Central nervous system (CNS) treatment is an essential component

Transcription

1 2608 A Longitudinal Magnetic Resonance Imaging Study of the Brain in Survivors of Childhood Acute Lymphoblastic Leukemia Arja H. Harila-Saari, M.D., Ph.D. 1 Eija L. Pääkkö, M.D., Ph.D. 2 Leena K. Vainionpää, M.D., Ph.D. 1 Juhani Pyhtinen, M.D., Ph.D. 2 B. Marjatta Lanning, M.D., Ph.D. 1 1 Department of Pediatrics, Oulu University Central Hospital, Oulu, Finland. 2 Department of Diagnostic Radiology, Oulu University Central Hospital, Oulu, Finland. BACKGROUND. The objective of this study was to evaluate changes in magnetic resonance imaging (MRI) of the brain in children with acute lymphoblastic leukemia (ALL) during the first 5 years after the cessation of therapy and to correlate MRI abnormalities with neuropsychologic outcome. METHODS. Thirty-two children with ALL were studied at the end of treatment and 5 years later by brain MRI and the results were compared with the neuropsychologic findings. Fifteen patients had received chemotherapy alone and 17 had received chemotherapy plus cranial radiation. RESULTS. MRI of the brain was abnormal in 6 of 30 patients at the end of treatment and in 8 of 32 patients 5 years later. White matter changes (WMC) were found in 3 patients at the end of treatment and in 4 patients 5 years later. Two patients had developed new mild changes, whereas in one case WMC had normalized during the follow-up. Two patients had old hemorrhages or calcifications at each examination, with some improvement after follow-up, although one case revealed a new calcification or hemorrhage. Signs of cortical atrophy were observed in five patients at both evaluations. The patients with abnormal MRI findings did not differ significantly in their performance in the neuropsychologic tests from the patients with normal MRI findings, but the two patients with persistent WMC had a depression of verbal functions. CONCLUSIONS. Abnormalities in brain MRI were infrequent at the end of treatment for childhood ALL and 5 years later. They did not appear to correlate significantly with neuropsychologic outcome. Brain MRI is not very informative as a routine follow-up method during the first 5 years after treatment. Cancer 1998;83: American Cancer Society. Presented in part at the 29th Meeting of the International Society of Paediatric Oncology, Istanbul, Turkey, September 23 27, Supported by the Fund for Children s Cancer, Oulu, Finland, the Foundation for Pediatric Research in Finland, and the Finnish Cancer Foundation. Address for reprints: Arja Harila-Saari, M.D., Ph.D., Department of Pediatrics, Oulu University Hospital, Oulu, Finland. Received February 3, 1998; revision received May 5, 1998; accepted May 5, KEYWORDS: acute lymphoblastic leukemia, children, magnetic resonance imaging, brain, neuropsychology, intelligence quotient, late effects. Central nervous system (CNS) treatment is an essential component of the successful management of acute lymphoblastic leukemia (ALL) in children. However, both radiotherapy and chemotherapy, either alone or in combination, may lead to both structural and functional changes in the CNS. 1,2 The main types of the structural brain damage recorded have been white matter destruction, vascular damage leading to hemorrhage and calcifications, and enlargement of ventricles and/or sulci, a sign of cortical atrophy. 3 9 Magnetic resonance imaging (MRI) has been found to be more sensitive than computed tomography (CT) in identifying these treatment-related changes, with the exception of calcifications, which are detected more readily by CT. 5,7,8,10 The frequency of abnormalities in MRI scans of patients with ALL has varied 1998 American Cancer Society

2 Brain MRI after Treatment for ALL/Harila-Saari et al widely, 5,6,9,11,12 most likely because of the differences in patient recruitment, CNS treatment, and time elapsing since therapy. To our knowledge, few attempts have been made to compare MRI changes with neuropsychologic findings, and little or no correlation has been found. 5,6,8,11 The development of brain changes may be delayed in patients treated for ALL, and there is evidence that neuropsychologic deficits may progress with time. 13,14 To our knowledge, there are no systematic follow-up studies of the MRI changes after treatment, and the natural history of these changes is unknown. We performed 5-year follow-up brain MRI examinations on 32 consecutive survivors of ALL after treatment to determine the incidence of MRI abnormalities with time and also to study the relation between these abnormalities and the neuropsychologic outcome. PATIENTS AND METHODS Patients Thirty-four patients admitted to the Department of Pediatrics at the University of Oulu for initial ALL treatment between February 1986 and May 1990 whose disease still was in complete, continuous remission 