risk of brain tumors in patients treated in era 3 remains unexplained.

Genetic factors may be important in the development of multiple malignancies. Patients with leukemia may be at increased risk of developing gliomas regardless of therapy. 1' 23- 25 ALL and brain tumors may both be components of familial, hereditary cancer syndromes (eg, LiFraumeni) 2 ,23 ,26- 30 and reports have described the diagnosis of glioma coincident with or shortly after the diagnosis of ALL.24 ' 26.30 '31 However, we did not attempt to demonstrate a role for genetic factors in the development of secondary brain tumors in this population (ie, screening for p5 3 mutations and similar testing) so the role genetics might play in this series is unknown. The incidence of second malignancies, including brain tumors, is changing over time as a result of differences in therapy and outcome in ALL. The incidence of high-grade gliomas following treatment for ALL may be nearing a peak since the widespread use of higher dose (24 Gy) cranial radiation in front-line therapy is now 10 to 15 years in the past, exceeding the median latency for this tumor.3 2 In contrast, the diagnosis of meningiomas after ALL therapy will probably continue to increase, given that the latency period frequently extends beyond two decades following diagnosis and meningiomas may be more frequent following lower dose radiation. 9 33 A continuing challenge in the development of riskadapted therapies for ALL is to balance the need for effective CNS prophylaxis and treatment with the risks of subsequent relapse and of unacceptable sequelae. The substitution of more intensive intrathecal chemotherapy for routine cranial radiation in many centers was designed to achieve this goal. If this approach leads to higher CNS relapse rates, and thus to the need for intensive salvage therapy, including highdose radiation, it may be necessary to explore other approaches. It will also be important to document the impact of intrathecal chemotherapy in long-term survivors who have received no or low-dose cranial radiation.

ACKNOWLEDGMENT We thank data managers, Mary Heim and Kim Juneau for their excellent work, the ladies of the Tumor Registry for their diligence, and Christy Wright for editorial assistance.

REFERENCES 1. Meadows AT, D'Angio GJ, Mike V, et al: Patterns of second malignant neoplasms in children. Cancer 40:1903-1911, 1977 2. Hawkins MM, Draper GJ, Kingston JE: Incidence of second primary tumours among childhood cancer survivors. Br J Cancer 56:339-347, 1987 3. Hudson MM, Jones D, Boyett J, et al: Late mortality of long-term survivors of childhood cancer. J Clin Oncol 15:2205-2213, 1997 4. Neglia JP, Meadows AT, Robison LL, et al: Second neoplasms after acute lymphoblastic leukemia in childhood. N Engl J Med 325:1330-1336, 1991

5. Rivera GK, Pinkel D, Simone JV, et al: Treatment of acute lymphoblastic leukemia. 30 years' experience at St Jude Children's Research Hospital. N Engl J Med 329:1289-1295, 1993 6. Kalbfleisch JD, Prentice RL: Statistical Analysis of Failure Time Data. New York, NY, Wiley, 1980 7. Gray RJ: A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat 16:1141-1154, 1988 8. Salvati M, Artico M, Caruso R, et al: A report on radiationinduced gliomas. Cancer 67:392-397, 1991 9. Olsen JH, Garwicz S, Hertz H, et al: Second malignant neoplasms

Downloaded from ascopubs.org by 37.44.207.97 on January 28, 2017 from 037.044.207.097 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

