Neuroradiology
Cerebrospinal Fluid Shunt Function and Hydrocephalus in the Pediatric Age Group A Radiographic/Clinical Correlation 1
F. Reed Murtagh, M.D.,2 Robert M. Quencer, M.D., and Catherine A. Poole, M.D. Eighty-four pediatric patients were evaluated clinically and radiographically on 112 separate admissions for suspected increased intracranial pressure and possible shunt malfunction. The shunt system was tested clinically in each patient and correlated with the ventricular size as determined by CT. Of the patients with enlarged ventricles, 870/0 had an improperly functioning shunt and 13 % had a normally functioning shunt. Of the patients with normal or small ventricles, 93 % had a normally functioning shunt. Thus the "false negative" rate was 4 % (small ventricles with a nonfunctioning shunt) and the "false positive" rate was 13 % (large ventricles with a functioning shunt). INDEX TERMS: Cerebrospinal fluid, flow dynamics • Computed tomography, head, 1[0] .1211 • Computed tomography, pediatric • Hydrocephalus, 1[0].145; 1[6].145 • (Skull andcontents, postoperative shunt, 1[0].451) Radiology 132:385-388, August 1979
TABLE I: VENTRICULAR SIZE vs. CLINICAL STATUS OF SHUNT
ALTHOUGH many articles have been written about the 1-\ radiographic evaluation of hydrocephalus in the pe ... diatric age group (1-6), few have attempted to correlate the clinical and radiographic features of shunt malfunction (7-11). Because computed tomography (CT) is an ideal method of examining the ventricular system, we decided to review a large number of patients with clinical signs of increased intracranial pressure who were thought to have shunt malfunction. Our object was to relate shunt function to ventricular size and to determine the reliability of CT as a measure of shunt malfunction.
Functioning Nonfunctioning Total Shunt Shunt ( %) Ventricles smallor normal in size Ventricles large or enlarging Total
56* 8
64
60
4 44 48
52
m
--------~------~---
* Includes 7 functioning shunts in patients with increased intracranial pressure due to an expanding intracranial mass, consisting of a subarachnoid cyst in 3 cases, a tumor in 2, and a subdural hematoma in
2
METHODS AND MATERIALS
Eighty-four consecutive patients aged 13 and under, seen on 112 separate hospital admissions, have been evaluated over the past two years because of clinical signs of increased intracranial pressure (ICP) and possible shunt malfunction. Thirty-three had intraventricular obstructive hydrocephalus due to aqueductal stenosis, tumor, or a subarachnoid cyst; 47 had extraventricular obstructive hydrocephalus as the result of hemorrhage, underdevelopment, myelomeningocele, or trauma; and 4 had miscellaneous forms of hydrocephalus. All presented with vomiting, incontinence, headache, blurred vision, seizures, decreased consciousness, increased head circumference, papilledema, and/or bulging fontanelles. Shunting devices were tested in all patients upon admission, and shunt malfunction was diagnosed when the device flushed or pumped improperly. Ventriculoperitoneal (VP) (96 cases), ventriculoatrial (VA) (13 cases), and a combination of VP
and VA (3 cases) were the only types of shunts present in the patients we studied. No attempt is made in this paper to relate the type of shunt to malfunction. The radiographic evaluation included CT of the brain, skull series, and chest radiographs, supplemented by abdominal radiographs in patients with VP shunts. A GE CT/T . or C'I /EMI 1005 scanner was employed. Radionuclide . studies were carried out in 4 cases and ventriculography in 1. Although radionuclide cisternography or direct injection of 1111n into the shunt reservoir is the most definitive means of demonstrating CSF physiology and shunt patency, it is interesting to note that the clinicians used it in only 4 cases in our study (12). The lateral ventricles were considered enlarged when the blcaudatedlstance was greater than one-sixth the diameter of the brain (13) or as enlarging when there was clear-cut enlargement compared to previous scans; they were considered small when they could not be identified or were "slit-like."