5 years after cessation of therapy were recruited. The initial evaluations had been performed between February 1989 and January 1993, and the results concerning white matter changes (WMC) visible in MRI scans of the brain had been reported earlier for 27 of these patients, 10 1 of whom, a patient with with moderate WMC, had since died and one patient with normal MRI findings who had had a recurrence. The reevaluations were performed between April 1994 and May Two patients who had moved away from the district refused to take part in the 5-year evaluations and were excluded from the series. The research was approved by the Ethical Committee of the Medical Faculty of the University of Oulu, and an informed consent was obtained from all patients and from the parents of those patients age 18 years. The group was comprised of 10 boys and 22 girls, ages 8 24 years (median, 13.2 years) at reevaluation. Two of the patients had had CNS leukemia with more than five blasts in the cerebrospinal fluid at the time of diagnosis but no clinical signs of CNS involvement or later CNS leukemia. One patient had been born prematurely at the 31 weeks gestation with a birth weight of 2100 g, but had normal developmental data and normal neurologic findings at the time of diagnosis. One patient had rheumatoid oligoarthritis, and one had had asthma at the time of diagnosis. One patient had had a slight delay in her verbal and motor development. All the other patients had a normal developmental history. Three of the patients had a history of seizures: one had had a single seizure during induction therapy at a time of septic infection and hyponatremia, one had had a seizure during oral maintenance therapy without causal explanation and no further seizures, and one had developed epilepsy 2 years after cessation of therapy. The patients were divided into standard, intermediate, and high risk treatment groups according to international criteria used in all the Nordic countries. 15 Ten of the 13 standard risk patients had been treated according to the Nordic protocol, 15,16 in which the CNS treatment is comprised of 8 intravenous infusions of methotrexate, 1.0 g/m 2 /24 hours, and 13 intrathecal injections of methotrexate, 12 mg/m 2, but no cranial irradiation. The remaining three standard risk patients received only 3 infusions and 7 intrathecal injections of methotrexate, in accordance with an earlier Nordic protocol. The 7 intermediate risk patients and 9 of the 12 high risk patients were treated according to the BFM-83 protocol, 17 in which the CNS treatment includes 4 or 6 pulses of intravenous methotrexate, 500 mg/m 2 / 24 hours and 9 or 11 intrathecal injections of methotrexate (8 12 mg), for the intermediate risk and high risk patients, respectively, as well as cranial irradiation. Three high risk patients had been treated according to the Nordic protocol, which includes 2 or 4 (for children age 5 years) pulses of intravenous infusions, 8 g/m 2 /24 hours, and maximum of 17 or 15 intrathecal injections of methotrexate, respectively, as well as cranial irradiation for those patients age 5 years at diagnosis. Two intermediate risk/high risk patients did not receive cranial irradiation because of their young age. The dose of cranial irradiation was 18 gray (Gy) in 5 cases, 22 Gy in 1 case, 24 Gy in 10 cases, and 30 Gy in 1 case. The treatment protocols have been presented in detail in our previous reports. 18,19 The characteristics of the study patients are summarized in Table 1. MRI of the Brain MRI imaging was performed using a 1.0-tesla (T) scanner at the end of treatment as reported previously, 10 and with either a 1.0-T (Magnetom; Siemens, Erlangen, Germany) scanner (28 patients) or a 1.5-T (Signa Horizon Echo Speed; General Electric, Milwaukee, WI) scanner (4 patients) at follow-up. T2-weighted axial and coronal (turbo spin echo or fast spin echo) images repetition time [TR] 3500, echo time [TE] 14 15, 93 98, 1 excitation, 5-mm slice, matrix, 23-cm field of view [FOV]) together with sagittal T1- weighted images (TR , TE 9 15, 2 excitations, 3- or 5-mm slice, matrix, cm

3 2610 CANCER December 15, 1998 / Volume 83 / Number 12 TABLE 1 Characteristics of the 32 Survivors of Childhood ALL Variable No. of patients Gender Male 10 Female 22 Patients in risk groups Standard risk 13 Intermediate risk 7 High risk 12 Patients treated with chemotherapy alone 15 Patients treated with chemotherapy and cranial irradiation 17 Patients age 5 years at diagnosis 18 Patients age 5 years at diagnosis and treated with cranial irradiation 10 Years (SD) Mean age at diagnosis (SD) 5.3 (3.5) Mean age at the end of treatment 8.4 (3.5) Mean age at neuropsychologic assessment and second brain MRI 13.4 (3.6) Mean time elapsing from treatment at reevaluation 5.0 (0.4) ALL: acute lymphoblastic leukemia; SD: standard deviation; MRI: magnetic resonance imaging. FOV) were obtained in all patients except five who did not undergo the coronal scan. MRI studies were reviewed independently by two radiologists (one of whom is a neuroradiologist [J.P.] whereas the other has a special interest in pediatric neuroradiology [E.P.]) and a consensus was reached. Special attention was paid to parenchymal signal abnormalities, especially increased signals in the white matter on the T2-weighted images, 10 and to the sizes of the ventricles and sulci. The degree of white matter abnormality was graded as mild, moderate, or severe if 0 25%, 25 50%, or 50%, respectively, of the frontal, parietal, and/or occipital white matter was involved. When enlargement of the ventricles or sulci was observed, the finding was compared with the initial CT finding at diagnosis, which was available for all patients. Only enlargement that had developed after diagnosis were regarded as treatment-related change. The cranial CT revealed a small hemorrhage in the lenticular nucleus in one patient at diagnosis; 6 months later her CT scans were normal as were her MRI scans. Two patients had not undergone MRI of the brain at the cessation of therapy, but had normal CT scan findings at diagnosis as well as during and at the end of therapy, and their MRI findings at 5-year reevaluation were normal. Therefore they were included in this series. Neuropsychologic Evaluations A clinical neuropsychologic assessment was performed in four separate sessions of approximately 45 minutes by a neuropsychologist who was aware of the patient s history but was unaware of the results of the other examinations. The patient and the parents were interviewed to assess how the patient was coping at school. An age-appropriate Wechsler intelligence test Wechsler Intelligence Scale for Children (WISC) 20 was performed in patients ages 6 15 years, and a Wechsler Adult Intelligence Scale (WAIS) 21 was performed on those patients age 16 years. The revised version of the WISC or WAIS was not used because a group of these patients also were taking part in a neuropsychologic follow-up study initiated in our hospital in 1986 using the old versions of the tests. The Finnish versions of the tests and the Finnish standards were used. A neuropsychologic assessment for children, NEPSY, 22 was obtained. It is comprised of 37 tests that measure various aspects of attention, language, motor and sensory functions, visuospatial functions, and memory. Tests covering all these areas were included according to the patient s age. The Developmental Test of Visual-Motor Integration 23 also was administered. The NEPSY has been standardized in Finland in healthy children from different age groups. 24 Standard deviation scores (z scores) based on the distribution of the Finnish norm group of corresponding age were calculated. Neuropsychologic function was interpreted as deficient if the score on three or more tests of that area fell below one standard deviation (z score -1), or two or more tests were impaired and at least one test fell below two standard deviation scores (z score -2). Statistical Analysis The Statistical Package for Social Sciences (SPSS for Windows), version 7.0, was used to analyze the data. The Fisher s exact probability test was used to assess the differences in the number of abnormalities between the groups, and the Mann-Whitney U test was used to assess the differences between the groups in neuropsychological test scores. Two-tailed probability values 0.05 were considered significant. RESULTS MRI Six of the 30 patients (20%) had treatment-related abnormalities found in the MRI scan at the end of treatment and 8 of the 32 patients (25%) had treatment-related abnormalities detected on MRI scan 5 years later. Six patients had changes at both evaluations. Three patients had WMC in the MRI scans obtained at the end of treatment, 10 with 1 case being classified as moderate and 2 as mild (the moderate

4 Brain MRI after Treatment for ALL/Harila-Saari et al FIGURE 1. (A) Periventricular high intensity white matter changes on a T2-weighted image at the end of treatment in a 10-year-old patient with high risk acute lymphoblastic leukemia treated with chemotherapy and cranial irradiation. (B) Five years later the lesions had diminished. case occurring in the chemotherapy group). One case of mild WMC had normalized during follow-up, whereas the remaining two cases had persisted, although some improvement was observed (Fig. 1). Two new patients had developed new mild WMC after the end of treatment, one in the irradiated group and one in the chemotherapy group (Fig. 2) (Table 2). One patient had a small atrophic area in the brain parenchyma that had increased in size during followup. It was not classified as WMC because it may have been a consequence of hemorrhage or ischemia and no other