BRAIN TUMORS AFTER THERAPY FOR ALL after cancer in childhood or adolescence. Nordic Society of Paediatric Haematology and Oncology Association of the Nordic Cancer Registries. BMJ 307:1030-1036, 1993 10. Rosso P, Terracini B, Fears TR, et al: Second malignant tumors after elective end of therapy for a first cancer in childhood: A multicenter study in Italy. Int J Cancer 59:451-456, 1994 11. Liwnicz BH, Berger TS, Liwnicz RG, et al: Radiation-associated gliomas: A report of four cases and analysis of postradiation tumors of the central nervous system. Neurosurgery 17:436-445, 1985 12. Bernstein M, Laperriere N: Radiation-induced tumors of the nervous system, in Gutin PH, Leibel SA, Sheline GE (eds): Radiation Injury to the Nervous System. New York, NY, Raven, 1991, pp 455-472 13. Harrison MJ, Wolfe DE, Lau TS, et al: Radiation-induced meningiomas: Experience at the Mount Sinai Hospital and review of the literature. J Neurosurg 75:564-574, 1991 14. Ghim TT, Seo JJ, O'Brien M, et al: Childhood intracranial meningiomas after high-dose irradiation. Cancer 71:4091-4095, 1993 15. Bondy M, Ligon BL: Epidemiology and etiology of intracranial meningiomas: A review. J Neurooncol 29:197-205, 1996 16. Ron E, Modan B, Boice JD Jr, et al: Tumors of the brain and nervous system after radiotherapy in childhood. N Engl J Med 319:1033-1039, 1988 17. Cavin LW, Dalrymple GV, McGuire EL, et al: CNS tumor induction by radiotherapy: A report of four new cases and estimate of dose required. Int J Radiat Oncol Biol Phys 18:399-406, 1990 18. Kaschten B, Flandroy P, Reznik M, et al: Radiation-induced gliosarcoma. Case report and review of the literature. J Neurosurg 83:154-162, 1995 19. Sposto R, Ertel IJ, Jenkin RD, et al: The effectiveness of chemotherapy for treatment of high grade astrocytoma in children: Results of a randomized trial. A report from the Childrens Cancer Study Group. J Neurooncol 7:165-177, 1989 20. Finlay JL, Boyett JM, Yates AJ, et al: Randomized phase III trial in childhood high-grade astrocytoma comparing vincristine, lomustine, and prednisone with the eight-drugs-in-l-day regimen. Childrens Cancer Group. J Clin Oncol 13:112-123, 1995

3767 21. Soffer D, Gomori JM, Siegal T, et al: Intracranial meningiomas after high-dose irradiation. Cancer 63:1514-1519, 1989 22. Rubinstein AB, Shalit MN, Cohen ML, et al: Radiation-induced cerebral meningioma: A recognizable entity. J Neurosurg 61:966-971, 1984 23. Farwell J, Flannery JT: Cancer in relatives of children with central-nervous-system neoplasms. N Engl J Med 311:749-753, 1984 24. Malone M, Lumley H, Erdohazi M: Astrocytoma as a second malignancy in patients with acute lymphoblastic leukemia. Cancer 57:1979-1985, 1986 25. Nelson DF, Cooper S, Weston MG, et al: Second malignant neoplasms in patients treated for Hodgkin's disease with radiotherapy or radiotherapy and chemotherapy. Cancer 48:2386-2393, 1981 26. Ahsan H, Neugut AI, Bruce JN: Association of malignant brain tumors and cancers of other sites. J Clin Oncol 13:2931-2935, 1995 27. Moll AC, Imhof SM, Bouter LM, et al: Second primary tumors in patients with hereditary retinoblastoma: a register-based follow-up study, 1945-1994. Int J Cancer 67:515-519, 1996 28. Meadows AT, Baum E, Fossati-Bellani F, et al: Second malignant neoplasms in children: An update from the Late Effects Study Group. J Clin Oncol 3:532-538, 1985 29. Tucker MA, D'Angio GJ, Boice JD Jr, et al: Bone sarcomas linked to radiotherapy and chemotherapy in children. N Engl J Med 317:588-593, 1987 30. Heyn R, Haeberlen V, Newton WA, et al: Second malignant neoplasms in children treated for rhabdomyosarcoma. Intergroup Rhabdomyosarcoma Study Committee. J Clin Oncol 11:262-270, 1993 31. Draper GJ, Sanders BM, Kingston JE: Second primary neoplasms in patients with retinoblastoma. Br J Cancer 53:661-671, 1986 32. Nesbit ME Jr, Sather HN, Robison LL, et al: Presymptomatic central nervous system therapy in previously untreated childhood acute lymphoblastic leukaemia: Comparison of 1800 rad and 2400 rad. A report for Children's Cancer Study Group. Lancet 1:461-466, 1981 33. Bhatia S, Robison LL, Oberlin O, et al: Breast cancer and other second neoplasms after childhood Hodgkin's disease. N Engl J Med 334:745-751, 1996

Downloaded from ascopubs.org by 37.44.207.97 on January 28, 2017 from 037.044.207.097 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.