High Resolution Sonographic Determination of the Normal Dimensions of the Intracranial Extraaxial Compartment in the Newborn Infant

Transcription

1 High Resolution Sonographic Determination of the Normal Dimensions of the Intracranial Extraaxial Compartment in the Newborn Infant Daniel A. Frankel, MD, David P. Fessell, MD, Wayne P. Wolfson, MD Prominence of the extraaxial space occasionally is encountered in infants referred for ultrasonography to exclude hydrocephalus. The interpretation of this finding can be problematic. We examined the width of the extraaxial compartment in 82 normal newborn infants. Scanning technique was optimized for viewing the near field, and the extraaxial space was measured over the cerebral convexities. Correlation was made with demographic variables. Measurements varied from 0 to 3.3 mm (mean, 1.6 mm), with slight negative linear relationship to gestational age. We conclude that small amounts of extraaxial fluid, up to 3.3 mm in width on scans, are common and normal in newborn infants. KEY WORDS: Brain, ultrasonography; Subarachnoid space; Extraaxial space; Infant, newborn. The standard clinical neonatal head ultrasonographic examination uses a 5.0 or 7.5 MHz transducer, or both, and the technique is optimized for evaluation of the deeper structures. The inherent focal zones of these transducers are ABBREVIATIONS CT, Computed tomography; SD, Standard deviation Received August 22, 1997, from the Department of Radiology, Henry Ford Hospital, Detroit, Michigan. Revised manuscript accepted for publication March 28, Address correspondence and reprint requests to Wayne P. Wolfson, MD, Department of Radiology, Henry Ford Hospital, 2799 West Grand Boulevard, Detroit, MI approximately 3 to 5 cm beneath the probe surface. Since much of the extraaxial space is located less than 1 cm from the skin surface, a routine scan may overlook even sizable collections of fluid in that area (Fig. 1). Experience with the normal appearance of the fluid-filled extraaxial space among many practicing radiologists is therefore relatively limited; furthermore, values for the normal dimensions of this space derived from studies of substantial numbers of newborn infants have not been published. When dedicated sonographic interrogation of the superficial extraaxial space is performed, neonates referred for evaluation of increased head circumference frequently are found to have small extraaxial fluid collections with otherwise normal scans. As stated previously, this finding often is interpreted with equivocation. When these collections are large, the differential diagnosis has traditionally included pathologic conditions requiring further diagnostic investigation, such as trauma or child abuse, blood 1998 by the American Institute of Ultrasound in Medicine J Ultrasound Med 17: , /98/$3.50

2 412 NEONATAL INTRACRANIAL EXTRAAXIAL COMPARTMENT J Ultrasound Med 17: , 1998 A B Figure 1 A, A routine coronal sonogram from a 5 month old infant referred for evaluation of increasing head circumference. The sonogram is optimized for evaluation of the cerebral ventricles and deeper structures; an extraaxial fluid collection over the frontal convexities is easily overlooked. B, The extraaxial fluid (calipers) is demonstrated more clearly when focal zones are adjusted for imaging the near field. dyscrasia, and hygroma. However, it is our premise that small collections of extraaxial fluid are common and normal, and by determining the range of normal values we hope to obviate additional imaging and the erroneous pursuit of suspected pathologic entities in many cases. This study also examines the relationship of extraaxial fluid measurements to sex, gestational age at birth, method of delivery, head circumference, and birth weight. MATERIALS AND METHODS Eighty-two newborn infants were examined prospectively (34 girls, 48 boys; age range, 0 to 2 days). Owing to logistical constraints, examinations were not performed on consecutive deliveries. The neonates were delivered without complication and without clinical suspicion of abnormality requiring more than routine postnatal care. Gestational ages ranged from 34.5 to 42 weeks (mean, 39.3 weeks); four newborn infants of 36 to 38 weeks gestation and two of less than 36 weeks were included. Clinical follow-up results ranging from 2 weeks to 6 months after discharge were available in 79 subjects. The study criteria mandated exclusion if any neurologic abnormality requiring additional or nonroutine care was detected on a subsequent well-child visit. The sonographic imaging technique is summarized as follows. Images were obtained in coronal and sagittal projections through the anterior fontanelle using a 10.0 MHz linear array or curved transducer (Acoustic Imaging 5200S scanner, Tempe, AZ). The neonates were scanned in the supine position, facing forward (anteroposterior position), by one of two radiologists (D.A.F., D.P.F.) experienced in cranial sonography; all images were then reviewed by a third radiologist (W.P.W.). Newborn infants are scanned most easily when sleeping or after feeding, and cine mode can be of great help in obtaining measurements in active patients. A generous amount of coupling gel was used to minimize pressure transmitted by the probe. The technique was optimized for viewing the near field by minimizing the depth of the image and adjusting the time-gain compensation pods appropriately. Reverberation artifact can be eliminated and near-field gain optimized by this adjustment; a standoff pad is not necessary. We defined the dimensions of extraaxial compartment using the craniocortical and sinocortical widths of Govaert and coworkers 1,2 (Fig. 2). These are, respectively, the distance from the inner table of the calvarium to the surface of the immediately subjacent cerebral cortical gyrus and the distance from the wall of the superior sagittal sinus to the cortical surface. Sinocortical and craniocortical widths were obtained bilaterally, and these four measurements were averaged to obtain a mean fluid width. Craniocortical measurements were generally made over the medial aspects of the convexities owing to field of view limitations posed by the acoustic window (the anterior fontanelle). The deeper structures also were examined to exclude any gross abnormality, such as hemorrhage or ventricular dilation. Owing to subjectivity in obtaining the interhemispheric width at differing depths and to the normal anatomic variability in the dimensions of the superior