WMCs were observed (Table 2). Two cases of old hemorrhage or calcification were found at each examination, with some improvement after follow-up in both cases, although one case revealed a new small calcification or hemorrhage. Enlarged ventricles and/or sulci as a sign of brain atrophy were observed at both evaluations in five patients, and the finding remained unchanged. Atrophy and hemorrhages or calcifications were observed only in the irradiated patients (Table 2). There generally were more abnormalities in the irradiated group of patients, but the difference was not statistically significant. Both patients with CNS leukemia at diagnosis had normal MRI scans. Neuropsychologic Findings All the patients were able to attend a normal school. Six patients had completed their compulsory education; of these patients three were continuing their education in an upper secondary school, one was going to a vocational school, and two already had graduated from a vocational school, with one having taken up full-time work whereas the other was unemployed. Twelve patients (38%) were reported to have had learning difficulties at school, including 9 patients (28%) who required special arrangements, usually remedial education or special rehabilitation (e.g., neuropsychologic or speech therapy), and one patient who had had to repeat a year at school. Four of the 8 patients with abnormal MRI findings (50%) had learning difficulties: both of the 2 patients with persistent WMC, 1 patient with WMC at the end of treatment, and 1 patient with cortical atrophy. However, only those 4 of the 12 patients with reported learning difficulties at school (33%) had abnormal MRI findings. None of the patients had a total or performance IQ 90, although 4 patients had a verbal IQ 90, 3 of whom also had abnormal MRI findings (2 with persistent WMC and 1 with atrophy). The mean IQ test

5 2612 CANCER December 15, 1998 / Volume 83 / Number 12 FIGURE 2. (A) Normal white matter on a coronal T2-weighted image at the end of treatment in a 9-year-old patient with high risk acute lymphoblastic leukemia treated with chemotherapy and cranial irradiation. (B) Five years later a small high intense area appeared in the left peritrigonal white matter. results from the eight patients with treatment-related brain MRI abnormalities did not differ significantly from those of the patients with normal MRI findings. The verbal impairment in the two patients with persistent WMC was most notable in the Information, Arithmetic and Digit span subtests (Table 3). The IQ scores of the other subgroups with atrophy, calcifications, or hemorrhage, or the group with transient WMC, did not differ from the other patients. The irradiated patients performed at a lower level on the IQ tests than the patients treated with chemotherapy alone, except in the digit span test measuring short term auditive memory, in which the chemotherapy group patients scored lower. The irradiated patients scored significantly lower on verbal IQ, information, and comprehension (Table 3). The IQ scores of patients who were age 5 years at diagnosis did not show a statistically significant difference compared with the IQ scores of the older patients, nor did the 10 patients who had undergone cranial irradiation and been diagnosed before age 5 years differ in this respect from the other patients as a whole, or from the 8 patients who were treated at a similar age without cranial irradiation. The visuospatial and memory functions were the most commonly impaired areas of the neuropsychologic test battery. Verbal and motor functions were impaired less frequently, and only three patients were found to have attention difficulties. Again there was no difference between the patients with a normal or abnormal MRI scan, or between the irradiated or nonirradiated patients (Table 4). DISCUSSION The overall incidence of treatment-related abnormalities in the brain MRI after treatment for ALL was 25% in this series. Previous studies including survivors of ALL have reported these abnormalities in 0 53% of cases after treatment, but there has been great variation in patient recruitment, time elapsing since treatment, and in the treatment regimens, and an additional common deficiency has been the small number of patients (Table 5). 5,6,9,11,12,25 28 The incidence of persistent WMC also was low because there were only two cases, one in each CNS treatment group. The patient with persistent WMC in the chemotherapy group had been born prematurely, which may have acted as a predisposing factor. The doses of both intrathecal and intravenous methotrexate and those of radiation therapy were comparable to those reported elsewhere. 5,6,9,11,12,25 28 Younger age at diagnosis was believed to be one predisposing factor for WMC in two previous studies, 8,27 but it was not confirmed in the current study.

