Neurodevelopmental outcome following prenatal diagnosis of an isolated anomaly of the corpus callosum.

OBJECTIVES: To assess the ability of prenatal ultrasound and magnetic resonance imaging (MRI) to diagnose isolated anomalies of the corpus callosum (ACC) and to further document the long-term prognosis following diagnosis. METHODS: This was a prospective case-control study carried out between 1999 and 2004. Diagnosis was made by a combination of ultrasound and MRI. All infants were examined by a neuropediatrician and parents consented to answer questionnaires (CDI, Ireton's Child Developmental Inventory) in 22 cases, which were matched with 44 control infants. The CDI was used to assess neurodevelopmental outcome in cases and controls. Mean DQ-CDI (development quotient calculated from CDI) values and frequencies of abnormal results were compared between groups, and a meta-analysis of previous studies was performed. RESULTS: The diagnosis of ACC was made prenatally and confirmed postnatally in 175 cases. The diagnosis was thought to be isolated ACC in 88/175 (50%) cases. Sixty of these 88 cases (68%) underwent termination of pregnancy and one died in utero. Twenty-seven were liveborn, of which 26 were followed up for a median of 50 (range, 30-74) months. Additional anomalies were diagnosed postnatally in four (15%) of these 26 neonates. The control group was significantly better (P < 0.05) compared with the cases diagnosed prenatally with isolated ACC with respect to gross motor, fine motor, language comprehension, numbers and general development, and it was marginally better for letters (P = 0.066). Seven of 26 (27%) (95% CI, 13-46%) infants with ACC over the age of 30 months had neurodevelopmental delay, compared with only one case with borderline developmental delay among the 44 controls (P = 0.006). CONCLUSION: Prenatal diagnosis of ACC by a combination of ultrasound and MRI is reliable. However, the isolated nature of the anomaly could only be assessed in 85% of our cases. Since counseling is provided at the time of prenatal diagnosis, our population of isolated ACC included the cases that were missed prenatally as being ACC with associated anomalies. A meta-analysis of nine studies suggests that the development of children diagnosed prenatally with isolated ACC is normal in up to 70% (CI 95%, 56-83%) of cases. This means that the prospective risk of neurodevelopmental delay for a fetus with ACC described as isolated prenatally is 27%, compared with 15% for an infant whose diagnosis of isolated ACC is confirmed postnatally.

Ultrasonography of the fetal brainstem: a biometric and anatomical, multioperator, cross-sectional study of 913 fetuses of 21-36 weeks of gestation.

OBJECTIVES: To establish the ultrasonographic fetal growth charts of the pons and the vermis/pons ratio on a multioperator basis in low-risk pregnancies and provide a detailed description of the anatomical and ultrasonographic criteria of normal brainstem growth. METHODS: A prospective, multicenter, multioperator, ultrasonographic study was conducted on 913 fetuses aged 21-36 weeks. The anteroposterior diameter of the pons and the greatest vermal height were measured to establish a growth chart, using a mid-sagittal plane with a posterior transfontanellar approach. The LMS semiparametric statistical method was used to construct the growth charts. Three morphological structures were also examined: the pons arch and its echostructure, the bulbo-protuberential sulcus and the primary vermal fissure. RESULTS: The anteroposterior diameter of the pons and the greatest vermal height were measured in 96.7% of cases. The anteroposterior diameter of the pons and vermis increased linearly with gestational age. The vermis/pons ratio was stable during pregnancy. CONCLUSION: We have drawn the growth charts for the pons and vermis during pregnancy and described the normal ultrasound morphology of the brainstem. Knowledge of these morphological and biometric data could facilitate early screening for pontocerebellar hypoplasia.

Intellectual status of working-class children adopted early into upper-middle-class families.

Failure rates observed (13 +/- 6 percent for school failures, 17 +/- 5 percent for scores below 95 on a collective IQ test) were far below those expected from the social class of birth (55 percent, 51 percent) or observed in a control group (56 +/- 8 percent, 49 +/- 9 percent) but close to those expected from the social class of adoption (15 percent, 15 percent).

[Ultrasonography of fetal esophagus: healthy appearance and prenatal diagnosis of a case of esophagus atresia with esotracheal fistula]. / Echographie de l'oesophage foetal: aspect physiologique et application au dépistage anténatal d'une atrésie oesophagienne.