3 J Ultrasound Med 17: , 1998 FRANKEL ET AL 413 A B Figure 2 Examination technique optimized for measurement of the extraaxial compartment in a normal newborn infant with 10 MHz probe. Coronal (A) and sagittal (B) sonograms demonstrate the craniocortical (x calipers) and sinocortical (+ calipers) widths. sagittal sinus, the width of the interhemispheric fissure was not considered a reproducible measurement and was not used in our study. It should be noted, however, that qualitative widening of the interhemispheric fissure may be apparent on a routine sonogram, and it may be the only indication that dedicated scanning of the superficial extraaxial space is indicated. Several scanning pitfalls should be avoided. Cortical sulci must be demonstrated to confirm that the subarachnoid space is being evaluated. Measurements should be made from the cranial surface to a cortical gyrus, not to a sulcus that may vary in dimension. The superior sagittal sinus should not be confused with fluid in the subarachnoid space. Again, visualization of sulci will eliminate this pitfall, and color or pulsed Doppler imaging can also be of value. 3 To determine the upper limit of normal, data were first evaluated for normality. As the distribution satisfied this assumption, the upper limit was calculated as x s, which excludes the top 5% of the data. The relationship between mean fluid measurement and each of the continuous variables (gestational age, birth weight, head circumference, and Apgar scores) was explored with Pearson correlation coefficients. Sex and type of delivery (vaginal versus cesarean section) were evaluated for a relationship with mean fluid width using a two-sided Student s t-test. the extraaxial space as detailed earlier. Figure 2B shows a sagittal image from the same patient, demonstrating the correct measurement of the depth of the fluid over a cortical gyrus. In contrast, Figure 3 demonstrates an inaccurate measurement made between the calvarium and a cortical sulcus. The range of values obtained for the width of the extraaxial space was 0 to 3.3 mm (Table 1). Seventyseven of 79 newborn infants had measurable quantities of extraaxial fluid. The mean width of the fluid space was 1.6 ± 0.8 mm. Figure 3 Sagittal sonogram with calipers (+) applied to the depth of a cortical sulcus. This measurement overestimates the true width of the extraaxial compartment. The craniocortical width is correctly measured over the surface of a cortical gyrus as shown by other calipers (x). RESULTS Figure 2A is a representative scan obtained in the coronal plane, with settings optimized for imaging