6 Brain MRI after Treatment for ALL/Harila-Saari et al TABLE 2 Treatment-Related Abnormalities in Brain MRI in Patients with ALL at the End of Treatment and 5 Years Later Gender Age at diagnosis (yrs) Treatment protocol CRT MRI at the end of treatment MRI at the 5-year reevaluation VIQ PIQ Clinical remarks M 4.77 SR Moderate WMC bilaterally periventricularly in the frontoparietal region F 7.21 HR 24 Mild WMC bilaterally in the frontoparietal region, widened sulci M 5.71 HR 24 Mild WMC, old hemorrhage, widened sulci Mild WMC bilaterally periventricularly in the frontoparietal region Mild WMC bilaterally in the frontoparietal region, widened sulci No WMC, hemorrhage diminished, widened sulci Born prematurely. Attention, verbal and memory deficits. Motor clumsiness. Epilepsy Verbal, visuospatial, and memory deficits Attention deficit. F 3.52 SR Normal Mild WMC in the corpus callosum F 6.33 HR 24 Normal Mild WMC in the left trigonum M 2.53 IR 18 Old hemorrhage, widened Old hemorrhage ventricles diminished, new hemorrhage in the left capsula interna, widened ventricles F 2.14 HR 24 Widened ventricles Widened ventricles Delayed development before treatment. Verbal, memory, motor, and visuospatial deficits. Motor clumsiness. F 7.60 HR 24 Small atrophic area in the right frontal region of the brain parenchyma, widened sulci Small atrophic area in the brain parenchyma increased in size, widened sulci Memory deficit. MRI: magnetic resonance imaging; ALL: acute lymphoblastic leukemia; CRT: cranial radiation therapy; VIQ: verbal intelligence quotient; PIQ: performance intelligence quotient; M: male; F: female; SR: standard risk; WMC: white matter changes; HR: high risk; IR: intermediate risk. Abnormalities were more common in our irradiated patients, but the difference was not significant. Conversely, all the patients with cortical atrophy and hemorrhage or calcification belonged to the irradiated group. Atrophy of the brain is a known late finding after irradiation, and is believed to be related to a diffuse white matter injury, but other studies have suggested that the atrophic changes may be related to chemotherapy and cranial irradiation playing a lesser role. 9,26 Whether the brain atrophy in the irradiated patients results from a more serious disease, the more intensive chemotherapy treatment, or the cranial irradiation remains an open question. Microangiopathy is a known complication of cranial irradiation 2 and may be the predisposing factor for the hemorrhage and calcifications observed in the current study. Calcifications cannot be evaluated reliably or differentiated from hemorrhage by the current MRI technique used and therefore this study does not reveal their true incidence. The number of MRI abnormalities did not change significantly during the 5 years of follow-up, although all the patients with WMC at the end of treatment improved. Whether the new mild WMC in the two patients mentioned earlier has any clinical significance remains to be seen. The lesions were located in areas that are not common sites of treatment-related WMC (one in the trigonum and the other in the corpus callosum) and they were very small. Both patients had normal findings in the neuropsychologic evaluation. The findings of atrophy remained unchanged, but the cases of old hemorrhage or calcification improved somewhat, although one patient had developed a new sign of this kind. The results of the neuropsychologic assessment of patients with abnormal MRI findings did not differ

7 2614 CANCER December 15, 1998 / Volume 83 / Number 12 TABLE 3 Mean (SD) Intelligence Test Scores and Age at Diagnosis in the Patients with ALL 5 Years after Cessation of Therapy Subtest Patients with normal brain MRI (n 24) Patients with abnormal brain MRI (n 8) Patients with persistent WMC (n 2) Irradiated patients (n 17) Nonirradiated patients (n 15) Age at diagnosis (yrs) (SD) 5.4 (3.9) 5.0 (2.1) 6.0 (1.7) 6.0 (4.3) 4.6 (2.2) Total IQ (10.6) (16.7) 94.0 (2.8) (13.4) (10.8) Verbal IQ (10.2) (20.8) 84.5 (6.4) (14.1) a (11.8) a Information 10.0 (2.8) 9.7 (4.7) 6.0 (2.8) 9.0 (3.3) a 10.9 (3.0) a Similarities 12.8 (2.6) 11.8 (3.2) 9.0 (1.4) 12.0 (2.8) 13.1 (2.7) Arithmetic 9.8 (2.1) 8.8 (4.2) 6.0 (0.0) 9.4 (3.1) 9.8 (2.5) Vocabulary 11.4 (2.8) 10.4 (3.9) 8.0 (2.8) 10.1 (2.7) 12.1 (3.2) Comprehension 11.5 (3.3) 10.3 (3.2) 10.5 (0.7) 9.6 (2.6) b 13.0 (3.1) b Digit span 9.3 (2.4) 9.6 (4.5) 5.5 (0.7) 10.0 (3.7) 8.6 (1.6) Performance IQ (10.9) (10.2) (0.7) (10.1) (11.5) Picture completion 11.0 (2.8) 11.8 (2.5) 12.5 (2.2) 11.2 (2.1) 11.2 (3.4) Picture arrangement c 