The thoracic part of a fetal esophagus is generally overlooked by usual prenatal ultrasonography. However, screening it might improve the detection rate of esophageal malformations for which prenatal diagnosis remains far from accurate. In this article, we describe the technique which makes it possible to get a precise image of a fetal thoracic esophagus in its more sensitive part: between the trachea and the aorta. After describing the appearance of a healthy thoracic esophagus, we will show how this technique can be used for prenatal detection of esophagus malformations. For this purpose, we provide the case report of a prenatal diagnosis of esophagus atresia with esotracheal fistula.

[Chart for estimation of fetal weight 2014 by the French College of Fetal Sonography (CFEF)]. / Courbe d'estimation de poids fœtal 2014 par le Collège français d'échographie fœtale (CFEF).

OBJECTIVE: To establish a reference chart for estimated fetal weight (EFW) using the Hadlock formula based on recent biometric data (2012-2013). MATERIAL AND METHODS: A prospective multicentric longitudinal study was carried out. Biometric parameters as the head circumference (HC), abdominal circumference (AC) and the femur length were measured in multiple areas of France from January 2012 until December 2013. EFW was calculated using the predictive formula of Hadlock using three parameters. The accurate gestational age was the main inclusion criteria calculated in weeks of gestation (WG). A polynomial regression approach was used to calculate the mean and standard deviation for every WG adjusted to raw data. Centiles of EFW were calculated from the z score that corresponds to the -1.88, -1.28, 0, +1.28, +1.88 respectively for the 3rd, 10th, 50th, 90th, et 97th percentile in order to establish a new chart of EFW. RESULTS: Measurements were obtained for 33,143 fetus between 17 et 38 WG. Reference charts with the 3rd, 10th, 50th, 90th et 97th percentiles were presented. CONCLUSION: The reference Chart 2014 is an in utero chart for EFW based on ultrasound measurements data reliable and homogenous from a sample of 33,143 fetus of a general population. It offers a tool to use in routine ultrasound examination for the survey of the fetal growth and to diagnose fetus that are small for gestational age or presenting a restriction in growth.

Predicting late-onset growth abnormalities using growth velocity between trimesters.

OBJECTIVES: To determine whether growth velocity parameters derived from routine prenatal ultrasound measurements at first, second and third trimester can identify normal growth at term as well as late-onset growth abnormalities. MATERIAL AND METHODS: Longitudinal study of fetal growth in normal singleton pregnancies with three normal ultrasound examinations and delivered at term. Fetuses were classified into 3 groups (<10th percentile, 10-90th percentile, >90th percentile) based on birth weight. Multiple regression on birth weight classification was used to build up a prediction equation of fetal growth potential (FGP) based on fetal biometry and fetal growth velocity parameters between ultrasound examinations. Best cut-off value for FGP predicting growth restriction and macrosomia were defined. RESULTS: 356 pregnancies were included. Fetal biometry growth velocities between examinations were calculated for all measurements. Using best cut-off values, the estimated sensitivity, specificity and odds ratio were: 60% [44;74], 91% [89;92] and 14.55 [6.30;33.98] and 53% [36;69], 89% [88;91] and 10 [4.27;23.49] for the prediction of growth restriction and macrosomia, respectively. DISCUSSION: Fetal growth potential can be derived and calculated from standard ultrasound measurements. It can improve identification of these fetuses at risk for late-onset growth abnormalities and their related morbidity.

[Linear insertion of the atrioventricular valves without defect]. / Insertion linéaire des valves atrioventriculaires sans défect. A propos d'une nouvelle série anatomique de 213 coeurs de foetus trisomiques 21.

On a first anatomical series of 52 hearts of trisomic 21 fetuses, published in June 2002, we described a new minor cardiac anomaly, belonging to the atrioventricular septal defect, with a linear insertion of the atrioventricular valves without defect. We want to confirm these data, on a larger series of 213 new hearts of trisomic 21 fetuses by adding a complementary section to the standard examination; 100% of controls have shown a normal insertion with an offsetting of the atrioventricular valves. On 113 out of these 213 hearts of trisomic 21 fetuses, with a so called "normal" heart at the standard examination showing no defect, the complementary section has shown that only 37.2% of these hearts have a normal insertion, whereas 62.83% show a linear insertion, without offsetting and without any septal defect. This linear insertion has been observed in all the different types of atrioventricular septal defect as a good hallmark for trisomy 21; but, since then, they have always been described associated with a septal defect, atrial or ventricular. Our hypothesis is that the linear insertion of the atrioventricular valves without defect is the minor form of the atrioventricular septal defect spectrum, taking place between the prior described partial types of atrioventricular septal defect, in which there is always a defect (ostium primum type atrial septal defect or inflow type ventricular septal defect), and the real normal heart. A precise description of the level of the complementary section and of the anatomic peculiarities of the linear insertion of the atrioventricular valves without defect would help its screening in fetal ultrasonography.