1 From the Department of Radiology, University of Miami School of Medicine, Miami, Fla. Presented at the Sixty-fourth Scientific Assembly and Annual Meeting of the Radiological Society of North America, Chicago, Ill., Nov. 26-Dec. 1, 1978. Received Dec. 13, 1978; accepted and revision requested March 30, 1979; revision received April 11. 2 Present address: Department of Radiology, Neuroradiology Section, Veterans Administration Hospital, 13000 N. 30thSt., Tampa, Fla. 33612; Department of Radiology (F.R.M., Assistant Professor), University of South Florida, Box 17, 12901 N. 30th St.,Tampa, Fla. 33612. sjh
385
386
F. REED MURTAGH AND OTHERS
Fig. 1. CT scan of the brain in a 6-year-old girl with aqueductal stenosis shows large lateral ventricles. The tip of the shunt tube is embedded in the frontal lobe parenchyma. This is an example of large ventricles with a nonfunctioning shunt.
August 1979
Fig. 2. CT scan of the brain in a 10-year-old boy with a myelomeningocele and an Arnold-Chiari malformation. Although the shunt was functioning normally, the ventricles were enlargedbecause of an extremely thin cortical mantle. This is an example of a "false positive" scan.
RESULTS
Fig. 3. eT scan of the brain in a 13-year-old boy. Clinically, the shunt was functioning well. An enlarging subarachnoid cyst of the quadrigeminal plate (arrowheads) was elevating the intracranial pressure, causing obstructive hydrocephalus. The cyst required a separate shunt tube for drainage. This and Figure 4 are examples of increased pressure due to an enlarging underlying mass, even though the shunt is functioning well.
The 112 examinations were divided into two groups based on the CT findings: those with small or normal-sized ventricles and those with enlarged or enlarging ventricles. These two groups were then subdivided into patients with a properly or improperly functioning shunt according to clinical testing. This categorization allowed us to correlate shunt function and ventricular size in patients with suspected increased ICP.
Our results are summarized in TABLE I. Of 52 patients with enlarged or enlarging ventricles, 44 (87 %) had improperly functioning shunts requiring surgical revision; but in 8 (13 % ) the shunt was functioning normally. Figure 1 shows a nonfunctioning shunt in a patient with enlarged ventricles. Because the tip of the catheter was outside the ventricular system and within the frontal lobe parenchyma, there was shunt malfunction. The patient in Figure 2 had huge ventricles and a myelomeningocele, but the ventricular size was unchanged from a previous CT scan and the shunt was working well. This patient was studied because of increasing seizure activity and had had many previous shunt revisions for hydrocephalus. In this case the seizures were due to insufficient medication. Of 60 patients with small or normal ventricles, 56 (93 % ) had normally functioning shunts and 4 (7 % ) had improperly functioning shunts requiring surgical revision. In 7 of the 56 patients with normal ventricles and a functioning shunt, the cause of the increased ICP was found by CT to be an expanding intracranial mass, consisting of a subarachnoid cyst in 3 cases, an intracranial tumor in 2, and a chronic subdural hematoma in 2. Figures 3 and 4 show patients with normal ventricular size and a properly functioning shunt who had increasing pressure due to other causes, while Figure 5 illustrates a nonfunctioning shunt in a patient with collapsed ventricles. DISCUSSION
Hydrocephalus is a dynamic abnormality involving an
Vol. 132
CSF SHUNT FUNCTION AND HYDROCEPHALUS
increase in the volume of the cerebrospinal fluid (CSF) in the ventricular system due to relative or absolute obstruction of the passage of CSF between its origin and the sites of absorption (14). This obstruction can occur anywhere along the course of the ventricular system and may be the result of tumor, inflammation, subarachnoid hemorrhage, or a congenital abnormality. Approximately 30% of the CSF comes from the choroid plexus, 30 % from the ventricular ependyma, 20 % from the intracranial subarachnoid lining, and 20 % from the spinal subarachnoid lining (15). Since 50 % is formed within the lateral ventricles, obstruction of flow will result in a rapid increase in the size of the ventricular system (14). In addition, the ependymal lining may tear, particularly at the angles of the frontal horns, causing periventricular edema to be visible on CT scans (16, 17). Shunting is the mechanical redirection of CSF from an obstructed cavity to an area capable of fluid reabsorption. Just about any body cavity capable of reabsorbing fluid has been used by surgeons as a repository for CSF (14). The proximal end of a Silastlc tube is placed into a lateral ventricle, usually the right, and the distal end is placed in an organ or cavity, most frequently the peritoneal cavity or right atrium. A Holter, Pudenz, Portnoy, or Hakim valve provides unidirectional flow of CSF and is capable of transcutaneous flushing. In this series, malfunction ranged from the tip of the proximal intraventricular catheter to the distal end of the shunt system. The inability of the ventricles to diminish significantly with adequate shunting in 13 % of our patients is explained by the relatively thin cortical mantle (5), which may have been due to long-standing hydrocephalus prior to adequate shunting or to frequent episodes of shunt obstruction resulting in pressure damage to the mantle. This group may be considered "false positive," since the ventricles were enlarged but shunt function was normal. In the 56 patients with small or normal-sized ventricles and normal shunt function, alternative explanations had to be found for the change in their clinical status: these included infection of the shunt or other organs, metabolic imbalance, improper medication, nutritional deficiencies, or an underlying mass (Figs. 3 and 4). In the 4 patients with normal or small ventricles and lack of shunt function, the tip of the shunt was located either within the lateral ventricle (2 cases) or outside the ventricular wall (2 cases) (Fig. 5); all 4 had shunt infection and ventriculitis followed by inflammatory ventricular fibrosis (5), which has been suggested as an explanation of noncompliance of the ventricular walls (7). Another theoretical cause of nondistensible ventricles might be prolonged complete ventricular collapse due to too much shunting of CSF from the system, possibly caused by a low-pressure shunt tube. The ventricles would then become coapted, much in the same way as the normal thalami in some cases (the massa intermedia), and be unable to dilate with increased CSF volume. What happens to the increased amount of CSF in the cranial vault in such cases is not known, but transependymal movement into the brain parenchyma may
387
Neuroradiology
Fig. 4. UnenhancedCT scan of the brain in a 4-year-old boy with myelomeningocele. The tip of the shunt tube is well placed for drainage of the lateral ventricles; the increased intracranial pressure was due to the bilateral subdural hematomas (arrows). Note the large posterior cisterna magna.
Fig. 5. CT scan of the brain in a 4-year-old girl with aqueductal stenosis and a clinically nonfunctioning shunt. The tip of the ventricular catheter lies across the midline and in the parenchyma of the left cerebral hemisphere. At surgery, it was held in place by fibrosis. Note the normal-sized ventricles with clinically increased intracranial pressure in this example of a "false negative" scan.
well be responsible for the increase in intracranial pressure. When shunt revision was performed in our 4 patients, it was difficult to remove the nonfunctioning shunt because of fibrosis around the tip of the catheter. This group may be considered' 'false negative," since the ventricles were normal or small but shunt malfunction was present. In these patients the CSF dynamics should not be considered normal simply because of normal or small ventricles on
388
F.
REED MURTAGH AND OTHERS
CT scanning: rather, the clinical status of the shunting system should be determined before the CT scan is interpreted.
CONCLUSION
Our study demonstrates that there is good correlation between ventricular size as seen on CT and shunt function. Specifically, 93 % of the patients whose CT scans showed small or normal ventricles had functioning shuntsand 87 % of those with large or enlarging ventricles had nonfunctioning shunts. However, a "false negative" rate of 4% (small ventricles with nonfunctioning shunt) and a "false positive" rate of 13 % (large ventricles with functioning shunt)were observed and must be taken into consideration when evaluating hydrocephalic patients for possible shunt malfunction. Long-standing hydrocephalus and a thin cortical mantle can result in a "false positive" scan, while ventricular fibrosis or coaptation can result in a "false negative" scan.