10.5 (1.5) 8.9 (2.5) 10.0 (2.8) 9.9 (1.5) 10.2 (2.2) Block design 12.3 (3.0) 12.8 (4.2) 10.5 (0.7) 12.1 (3.4) 12.8 (3.1) Object assembly c 11.7 (3.0) 12.6 (2.2) 11.5 (2.1) 11.8 (2.9) 12.0 (2.9) Coding 13.2 (3.5) 11.5 (4.1) 10.0 (5.7) 12.6 (3.1) 13.0 (4.4) SD: standard deviation; ALL: acute lymphoblastic leukemia; MRI: magnetic resonance imaging; WMC: white matter changes; IQ: intelligence quotient. a P b P c Results not available for the five patients tested with Wechsler Adult Intelligence Scale. TABLE 4 Impairment of Given Areas of Neuropsychologic Function in Survivors of ALL 5 Years after Cessation of Therapy No. and percentage (%) of patients with impairment Neuropsychologic function area Whole group (n 32) Patients with normal brain MRI (n 24) Patients with abnormal brain MRI (n 8) Irradiated patients (n 17) Nonirradiated patients (n 15) Attention 3 (9%) 1 (4%) 2 (25%) 2 (12%) 1 (7%) Language 9 (28%) 6 (25%) 3 (38%) 6 (35%) 3 (20%) Motor and sensory functions 6 (19%) 4 (17%) 2 (25%) 2 (12%) 4 (27%) Visuospatial functions 17 (53%) 15 (63%) 2 (25%) 9 (53%) 8 (53%) Memory 15 (47%) 11 (46%) 4 (50%) 8 (47%) 7 (47%) ALL: acute lymphoblastic leukemia; MRI magnetic resonance imaging. significantly from those in patients with normal MRI findings, but the two patients with persistent WMC had subnormal verbal IQ scores. It may be that the persistent WMC also may have clinical correlations whereas the transient mild changes in the white matter do not. Previous attempts to correlate MRI changes with neuropsychologic outcome in ALL patients have failed, 6,8 except in a report by Ciesielski et al., who used a neurodevelopmental approach and found a correlation between cerebellar hypoplasia and visuomotor coordination and memory deficits. 29 Intracranial calcifications on CT scans also have been correlated with lower IQ and memory deficits. 5,30 Ciesielski et al. suggested two reasons for the previous difficulties in finding a relation between structural changes and the profiles of neuropsychologic deficits: 1) variability in the localization of MRI deficits and difficulty in estimating the severity of the lesions and 2) a poor choice of neuropsychologic tools. The variability of the MRI changes also is one shortcoming of the current study when one is trying to find a relation between the changes and neuropsychologic deficits. However, our NEPSY test battery, used in addition to Wechsler s intelligence tests, should cover the different neuropsychologic function areas so that even specific cognitive dysfunctions could be found. Cranial irradiation 5,13,31-33 and young age at diagnosis 5,13,34 have been reported to be risk factors for neuropsychologic toxicity of leukemia treatment, although contradictory reports exist. 35 Cranial irradia-

8 Brain MRI after Treatment for ALL/Harila-Saari et al TABLE 5 MRI Studies of the Brain in Children with ALL after Treatment Authors No. of patients with ALL Treatment: CRT in addition to chemotherapy Time of evaluation Incidence of WMC Incidence of other changes Correlation with neuropsychologic evaluation Kramer et al Yes 12 months 12 years after therapy Biti et al Yes (on 22 patients) months after therapy Mulhern et al , age 2 years Yes (on 20 patients) Mean 7.3 years (SD 3.7 at diagnosis years) after therapy 0% Atrophy 10% No correlation 24% 30% Atrophy 0%, gray matter changes 5%, calcifications 15%, hemorrhage 5%, hemosiderin 10%, posterior fossa changes 10%, other abnormalities 10% Bakke et al No 3 8 years after therapy 53% Atrophy 13% Kingma et al , age 7 years at diagnosis Yes years after diagnosis 31% Atrophy 23%, calcifications or old hemorrhage 23%, cyst or tumor 6% Pääkkö et al Yes (on 25 patients) 2 20 years after therapy Laitt et al Yes 8 22 years after therapy Matsumoto et al Yes 1 month 10 years after diagnosis Seidel et al , age 5 years No 18 months 9.5 years at diagnosis after therapy Current study 32 Yes (on 17 patients) End of treatment and 5 years later 15% Atrophy 41%, meningioma 7%, old hemorrhage or calcification 30% 3% Atrophy 17%, brain tumors 9%, vasculopathy 6%, cystic infarcts 6%, vascular malformation 3% 18% 0% 0% 9% and 13% Atrophy 16%, calcifications or hemorrhage 6% No correlation with MRI. Calcifications on CT correlated with lower IQ No correlation Two patients with persistent WMC had deficient verbal IQ MRI: magnetic resonance imaging; ALL: acute lymphoblastic leukemia; CRT: cranial radiation therapy; WMC: white matter changes; SD: standard deviation; CT: computed tomography; IQ: intelligence quotient. tion also was related to a poorer verbal