4 414 NEONATAL INTRACRANIAL EXTRAAXIAL COMPARTMENT J Ultrasound Med 17: , 1998 A slight negative linear relationship (r = 0.23) was found between gestational age and mean fluid width (Table 2). No significant linear correlation was found between the extraaxial fluid width and birth weight, head circumference, and Apgar scores. Similarly, no significant relationship was noted for method of delivery or sex using Student s t-test at the 0.05 level of significance. DISCUSSION Traditionally, it had been thought that most extraaxial fluid collections in infants were pathologic. Widening of the fluid-filled cortical sulci in infants was first described with pneumoencephalography by Harwood-Nash and Fitz. 4 Similar findings on CT were first reported by Fukuyama and colleagues in a small number of infants. 5 A qualitative evaluation by Kleinman and associates involved diagnostic CT scans that were otherwise deemed to be normal. 6 Discrete and variably sized fluid-filled subarachnoid spaces were demonstrated in most children under 2 years of age, although neonates were not included in the study. Pedersen and coworkers studied the interhemispheric fissure width in a small number of infants with neurologic abnormalities and apparently normal CT scans and reported a 95th percentile measurement of 6.6 mm. 7 However, it should be noted that with axial CT, accurate measurement of the subarachnoid spaces near the anterior fontanelle is not possible and this result is therefore not directly comparable to sonographic measurements. Slovis and colleagues first reported the ability of ultrasonography to image extracerebral fluid collections in neonates with intracranial pathologic conditions. 8 A small (nine cases) series of neonates examined with a 10 MHz transducer by Govaert and associates exhibited a range of values for the sinocortical width of up to 2 mm, similar to the results reported here. 1 In the most comprehensive study to date, Libicher and coauthors studied 89 infants, most of whom were older than those in our study (mean age, 105 days; no neonates) with a 5 MHz transducer. 2 Their results suggested 95th percentile measurements of 3 mm for the sinocortical width and 4 mm for craniocortical width. With the wide availability of 10 MHz transducers, higher resolution scanning of the extraaxial space can now be performed routinely. Our results, obtained in a sample of term newborn infants and using a 10 MHz transducer exclusively, show that the normal extraaxial space consistently measures less than 3.3 mm in width. A slight negative linear relationship of width of the fluid cavity to gestational age at birth was also documented. Although it was statistically significant, this may not be clinically relevant. A similar inverse relationship of gestational age to the size of the subarachnoid spaces has been observed in prior anatomic studies. 9 In conclusion, we have shown that small amounts of extraaxial fluid are common and normal in newborn infants. The upper limits of normal for the width of the extraaxial space in term neonates is at least 3.3 mm. Extraaxial fluid collections in this size range should be considered incidental and benign when seen as an isolated finding in term newborn infants. As our data included only term or near-term neonates and sonograms that were otherwise normal, we cannot comment on the significance of extraaxial fluid in premature infants or in association with other intracranial structural abnormalities. Our results show that high resolution ultrasonography using a 10 MHz transducer is a simple, noninvasive, and cost-effective method of screening the extraaxial space when this normal variant is recognized. In most cases, more costly evaluation requiring general anesthesia or exposure to ionizing radiation, or both, will thereby be obviated or at least selected more judiciously. Table 1: Demographic Data and Mean Width of the Fluid Space Variable Mean (SD) Range Gestational age (wk) 39.3 (1.6) Head Circumference (cm) 34.1 (1.5) Weight (g) 3324 (471) Apgar score, 1 min 8.5 (0.7) 6 9 Apgar score, 5 min 9.0 (0.3) 8 10 Mean volume of fluid (mm) 1.57 (0.78) Table 2: Statistical Data Variable r P value Gestational age Head circumference Weight Apgar score, 1 min Apgar score, 5 min

5 J Ultrasound Med 17: , 1998 FRANKEL ET AL 415 REFERENCES 1. Govaert P, Pauwels W, Vanhaesebrouck P, et al: Ultrasound measurement of the subarachnoid space in infants. Eur J Pediatr 148:412, Libicher M, Tröger J: Ultrasound measurement of the subarachnoid space in infants: Normal values. Radiology 184:749, Bezinque SL, Slovis TL, Touchette AS, et al: Characterization of superior sagittal sinus blood flow velocity using color flow Doppler in neonates and infants. Pediatr Radiol 25:175, Harwood-Nash DC, Fitz CR (Eds): Pneumoencephalography. In Neuroradiology in Infants and Children. St. Louis, CV Mosby, 1976, p Fukuyama Y, Miyao M, Ishizu T, et al: Developmental changes in normal cranial measurements by computed tomography. Dev Med Child Neurol 21:425, Kleinman PK, Zito JL, Davidson RI, et al: The subarachnoid spaces in children: Normal variations in size. Radiology 147:455, Pedersen H, Gyldensted M, Gyldensted C: Measurement of the normal ventricular system and supratentorial subarachnoid space in children with computed tomography. Neuroradiology 17:231, Slovis TL, Kelly JK, Eisenbrey AB, et al: Detection of extracerebral fluid collections by real-tine sector scanning through the anterior fontanelle. J Ultrasound Med 1:41, Fitz CR: Developmental anomalies of the brain. In Rosenberg R (Ed): The Clinical Neurosciences. New York, Churchill Livingstone, 1984, p 215 ERRATA In the article Use of Intraoperative Ultrasonography During Hepatic Transplantation (J Ultrasound Med 17:1,1998), the list of authors should have read as follows: David L. Waldman, MD, PhD, David E. Lee, MD, Debra J. Ruebens, MD, Oscar Bronsther, MD, Mark S. Orloff, MD Dr. Debra J. Ruebens was inadvertently omitted from the list that appeared with the article. In the article Color Doppler Hemodynamic Evaluation of Flow to the Normal Hip (J Ultrasound Med 17:275, 1998), the vein labeled with an arrow in Figure 2B should be the lateral, not the medial circumflex vein.