Department of Radiology Neuroradiology Section Veterans Administration Hospital 13000 N. 30th St. Tampa, Fla. 33612
REFERENCES 1. Altman J, James AE Jr: Ventriculo-venous cerebrospinal fluid shunts. Roentgenologic analysis. Am J Roentgenol 112:237-250, Jun 1971 2. Fitz CR, Harwood-Nash DC: Computed tomography in hydrocephalus. J Comput Tomogr 2:91-107, Jun 1978
August 1979
3. Kingsley D, Kendall BE: The value of computed tomography in the evaluation of the enlarged head. Neuroradiology 15:59-71, 27 Apr 1978 4. Larson EB, Omenn GS, Magno J: Impact of computed tomography on the care of patients with suspected hydrocephalus. Am J RoentgenoI131:41-44, Jul 1978 5. Naidich TP, Epstein F, Lin JP, et al: Evaluation of pediatric hydrocephalus by computed tomography. Radiology 119:337-345, May 1976 6. New PFJ, Weiner MA: The radiological investigation of hydrocephalus. Radiol Clin North Am 9: 117-140, Apr 1971 7. Epstein F, Naidich T, Kricheff I, et al: Role of computerized axial tomography in diagnosis, treatment and follow-up of hydrocephalus. Preliminary communication. Child's Brain 3:91-100, 1977 8. Forrest DM, Cooper DGW: Complications of ventriculo-atrial shunts. A review of 455 cases. J Neurosurg 29:506-512, Nov 1968 9. Kurlander GJ, Chua GT: Roentgenology of ventriculo-atrial shunts for the treatment of hydrocephalus. Am J Roentgenol 101: 157-167, Sep 1967 10. Little JR, Rhoton AL Jr, Mellinger JF: Comparison of ventriculoperitoneal and ventriculoatrial shunts for hydrocephalus in children. Mayo Clin Proc 47:396-401, Jun 1972 11. Murtagh F, Lehman R: Peritoneal shunts in the management of hydrocephalus. JAMA 202: 1010-1 0 14, 11 Dec 1967 12. Gilday DL, Kellam J: 111In_DTPA evaluation of CSF diversionary shunts in children. J Nucl Med 14:920-923, Dec 1973 13. Haug G: Age and sex dependence of the size of normal ventricles on computed tomography. Neuroradiology 14:201-204, 31 Dec 1977 14. Harwood-Nash DC, Fitz CR: Neuroradiology in Infants and Children. St. Louis, Mo. Mosby, 1976, Vol 2, Chapt 10, pp 609-667 15. Milhorat TH: The third circulation revisited. J Neurosurg 42: 628-645, Jun 1975 16. James AE Jr, Strecker E-P, Sperber E, et al: An alternative pathway of cerebrospinal fluid absorption in communicating hydrocephalus. Transependymal movement. Radiology 111:143-146, Apr 1974 17. Pasquini U, Bronzini M, Gozzoli E, et al: Periventricular hypodensity in hydrocephalus: a clinico-radiological and mathematical analysis using computed tomography. J Comput Assist Tomogr 1: 443-448,Oct. 1977 18. Taveras JM, Wood EH: Diagnostic Neuroradiology. Baltimore, Williams & Wilkins, 2d Ed, 1976, Vol 1, pp 361-398
Cerebrospinal Fluid Shunt Function and Hydrocephalus in the Pediatric Age Group A Radiographic/Clinical Correlation 1
F. Reed Murtagh, M.D.,2 Robert M. Quencer, M.D., and Catherine A. Poole, M.D. Eighty-four pediatric patients were evaluated clinically and radiographically on 112 separate admissions for suspected increased intracranial pressure and possible shunt malfunction. The shunt system was tested clinically in each patient and correlated with the ventricular size as determined by CT. Of the patients with enlarged ventricles, 870/0 had an improperly functioning shunt and 13 % had a normally functioning shunt. Of the patients with normal or small ventricles, 93 % had a normally functioning shunt. Thus the "false negative" rate was 4 % (small ventricles with a nonfunctioning shunt) and the "false positive" rate was 13 % (large ventricles with a functioning shunt). INDEX TERMS: Cerebrospinal fluid, flow dynamics • Computed tomography, head, 1[0] .1211 • Computed tomography, pediatric • Hydrocephalus, 1[0].145; 1[6].145 • (Skull andcontents, postoperative shunt, 1[0].451) Radiology 132:385-388, August 1979
TABLE I: VENTRICULAR SIZE vs. CLINICAL STATUS OF SHUNT
ALTHOUGH many articles have been written about the 1-\ radiographic evaluation of hydrocephalus in the pe ... diatric age group (1-6), few have attempted to correlate the clinical and radiographic features of shunt malfunction (7-11). Because computed tomography (CT) is an ideal method of examining the ventricular system, we decided to review a large number of patients with clinical signs of increased intracranial pressure who were thought to have shunt malfunction. Our object was to relate shunt function to ventricular size and to determine the reliability of CT as a measure of shunt malfunction.