performance score in our series, but no difference was found between the age groups. However, the statistical power was too low to find small differences between the groups because of the small number of patients throughout. The higher-than-average mean IQ scores in our patients is explained by the fact that the old version of Wechsler s intelligence test (WISC) was used in this study, although the version used should not make any difference in the comparison between the groups. The typical impaired neuropsychologic function areas were visuospatial and memory, occuring in 50% of the patients and equally often in both treatment groups. This supports a previous report suggesting that another factor in addition to cranial irradiation may be responsible for impairments of this type. 36 Our results indicate that a minority of the patients treated for childhood ALL have treatment-related brain MRI abnormalities, whereas the majority of the patients had specific neuropsychologic deficits. Conventional MRI of the brain does not appear to be sensitive enough to detect the neurotoxic effects of leukemia treatment, and it may not be specific enough to detect a correlation between the various abnormalities and neuropsychologic assessment. New MRI techniques such as MRI spectroscopy or functional MRI may be more sensitive and specific methods in this respect. Conventional MRI of the brain does not appear to be very informative as a routine follow-up method in clinical practice, although it is needed as a screening method in the late follow-up of irradiated patients because of their increased risk for brain tumors. 9,26 The treatment given to these patients for ALL entailed a fairly low neurotoxicity. Specific neuropsychologic deficits were observed in the majority of the patients, but all of the patients still were able to attend

9 2616 CANCER December 15, 1998 / Volume 83 / Number 12 a normal school. Our goal has to be a form of treatment that avoids neuropsychologic sequelae, and we have to strive for the prevention of such deficits. However, when one considers that we are treating a lifethreatening disease, the cost of cure in these patients appears to be reasonable. REFERENCES 1. Ochs JJ. Neurotoxicity due to central nervous system therapy for childhood leukemia. Am J Pediatr Hematol Oncol 1989;11: Bleyer WA. Neurologic sequelae of methotrexate and ionizing radiation: a new classification. Cancer Treat Rep 1981; 65(Suppl 1): Ochs JJ, Parvey LS, Whitaker JN, Bowman WP, Ch ien L, Campbell M, et al. Serial cranial computed-tomography scans in children with leukemia given two different forms of central nervous system therapy. J Clin Oncol 1983;1: Vainionpää L, Laitinen J, Lanning M. Cranial computed tomographic findings in children with newly diagnosed acute lymphoblastic leukemia: a prospective follow-up study during treatment. Med Pediatr Oncol 1992;20: Mulhern RK, Kovnar E, Langston J, Carter M, Fairclough D, Leigh L, et al. Long-term survivors of leukemia treated in infancy: factors associated with neuropsychologic status. J Clin Oncol 1992;10: Kingma A, Mooyaart EL, Kamps WA, Nieuwenhuizen P, Wilmink JT. Magnetic resonance imaging of the brain and neuropsychological evaluation in children treated for acute lymphoblastic leukemia at a young age. Am J Pediatr Hematol Oncol 1993;15: Packer RJ, Zimmerman RA, Bilaniuk LT. Magnetic resonance imaging in the evaluation of treatment-related central nervous system damage. Cancer 1986;58: Wilson DA, Nitschke R, Bowman ME, Chaffin MJ, Sexauer CL, Prince JR. Transient white matter changes on MR images in children undergoing chemotherapy for acute lymphocytic leukemia: correlation with neuropsychologic deficiencies. Radiology 1991;180: Pääkkö E, Talvensaari K, Pyhtinen J, Lanning M. Late cranial MRI after cranial irradiation in survivors of childhood cancer. Neuroradiology 1994;36: Pääkkö E, Vainionpää L, Lanning M, Laitinen J, Pyhtinen J. White matter changes in children treated for acute lymphoblastic leukemia. Cancer 1992;70: Kramer JH, Norman D, Brant-Zawadzki M, Ablin A, Moore IM. Absence of white matter changes on magnetic resonance imaging in children treated with CNS prophylaxis therapy for leukemia. Cancer 1988;61: Biti GP, Magrini SM, Villari N, Caramella D, Guazzelli G, Rosi A, et al. Brain damage after treatment for acute lymphoblastic leukemia. A report on 34 patients with special regard to MRI findings. Acta Oncol 1989;28: Meadows AT, Gordon J, Massari DJ, Littman P, Fergusson J, Moss K.. Declines in IQ scores and cognitive dysfunctions in children with acute lymphocytic