Is a dot-filling group test a good tool for assessing manual performance in children?

A dot-filling test was used to assess hand performance in two populations of French children. In the first experiment, the effect of sex, age and handedness on hand performance was examined (1742 children). Age was related to the degree of laterality, but girls were less lateralized than boys. The second experiment (200 children) showed that the surprising sex differences cannot be explained by the group testing procedure. Underlying processes leading to performance in the dot-filling task are probably different from those involved in other tests generally used for children.

Hemispheric asymmetry and corpus callosum morphometry: a magnetic resonance imaging study.

Previous post-mortem studies (Aboitiz, F., Scheibel, A.B., Fisher, R.S., Zaidel, E., 1992. Brain Res. 598, 154-161 and Aboitiz, F., Scheibel, A.B., Zaidel, E., 1992. Brain 115, 1521-1541) have shown an inverse association between asymmetry in perisylvian areas and the size of a specific segment, the isthmus, of the corpus callosum (CC) in males. The purpose of this work was to study in vivo the association between hemispheric asymmetry and the total size of the CC in 35 right-handed subjects (16 males, 19 females; mean age 24.9 +/- 3.9). An MRI scan was performed for each subject. The area of the right (RH) and left (LH) hemispheres were measured from images in the sagittal plane and the area of the CC from images in the mid-sagittal plane. The index of hemispheric asymmetry was absolute value((LH - RH)/[(LH + RH)/2]). There was a negative correlation between the absolute value of hemispheric asymmetry and the size of the CC in males (r = -0.55, P = 0.03) but not in females (r = -0.20, P = 0.42). These findings, like those of Aboitiz et al. (Aboitiz, F., Scheibel, A.B., Zaidel, E., 1992. Brain 115, 1521-1541), suggest a sex-dependent decrease in interhemispheric connectivity with increasing hemispheric asymmetry.

Are the brains of monozygotic twins similar? A three-dimensional MR study.

PURPOSE: The role of genetic mechanisms and the influence of environmental events in human brain development have been difficult to evaluate. The purpose of this study was to compare the cerebral cortical morphology and midline structures of monozygotic twin pairs using MR imaging. METHODS: Six observers, blinded to twin pairings, evaluated the 3-D renderings of the cortical surface and midline structures from MR images of seven monozygotic twin pairs. A morphometric analysis of the corpus callosum and of the distance between the anterior and posterior commissures was also performed. RESULTS: Despite surprising anatomic differences, the brains of the twin pairs were similar enough to enable the observers to distinguish twin pairs from unrelated subjects. Five of six observers correctly identified the brains of all seven twin pairs; the remaining observer failed to make a correct match in only one of seven pairs. Three of six observers identified the midline sagittal images of the related twins in all seven pairs, and the other three identified the related midline sagittal images in five of seven pairs. The results were statistically significant. CONCLUSION: Although the observed differences in morphologic characteristics between twins necessarily reflect nongenetic influences, the cortical patterns and midline structures of monozygotic twins probably are genetically similar.

MR determination of hippocampal volume: comparison of three methods.

PURPOSE: To determine whether measurements of the volume of the hippocampal formation obtained from a three-dimensional acquisition not perpendicular to the hippocampus are statistically different from those obtained from a perpendicular acquisition. METHODS: Both hippocampi were studied in 10 healthy volunteers with two three-dimensional acquisitions, allowing three different volume-calculation protocols: (a) on sections from a coronal 3-D acquisition not perpendicular to the axis of the hippocampal formation (NOPERP protocol), (b) on sections obtained with the same acquisition but reformatted perpendicular to the axis of the hippocampal formation (REFOR protocol), and (c) on sections from a coronal 3-D acquisition perpendicular to the axis of the hippocampal formation (PERP protocol) obtained with the patient's head tilted backward. To obtain measurements of the volume of the hippocampal formations, an accurate 3-D processing technique was used to segment the hippocampus. In all subjects, two hippocampal formation right-left asymmetry indexes were calculated by using each of the three protocols. RESULTS: For the right hippocampus, the mean volume was 3.42 cm3 (NOPERP protocol), 4.18 cm3 (REFOR protocol), and 3.91 cm3 (PERP protocol). For the left hippocampus, the mean volume was 3.29 cm3 (NOPERP protocol), 4.02 cm3 (REFOR protocol), and 3.74 cm3 (PERP protocol). For both hippocampi, the differences of the mean volumes were significant between each protocol. However, for both hippocampi, a high correlation was observed between volumes obtained with the different protocols. For the two asymmetry indexes, there were no significant differences for the means obtained with the three protocols. CONCLUSION: With the use of 3-D acquisitions in the study of hippocampal formation biometry, different procedures lead to significant variations in the absolute values of the volume of the hippocampal formation. However, there is a strong correlation between the results obtained by each method.

Extent of tumor-brain interface: a new tool to predict evolution of malignant gliomas.

OBJECT: Tumor size is one of the features commonly used in oncology to predict disease evolution. However, for most primary brain tumors it is not predictive of outcome. Taking advantage of a gene therapy trial in which recurrences of glioblastoma were targeted with suicide genes, the authors developed a new parameter: the extent of tumor-brain interface--also called surface of tumor volume (STV)--to better describe three-dimensional conformation and the relationship between tumors and the surrounding normal tissue. Correlations between the STV and the usual clinical parameters were analyzed. METHODS: Between 1995 and 1998, 16 patients presenting with recurrent glioblastomas were enrolled in this study. Preoperative magnetic resonance images were analyzed on a separate workstation; the interface between tumor and normal brain tissue was measured on each 3-mm-thick section to assess STV. The mean STV was 29.2 cm2, and the mean tumor volume (TV) was 23.8 cm3. The STV was significantly correlated with survival (Spearman test: r = -0.54, p = 0.03), but TV was not (Spearman test: r = -0.39, p = 0.15). A separate analysis of responding and nonresponding patients showed that, as expected, STV was negatively correlated with survival among nonresponding patients (p = 0.04), but that among responding patients there was a positive tendency between STV and survival. CONCLUSIONS: These findings indicate that STV may be a useful tool for predicting the evolution of malignant glioma. Moreover, in future gene therapy trials in which such in situ approaches are used, increasing density and improved distribution of transfer cells should be taken into consideration as an important issue for efficacy.

Perinatal mortality in twin pregnancies: a 3-year analysis in Seine Saint-Denis (France).

OBJECTIVES: Our purpose was to evaluate perinatal mortality in twins and to investigate factors associated with this mortality. STUDY DESIGN: A prospective study on the perinatal mortality was performed in the department of Seine-Saint-Denis. Out of all the perinatal deaths, we have retrospectively isolated those arising from twin pregnancies. RESULTS: There were 54 twin pregnancies complicated with perinatal death. The perinatal mortality rate in twin pregnancy was 78.0 per 1000 twin babies delivered. Out of 86 twin deaths, 38 (44.2%) were born before 28 weeks gestation and out of 82 twin perinatal deaths, 37 (45.1%) weighed less than 1000 g. Chorionicity was recorded in 44 twin pairs: 21 (47.7%) were dichorionic and 23 (52.3%) monochorionic. Finally, out of 48 twin sets there were four (8.3%) monoamniotic pregnancies. CONCLUSIONS: The present data show that extreme prematurity represents nearly half of perinatal mortality in twins. This study indicates also a significant proportion of monochorionic placentation among twin pregnancies with poor outcome.

Fetal biometric parameters, twin type and birth weight difference. A longitudinal study.

OBJECTIVE: To examine the relationship between twin type, divided by zygosity, chorionicity, and birth weight difference [more or less than 15% intertwin weight difference at delivery] on fetal biometric measurements, including biparietal diameter, transverse abdominal diameter, head and abdominal circumferences (AC) and head and abdominal areas, femur length (FL), transverse cerebellar diameter, cerebellar circumference and cerebellar area at 18, 23, 28 and 32 weeks of gestational age. STUDY DESIGN: A prospective and longitudinal study was done on 75 twin pairs that did not present the twin-twin transfusion syndrome (150 fetuses) divided into birth weight difference group I (<15%) and group II (>/=15%). The twin pairs were divided into three groups: dizygotic, monochorionic, and dichorionic-monozygotic. In each group the differences in various ultrasound indices were evaluated in reference to discordant growth or concordant growth. RESULTS: In our sample, there were no monozygotic twin pairs with dichorionic placentation with a birth weight difference of more than 15%. Significant associations of group and gestational age were found in dizygotic and monochorionic twins. Intrapair differences were significantly higher in group II (>/=15%) than in group I (<15%) for all fetal parameters studied except for AC, abdominal area and all cerebellar parameters. The value of these discrepancies increased according to gestational age except for FL. Significant interactions between group and zygosity type indicated that intrapair FL differences were more associated with group in monochorionic twins, whereas intrapair transverse cerebellar diameter differences were more associated with group in dizygotic twins. CONCLUSION: Our data show that most fetal biometric parameters are associated with birth weight discordancy. Abdominal area could be a relevant marker for twins with obstetric complications. Note that this is the first research that has studied a twin sample divided by both twin type and birth weight group.

Periventricular leukomalacia and mode of delivery in twins under 1500 g.

BACKGROUND: The optimal mode of delivery in twin gestations remains undefined, particularly for twins weighing less than 1500 g. OBJECTIVE: To evaluate the impact of the mode of delivery on neonatal outcome in twins below 1500 g. MATERIALS AND METHODS: In this multicenter cohort study during 1999, 66 sets of twins born in hospital and weighing below 1500 g formed our study group. Antenatal and neonatal parameters and their relationship to mode of delivery were studied, based on a factor analysis. Analysis of covariance was used to assess the effect of the mode of delivery on postnatal factors, with antenatal parameters used as covariates. RESULTS: Statistical analysis showed that infants delivered vaginally had significantly more periventricular leukomalacia than those children delivered by Cesarean section (p = 0.03). The estimated odds for leukomalacia were higher in the vaginal than in the Cesarean group when adjusted for covariates (OR = 4.7; 95% CI = 1.0, 25.15). CONCLUSION: Routine Cesarean section should be recommended in twin gestations with infants weighing less than 1500 g, regardless of gestational age or fetal presentation.

Biparietal diameter in twins at gestational weeks 18-32. Differences and similarities. Le Groupe Romulus.

OBJECTIVE: To examine twin similarities on biparietal diameter (BPD) measurements by zygosity (monozygotic [MZ] and dizygotic [DZ]) and chorionicity (monochorionic [MC] and dichorionic [DC]) and their evolution during pregnancy. METHOD: A sample of 54 pairs of twins (43 DZ sets, 11 MZ sets [7 MC and 4 DC]) was constructed using retrospective data. Despite the small sample size, our data were complete, and, for the first time we measured different fetal parameters on digital ultrasound images outside routine examination. The intraexaminer and interexaminer reliability of BPD measurement was significant (r = .95, P < .001). RESULTS: In this study, developmental results indicate significant linear regression coefficient (R) through the whole period of gestation (r = .96, P < .001), though product moment correlations comparing the periods of gestation two by two are weaker. The distribution of BPD values was slightly wider at the 28th week and markedly wider at the 32nd week than that at the 18th and 23rd weeks. The intraclass correlations of DZ and MZ (MC and DC) twins were examined at the 18th, 23rd, 28th and 32nd weeks of amenorhea. The intraclass correlations of DZ twins were significant through the whole period of gestation (r = .45, P = .001; r = .27, P = .04; r = .36, P = .008; and r = .42, P = .002, respectively), whereas the intraclass correlations of MZ twins were significant only at the 18th, 23rd and 28th weeks (r = .73, P = .002; r = .69, P = .005; r = .49, P = .047, respectively). We found significant within-variance differences not only between DZ and DC-MZ but also between DC-MZ and MC-MZ in late gestation. CONCLUSION: Our analysis of twin BPD development demonstrated that zygosity and chorionicity type are both important determinants of twin fetal development.

Magnetic resonance imaging and cerebral volumic evaluation: comparison of four post-processing techniques.

Four different MRI techniques of volumetric evaluation were applied in 10 healthy females and compared. The results have shown that not all techniques were interchangeable, but that the results of two groups of methods are not statistically different.

[Biometry 2000. Fetal growth charts by the French College of fetal ultrasonography and the Inserm U 155]. / Biométrie 2000. Tables de croissance foetale par le Collège français d'échographie foetale (CFEF) et l'Inserm U 155.

OBJECTIVE: The aim of Biometry 2000 was to establish reliable sonographic fetal auxology reference charts for biparietal diameter, head circumference, abdominal diameter, abdominal circumference and femur length. DESIGN: A prospective multicentric longitudinal and transversal study was carried out. SUBJECTS: The data of this study were obtained from 35,456 scans recorded in 10,043 women for a period of one year. METHODS: Sonographic measurements were performed using a single method for each parameter. Twins and unknown gestational ages were not included. No exclusions were made for small-for-age, congenital malformations or any event during pregnancy. Assuming measurements have a normal distribution, centiles were calculated after polynomial regression models from mean and standard deviation according to gestational age in exact weeks from last menstrual period. RESULTS: A total of 32,584 fetal biparietal diameters, 20,400 direct cephalic circumferences, 31,790 transverse abdominal diameters, 21,562 direct abdominal circumferences and 30,960 femur lengths were obtained. Fetal growth charts with means, standard deviations and centiles are presented. CONCLUSION: The multicentric study, Biometry 2000, has established sonographic fetal charts derived from a large sample of scans and unselected pregnant women assuming a normal distribution.

[Prevention of prematurity in 842 consecutive twin pregnancies]. / Prévention de la prématurité au cours du suivi de 842 grossesses gémellaires consécutives.

OBJECTIVE: To evaluate the effects of a specific management in prevention of prematurity in twin pregnancies. METHOD: The parameters involved in obstetrical follow-up of twin pregnancies were evaluated in a retrospective study of 842 twin pregnancies between 1979-1992. RESULTS: Five-hundred-and-fifteen pregnancies were spontaneous. Two-hundred followed treatment with ovulation induction agents and 127 were due to in vitro fertilization. Early diagnosis of twin pregnancies allowed preventive measures against prematurity. Management of twin pregnancies included rest at home and regular clinical examination of the cervix. Monthly consultations and monthly ultrasounds were performed at hospital. Weekly consultations were made by midwives at home. Hospitalization was not systematic, but was necessary in cases of maternal complications. The diagnosis of twin pregnancy was made before 16 weeks in 82.4% of the cases. The mean gestational age at delivery was 36.2 weeks. The prematurity rate was 45.96%. The mean weight of the neonates was 2,376 +/- 533 g for the first twin and 2,297 +/- 547 g for the second twin. The mean Apgar score at 5 minutes was 9.7 and 9.3, respectively for the first and the second twin. The perinatal mortality was 39.3 per 1,000. The main neonatal complications resulted from prematurity. Twins were hospitalized in the intensive care unit in 20.7 cases. CONCLUSION: The present study supports early diagnosis of multiple pregnancies with systematic ultrasound at 11-13 weeks for each pregnancy, information of the patients, rest at home and regular clinical examination of the cervix.

[First-line radiosurgery of brain metastases. Interest and results of reduced dose]. / Irradiation stéréotaxique de première intention des métastases cérébrales. Intérêt et résultats d'une réduction de la dose délivrée.

BACKGROUND: The minimal radiosurgical dose required to control cerebral metastases remains unknown. The aim of this study was to test whether a lower peripheral dose than usually delivered could effectively control these lesions or not. PATIENTS AND METHODS: One hundred and eighty patients presenting 356 lesions were give first-line radiosurgery between 1995 and 2001 in Pitié-Salpêtrière hospital using a 10 MV LINAC. Mean age was 59 years, sex-ratio was 1.65, mean KI was 70. The lung was the most frequent primary site (n=85), followed by melanoma (n=29), kidney (n=21), digestive tract (n=14), breast (n=11), and others (n=20). Seventy-six percent of the patients presented 1 or 2 lesions. Mean tumor Volume was 5.5 cm3. Mean peripheral dose was 14.8Gy, mean isocenter dose was 21.6Gy. RESULTS: Median survival was 7.6 months, local control rate was 90% at 6 months, 76% at 1 Year and 70% at 2 years. Median "neurological disease free" survival was 15 months. Multivariate analysis demonstrated the influence of two parameters on survival: number of lesions (p=0.001) and KI (p=0.04). The only parameter significantly correlated with disease-free survival was the number of isocenters (p=0.005). Morbidity (grade 2 RTOG) was 7.2% with no perimortality. CONCLUSIONS: Low peripheral doses delivered by radiosurgery may control brain metastases with the same efficacy and fewer side-effects as the doses usually reported in the literature.