Functioning Nonfunctioning Total Shunt Shunt ( %) Ventricles smallor normal in size Ventricles large or enlarging Total
56* 8
64
60
4 44 48
52
m
--------~------~---
* Includes 7 functioning shunts in patients with increased intracranial pressure due to an expanding intracranial mass, consisting of a subarachnoid cyst in 3 cases, a tumor in 2, and a subdural hematoma in
2
METHODS AND MATERIALS
Eighty-four consecutive patients aged 13 and under, seen on 112 separate hospital admissions, have been evaluated over the past two years because of clinical signs of increased intracranial pressure (ICP) and possible shunt malfunction. Thirty-three had intraventricular obstructive hydrocephalus due to aqueductal stenosis, tumor, or a subarachnoid cyst; 47 had extraventricular obstructive hydrocephalus as the result of hemorrhage, underdevelopment, myelomeningocele, or trauma; and 4 had miscellaneous forms of hydrocephalus. All presented with vomiting, incontinence, headache, blurred vision, seizures, decreased consciousness, increased head circumference, papilledema, and/or bulging fontanelles. Shunting devices were tested in all patients upon admission, and shunt malfunction was diagnosed when the device flushed or pumped improperly. Ventriculoperitoneal (VP) (96 cases), ventriculoatrial (VA) (13 cases), and a combination of VP
and VA (3 cases) were the only types of shunts present in the patients we studied. No attempt is made in this paper to relate the type of shunt to malfunction. The radiographic evaluation included CT of the brain, skull series, and chest radiographs, supplemented by abdominal radiographs in patients with VP shunts. A GE CT/T . or C'I /EMI 1005 scanner was employed. Radionuclide . studies were carried out in 4 cases and ventriculography in 1. Although radionuclide cisternography or direct injection of 1111n into the shunt reservoir is the most definitive means of demonstrating CSF physiology and shunt patency, it is interesting to note that the clinicians used it in only 4 cases in our study (12). The lateral ventricles were considered enlarged when the blcaudatedlstance was greater than one-sixth the diameter of the brain (13) or as enlarging when there was clear-cut enlargement compared to previous scans; they were considered small when they could not be identified or were "slit-like."
1 From the Department of Radiology, University of Miami School of Medicine, Miami, Fla. Presented at the Sixty-fourth Scientific Assembly and Annual Meeting of the Radiological Society of North America, Chicago, Ill., Nov. 26-Dec. 1, 1978. Received Dec. 13, 1978; accepted and revision requested March 30, 1979; revision received April 11. 2 Present address: Department of Radiology, Neuroradiology Section, Veterans Administration Hospital, 13000 N. 30thSt., Tampa, Fla. 33612; Department of Radiology (F.R.M., Assistant Professor), University of South Florida, Box 17, 12901 N. 30th St.,Tampa, Fla. 33612. sjh
385
386
F. REED MURTAGH AND OTHERS
Fig. 1. CT scan of the brain in a 6-year-old girl with aqueductal stenosis shows large lateral ventricles. The tip of the shunt tube is embedded in the frontal lobe parenchyma. This is an example of large ventricles with a nonfunctioning shunt.
August 1979
Fig. 2. CT scan of the brain in a 10-year-old boy with a myelomeningocele and an Arnold-Chiari malformation. Although the shunt was functioning normally, the ventricles were enlargedbecause of an extremely thin cortical mantle. This is an example of a "false positive" scan.
RESULTS
Fig. 3. eT scan of the brain in a 13-year-old boy. Clinically, the shunt was functioning well. An enlarging subarachnoid cyst of the quadrigeminal plate (arrowheads) was elevating the intracranial pressure, causing obstructive hydrocephalus. The cyst required a separate shunt tube for drainage. This and Figure 4 are examples of increased pressure due to an enlarging underlying mass, even though the shunt is functioning well.
The 112 examinations were divided into two groups based on the CT findings: those with small or normal-sized ventricles and those with enlarged or enlarging ventricles. These two groups were then subdivided into patients with a properly or improperly functioning shunt according to clinical testing. This categorization allowed us to correlate shunt function and ventricular size in patients with suspected increased ICP.
Our results are summarized in TABLE I. Of 52 patients with enlarged or enlarging ventricles, 44 (87 %) had improperly functioning shunts requiring surgical revision; but in 8 (13 % ) the shunt was functioning normally. Figure 1 shows a nonfunctioning shunt in a patient with enlarged ventricles. Because the tip of the catheter was outside the ventricular system and within the frontal lobe parenchyma, there was shunt malfunction. The patient in Figure 2 had huge ventricles and a myelomeningocele, but the ventricular size was unchanged from a previous CT scan and the shunt was working well. This patient was studied because of increasing seizure activity and had had many previous shunt revisions for hydrocephalus. In this case the seizures were due to insufficient medication. Of 60 patients with small or normal ventricles, 56 (93 % ) had normally functioning shunts and 4 (7 % ) had improperly functioning shunts requiring surgical revision. In 7 of the 56 patients with normal ventricles and a functioning shunt, the cause of the increased ICP was found by CT to be an expanding intracranial mass, consisting of a subarachnoid cyst in 3 cases, an intracranial tumor in 2, and a chronic subdural hematoma in 2. Figures 3 and 4 show patients with normal ventricular size and a properly functioning shunt who had increasing pressure due to other causes, while Figure 5 illustrates a nonfunctioning shunt in a patient with collapsed ventricles. DISCUSSION
Hydrocephalus is a dynamic abnormality involving an
Vol. 132
CSF SHUNT FUNCTION AND HYDROCEPHALUS
increase in the volume of the cerebrospinal fluid (CSF) in the ventricular system due to relative or absolute obstruction of the passage of CSF between its origin and the sites of absorption (14). This obstruction can occur anywhere along the course of the ventricular system and may be the result of tumor, inflammation, subarachnoid hemorrhage, or a congenital abnormality. Approximately 30% of the CSF comes from the choroid plexus, 30 % from the ventricular ependyma, 20 % from the intracranial subarachnoid lining, and 20 % from the spinal subarachnoid lining (15). Since 50 % is formed within the lateral ventricles, obstruction of flow will result in a rapid increase in the size of the ventricular system (14). In addition, the ependymal lining may tear, particularly at the angles of the frontal horns, causing periventricular edema to be visible on CT scans (16, 17). Shunting is the mechanical redirection of CSF from an obstructed cavity to an area capable of fluid reabsorption. Just about any body cavity capable of reabsorbing fluid has been used by surgeons as a repository for CSF (14). The proximal end of a Silastlc tube is placed into a lateral ventricle, usually the right, and the distal end is placed in an organ or cavity, most frequently the peritoneal cavity or right atrium. A Holter, Pudenz, Portnoy, or Hakim valve provides unidirectional flow of CSF and is capable of transcutaneous flushing. In this series, malfunction ranged from the tip of the proximal intraventricular catheter to the distal end of the shunt system. The inability of the ventricles to diminish significantly with adequate shunting in 13 % of our patients is explained by the relatively thin cortical mantle (5), which may have been due to long-standing hydrocephalus prior to adequate shunting or to frequent episodes of shunt obstruction resulting in pressure damage to the mantle. This group may be considered "false positive," since the ventricles were enlarged but shunt function was normal. In the 56 patients with small or normal-sized ventricles and normal shunt function, alternative explanations had to be found for the change in their clinical status: these included infection of the shunt or other organs, metabolic imbalance, improper medication, nutritional deficiencies, or an underlying mass (Figs. 3 and 4). In the 4 patients with normal or small ventricles and lack of shunt function, the tip of the shunt was located either within the lateral ventricle (2 cases) or outside the ventricular wall (2 cases) (Fig. 5); all 4 had shunt infection and ventriculitis followed by inflammatory ventricular fibrosis (5), which has been suggested as an explanation of noncompliance of the ventricular walls (7). Another theoretical cause of nondistensible ventricles might be prolonged complete ventricular collapse due to too much shunting of CSF from the system, possibly caused by a low-pressure shunt tube. The ventricles would then become coapted, much in the same way as the normal thalami in some cases (the massa intermedia), and be unable to dilate with increased CSF volume. What happens to the increased amount of CSF in the cranial vault in such cases is not known, but transependymal movement into the brain parenchyma may
387
Neuroradiology
Fig. 4. UnenhancedCT scan of the brain in a 4-year-old boy with myelomeningocele. The tip of the shunt tube is well placed for drainage of the lateral ventricles; the increased intracranial pressure was due to the bilateral subdural hematomas (arrows). Note the large posterior cisterna magna.
Fig. 5. CT scan of the brain in a 4-year-old girl with aqueductal stenosis and a clinically nonfunctioning shunt. The tip of the ventricular catheter lies across the midline and in the parenchyma of the left cerebral hemisphere. At surgery, it was held in place by fibrosis. Note the normal-sized ventricles with clinically increased intracranial pressure in this example of a "false negative" scan.
well be responsible for the increase in intracranial pressure. When shunt revision was performed in our 4 patients, it was difficult to remove the nonfunctioning shunt because of fibrosis around the tip of the catheter. This group may be considered' 'false negative," since the ventricles were normal or small but shunt malfunction was present. In these patients the CSF dynamics should not be considered normal simply because of normal or small ventricles on
388
F.
REED MURTAGH AND OTHERS
CT scanning: rather, the clinical status of the shunting system should be determined before the CT scan is interpreted.
CONCLUSION
Our study demonstrates that there is good correlation between ventricular size as seen on CT and shunt function. Specifically, 93 % of the patients whose CT scans showed small or normal ventricles had functioning shuntsand 87 % of those with large or enlarging ventricles had nonfunctioning shunts. However, a "false negative" rate of 4% (small ventricles with nonfunctioning shunt) and a "false positive" rate of 13 % (large ventricles with functioning shunt)were observed and must be taken into consideration when evaluating hydrocephalic patients for possible shunt malfunction. Long-standing hydrocephalus and a thin cortical mantle can result in a "false positive" scan, while ventricular fibrosis or coaptation can result in a "false negative" scan.
Department of Radiology Neuroradiology Section Veterans Administration Hospital 13000 N. 30th St. Tampa, Fla. 33612
REFERENCES 1. Altman J, James AE Jr: Ventriculo-venous cerebrospinal fluid shunts. Roentgenologic analysis. Am J Roentgenol 112:237-250, Jun 1971 2. Fitz CR, Harwood-Nash DC: Computed tomography in hydrocephalus. J Comput Tomogr 2:91-107, Jun 1978
August 1979
3. Kingsley D, Kendall BE: The value of computed tomography in the evaluation of the enlarged head. Neuroradiology 15:59-71, 27 Apr 1978 4. Larson EB, Omenn GS, Magno J: Impact of computed tomography on the care of patients with suspected hydrocephalus. Am J RoentgenoI131:41-44, Jul 1978 5. Naidich TP, Epstein F, Lin JP, et al: Evaluation of pediatric hydrocephalus by computed tomography. Radiology 119:337-345, May 1976 6. New PFJ, Weiner MA: The radiological investigation of hydrocephalus. Radiol Clin North Am 9: 117-140, Apr 1971 7. Epstein F, Naidich T, Kricheff I, et al: Role of computerized axial tomography in diagnosis, treatment and follow-up of hydrocephalus. Preliminary communication. Child's Brain 3:91-100, 1977 8. Forrest DM, Cooper DGW: Complications of ventriculo-atrial shunts. A review of 455 cases. J Neurosurg 29:506-512, Nov 1968 9. Kurlander GJ, Chua GT: Roentgenology of ventriculo-atrial shunts for the treatment of hydrocephalus. Am J Roentgenol 101: 157-167, Sep 1967 10. Little JR, Rhoton AL Jr, Mellinger JF: Comparison of ventriculoperitoneal and ventriculoatrial shunts for hydrocephalus in children. Mayo Clin Proc 47:396-401, Jun 1972 11. Murtagh F, Lehman R: Peritoneal shunts in the management of hydrocephalus. JAMA 202: 1010-1 0 14, 11 Dec 1967 12. Gilday DL, Kellam J: 111In_DTPA evaluation of CSF diversionary shunts in children. J Nucl Med 14:920-923, Dec 1973 13. Haug G: Age and sex dependence of the size of normal ventricles on computed tomography. Neuroradiology 14:201-204, 31 Dec 1977 14. Harwood-Nash DC, Fitz CR: Neuroradiology in Infants and Children. St. Louis, Mo. Mosby, 1976, Vol 2, Chapt 10, pp 609-667 15. Milhorat TH: The third circulation revisited. J Neurosurg 42: 628-645, Jun 1975 16. James AE Jr, Strecker E-P, Sperber E, et al: An alternative pathway of cerebrospinal fluid absorption in communicating hydrocephalus. Transependymal movement. Radiology 111:143-146, Apr 1974 17. Pasquini U, Bronzini M, Gozzoli E, et al: Periventricular hypodensity in hydrocephalus: a clinico-radiological and mathematical analysis using computed tomography. J Comput Assist Tomogr 1: 443-448,Oct. 1977 18. Taveras JM, Wood EH: Diagnostic Neuroradiology. Baltimore, Williams & Wilkins, 2d Ed, 1976, Vol 1, pp 361-398