leukaemia treated with cranial irradiation. Lancet 1981;2: Rubenstein CL, Varni JW, Katz ER. Cognitive functioning in long-term survivors of childhood leukemia: a prospective analysis. J Dev Behav Pediatr 1990;11: Gustafsson G, Garwicz S, Hertz H, Johanesson G, Jonmundsson G, Moe PJ, et al. A population-based study of childhood acute lymphoblastic leukemia diagnosed from July 1981 through June 1985 in the five Nordic countries. Incidence, patient characteristics and treatment results. Acta Paediatr Scand 1987;76: Gustafsson G, Berglund G, Garwicz S, Hertz H, Jonmundsson G, Moe PJ, et al. A population-based study of children with standard risk acute lymphoblastic leukemia in the five Nordic countries. A follow-up of 230 patients. Acta Paediatr Scand 1989;78: Riehm H, Reiter A, Schrappe M, Berthold F, Dopfer R, Gerein V, et al. Corticosteroid-dependent reduction of leukocyte count in blood as a prognostic factor in acute lymphoblastic leukemia in childhood (therapy study ALL-BFM 83) [ in German]. Klin Padiatr 1987;199: Vainionpää L. Clinical neurological findings of children with acute lymphoblastic leukaemia at diagnosis and during treatment. Eur J Pediatr 1993;152: Harila-Saari AH, Ahonen AK, Vainionpää LK, Pääkko EL, Pyhtinen J, et al. Brain perfusion after treatment of childhood acute lymphoblastic leukemia. J Nucl Med 1997;38: Wechsler D. Wechsler Intelligence Scale for Children. Käsikirja [manual]. [Edited and translated by Psykologien kustannus]. Helsinki: Psykologien kustannus, Wechsler D. WAIS. Wechslerin aikuisten älykkyysasteikko. [Manual for the Wechsler adult intelligence scale.] [Edited and translated by Psykologien kustannus]. Helsinki: Psykologien kustannus, Korkman M. NEPSY. A proposed neuropsychological test battery for young, developmentally disabled children [published doctoral dissertation]. Helsinki: University of Helsinki, Beery KE. Visual-motor integration. Cleveland, OH: Modern Curriculum, Korkman M. NEPSU. Lasten neuropsykologinen tutkimus. Uudistettu versio. [NEPSY. Neuropsychological Assessment of Children. Revised version.] Helsinki: Psykologien kustannus, Bakke SJ, Fossen A, Storm-Mathiesen I, Lie SO. Long-term cerebral effects of CNS chemotherapy in children with acute lymphoblastic leukemia. Pediatr Hematol Oncol 1993;10: Laitt RD, Chambers EJ, Goddard PR, Wakeley CJ, Duncan AW, Foreman NK. Magnetic resonance imaging and magnetic resonance angiography in long term survivors of acute lymphoblastic leukemia treated with cranial irradiation. Cancer 1995;76: Matsumoto K, Takahashi S, Sato A, Imaizumi M, Higano S, Sakamoto K, et al. Leukoencephalopathy in childhood hematopoietic neoplasm caused by moderate-dose methotrexate and prophylactic cranial radiotherapy an MR analysis. Int J Radiat Oncol Biol Phys 1995;32: Seidel H, Nygaard R, Haave I, Moe PJ. Magnetic resonance imaging and neurological evaluation after treatment with high-dose methotrexate for acute lymphocytic leukaemia in young children. Acta Paediatr 1996;85: Ciesielski KT, Yanofsky R, Ludwig RN, Hill DE, Hart BL, Astur RS, et al. Hypoplasia of the cerebellar vermis and cognitive deficits in survivors of childhood leukemia. Arch Neurol 1994;51: Brouwers P, Riccardi R, Fedio P, Poplack DG. Long-term neuropsychologic sequelae of childhood leukemia: correlation with CT brain scan abnormalities. J Pediatr 1985;106:

10 Brain MRI after Treatment for ALL/Harila-Saari et al Moss HA, Nannis ED, Poplack DG. The effects of prophylactic treatment of the central nervous system on the intellectual functioning of children with acute lymphocytic leukemia. Am J Med 1981;71: Rowland JH, Glidewell OJ, Sibley RF, Holland JC, Tull R, Berman A, et al. Effects of different forms of central nervous system prophylaxis on neuropsychologic function in childhood leukemia. J Clin Oncol 1984;2: Copeland DR, Fletcher JM, Pfefferbaum-Levine B, Jaffe N, Ried H, Maor M, et al. Neuropsychological sequelae of childhood cancer in long-term survivors. Pediatrics 1985;75: Jannoun L, Chessells JM. Long-term psychological effects of childhood leukemia and its treatment. Pediatr Hematol Oncol 1987;4: Mulhern RK, Fairclough D, Ochs J. A prospective comparison of neuropsychologic performance of children surviving leukemia who received 18-Gy, 24-Gy, or no cranial irradiation [published erratum appears in J Clin Oncol 1991;9(10): 1922]. J Clin Oncol 1991;9: Waber DP, Tarbell NJ, Fairclough D, Atmore K, Castro R, Isquith P, et al. Cognitive sequelae of treatment in childhood acute lymphoblastic leukemia: cranial radiation requires an accomplice. J Clin Oncol 1995;13: