Effect of respiratory movement on cerebrospinal fluid dynamics in hydrocephalic infants with shunts HIROSHI YAMADA, M . D . , MASATAKA TAJIMA, M . D . ,

AND MASAHIRO NAGAYA, M.D.

Department of Neurosurgery and Department of Pediatric Surgery, Central Hospital, Aichi Prefecture Colony, Kasugai, Aichi, Japan v' The authors report a study of the effect of respiratory movement on intracranial, auricular, and intraperitoneal cerebrospinal fluid (CSF) pressure in hydrocephalic infants with shunts. Postoperative intraventricular pressures were also recorded for comparison. The intraventricular, right auricular, and intraperitoneal pressures rose during expiration and dropped during inspiration; the pressure changes were most marked while the infants were crying or straining. All pressures dropped simultaneously at the time of inspiration, but the auricular pressure was most significantly affected. It dropped to - 100 to -200 mm H~O when the patients cried, while intraventricular and intraperitoneal pressures remained above 0 mm H20. The postoperative intracranial pressures were in accord with these results; the pressures after ventriculoatrial shunt were significantly lower than those after ventriculoperitoneal shunt when the same pressure valves were used. KEY WoRDs 9 infantile hydrocephalus 9 ventriculoatriai shunt 9 ventriculoperitoneal shunt 9 respiratory movement 9 C S F pressure

V

ENTRICULOATRIAL (VA) shunt is currently the procedure most widely performed in the treatment of hydrocephalus in children; one of its most important complications involves the intracranial effects of long-standing decompression of the brain. 1,8,*,9-11,13,19,2~Relatively little has been written on the mechanism of these complications. Recently low intracranial pressures were shown to occur when hydrocephalic patients with shunts are in the erect posture. Fox, et al.) "8 and Portnoy, et al., ~6 have reported that posture is one of the most important factors effecting change in the cerebrospinal fluid (CSF) pressure in the brains of patients with shunts. The effects of 194

respiratory movement in this same group of patients has not been widely emphasized. 16 The purpose of this report is to elucidate the effects of respiratory movement on the intracranial, auricular, and intraperitoneal pressures and CSF dynamics in awake hydrocephalic infants with Shunts and to identify adequate related operative procedures. Material and Methods

Preshunting Studies The study was performed in six 2- to 4week-old infants before the shunting operation. With the unsedated infant lying supine, a No. 19 plastic needle was introduced J. Neurosurg. / Volume 42 / February, 1975

Effect of respiratory m o v e m e n t on CSF pressure in s h u n t s through the anterior fontanel into the right lateral ventricle. A standard Pudenz silicone VA shunt tube was cannulated into the right auricle through the internal jugular vein in the neck. Patency was maintained by intermittent irrigation with saline solution containing heparin. The intraperitoneal pressure was measured from water-filled open-tip catheters. A trocar needle puncture was made in the right hypochondrium, and a 15-cm silicone tube introduced through the trocar needle into the lower portion of the peritoneal cavity. The needle and silicone tubes were connected by plastic tubes to three-way stopcocks attached to an electromanometer with an inductance transducer. The respiratory pattern was recorded in the monitor, with all recordings made on a 4-channel polygraph. The transducer was placed at the level of the right auricle, approximately the same as the anterior fontanel level. Baseline ventricular, right auricle, and intraperitoneal CSF pressures were recorded while the patients were quiet. The pressure recordings were then made during infant activity; crying and straining were the most effective ways to elevate all pressures. We repeated the measurements at least three times in all infants, both while they were quiet and while they were crying. Post-shunting Studies Following the shunting procedure in 16 patients (five VA and 11 VP) the intraventricular pressure was recorded during infant activity. Postoperative intracranial pressure in the VA and VP shunts could thus be compared. Prior to any post-shunt recordings, the patients were kept lying flat for several hours to eliminate siphoning effects. This part of the study was carried out 4 to 18 days (average 9 days) after operation. Results

Awake hydrocephalic infants had wide fluctuations in intraventricular (IVP), right auricular (RAP), and intraperitoneal pressures (IPP) during the infants' activities (Figs. 1 and 2). IVP, RAP, and IPP simultaneously rose during expiration at crying or straining; IVP rose to 400 to 900 mm H20, RAP to 400 to 800 mm H20, and IPP to 400 to 750 mm H~O (Table 1). On the other hand, J. Neurosurg. / Volume 42 / February, 1975

these pressures dropped abruptly at inspiration. However, the IVP and IPP never dropped below zero during any infant's activity, while the RAP at inspiration dropped to - 1 0 0 or even - 2 0 0 mm Hg during crying or straining. The intraventricular pressure in postoperative patients showed significant contrast between those with VA and VP shunts (Table 2). In both groups the postoperative IVP during quiet as well as crying tended to be lower in the larger lateral ventricles. In patients with a VA shunt (Pudenz medium pressure) the opening pressures while the patients were quiet were below zero ( - 5 0 mm H20), but rose to 200 to 300 mm H20 during crying or straining (Fig. 3). Negative IVP appeared more significant in a patient with respiratory distress or one who cried vigorously. In these patients, overriding of the skull was more marked following the VA shunting procedure. On the other hand, in the VP shunt group (Pudenz medium pressure) the opening pressure of 50 to 100 mm H20 rose to 300 to 500 mm H20 during crying or straining (Fig. 4). This value is thought to be close to the normal infant IVP. In only two VP shunt patients was the pressure 0 mm H~O when they were quiet. These results were consistent with the previous preoperative recordings of IVP, RAP, and IPP. Discussion

Negative intraventricular pressures caused by shunt systems have been reported. The related complications include subdural hematoma, 4,8,9'18'19 ~1 overriding cranial bones, s~ craniosynostosis, TM and the conversion of a communicating to a noncommunicating hydrocephalus in the presence of a stenotic or occluded aqueduct? Recently a new factor has been introduced to explain this phenomenon; the length of the tubing from the ventricles to the heart or peritoneal cavity sets up a siphoning effect when the patient is in the upright position. This siphoning factor has been considered a significant way to reduce the intracranial pressure. Studies performed on hydrocephalic animals show a fairly direct relationship between the intraventricular pressure and the distance of the tubing below the ventricles, and a new valve has been devised that answers this clinical problem. 5,6,16 In the horizontal 195

H. Yamada, M. Tajima and M. Nagaya PRESSURE mmH20 CEREBRAL VENTRICLE 400 200 0 400

RIGHT ATRIUM

200

- 200 400

PERITONEAL CAVITY

200 O

-

-

RESPIRATORY PATTERN

Quiet

TIeR (SEC)

I I.I

Crying

I I t I k i I I I I I i i ~ t I i I 1 1 i

I I I

I I I I t k I I I I I I i

FIG. 1. Case 4. Graph showing record of intraventricular, right atrial, and intraperitoneal pressure in a 29-day-old hydrocephalic infant at rest and crying. Note the simultaneous elevation of the pressures during expiration, and the decrease during inspiration and crying. Right atrial pressure is most significantly affected. PRESSURE

r~n f120 CEREBRAL VENTRICLE

400 200 0

RICdlT ATRIUM

20O 0

~

-200

Quiet

TIeR (SEC)

Crying

I I I I I I I I t I $ l I L I I I 1 I ) I I I I I I I ~ I I I I I I I ~ I I I 1

FIG. 2. Case 5. Graph showing simultaneous intraventricular and right atrial pressure measurement in a 55-day-old hydrocephalic infant at rest and crying. Right atrial pressure during inspiration while crying drops to approximately -200 mm H~O. posture the siphoning apparently has no significant effect on the intracranial pressure. There are very few reports concerning the difference between postoperative intracranial pressure and pressure at the end of the distal tube. In this study we have demonstrated a significant difference in the IVP of patients ]96

with VP and VA shunts; postoperative IVP with VA shunts was lower than that of VP shunts when the same pressure valves were used. Both groups of patients had been kept recumbent before the pressure recording, so that the siphoning effect could be considered negligible. J. Neurosurg. / Volume 42 / February, 1975

Effect of respiratory m o v e m e n t on C S F pressure in s h u n t s TABLE 1 Simultaneous recording of right auricular, intraperitoneal, and intraventricular pressures in six cases Age 'Case (days), No. Sex

1 2 3 4 5 6

Diagnosis

45, M atresia of fourth ventricle 18, F Arnold-Chiari malformation 25, M Dandy-Walker malformation 29, F Arnold-Chiari malformation 55, F Arnold-Chiari malformation 27, F aqueductal stenosis

Right Auricular Intraperitoneal Intraventricular Pressure (RAP) Pressure (IPP) Pressure (IVP) (mm H20) (mm H20) (mm H~O) Quiet Crying Quiet Crying Quiet Crying InspiraExpiraInspira- ExpiraInspira- Expiration tion tion tion tion tion -- 104) --2004)

300-800 0-30 I00-150 400-700

150 200-300 500-900

- 10-0 - 2 0 0 - - 2 0

450-750 0-20

200 350-450 600-750

-- 5-0 -- 200-0

30-100 500-750

180 200-350 400-800

500--700not examined

- 10-0 - 2 0 0 - - 100 400-500 0-50

50-100 400-500

-- 10-4) -- 200-0

400-600 not examined

- 1(}4) - 1504)

350-800 0-30 50-150

100 200-250 350-400 120

400-600

200-300 400-600

250 300-500 500-850

TABLE 2 Summary of 16 cases of hydrocephalus with shunt, including pre- and postoperative bltraventricular pressure ( IVP) Case No. Sex

Diagnosis

Age at Head Preop Days Postop Time of Circum. at IVP When IVP Shunt Surgery (days) (cm) (mm H20) Recorded

Postop IVP (ram H~O) Quiet Crying

Group 1: VP shunt Arnold-Chiari malformation 1 F 2 F Arnold-Chiari malformation 3 M atresia of fourth ventricle 4 M aqueductal stenosis 5 M communicating hydrocephalus 6 F communicating hydrocephalus 7 M Arnold-Chiari malformation 8 F Arnold--Chiari malformation 9 F Arnold-Chiari malformation 10 F Arnold-Chiari malformation 11 F aqueductal stenosis

14 55 45 18 40 75 15 29 34 18 26

40.0 38.8 44.5 55.8 45.0 47.2 39.0 40.5 43.0 39.8 41.3

-150 150 145 110 130 -120 110 180 125

4 7 17 12 7 5 12 18 7 8 10

110 120 30 20 50 0 70 90 0 80 30

Group 12 13 14 15 16

74 18 25 9 13

51.8 38.2 46.2 38.9 42.5

120 -130 160 150

8 6 13 7 5

-35 5 -20 --30 --30

2: VA shunt F communicating hydrocephalus M Arnold-Chiari malformation M aqueductal stenosis F Arnold-Chiari malformation F Arnold-Chiari malformation

O u r study suggests that r e s p i r a t o r y m o v e ment, including coughing, sneezing, and straining, significantly lowers atrial pressure. A secondary effect will be aspiration o f C S F from the cerebral ventricle to the right atrium, with resulting negative i n t r a c r a n i a l pressure. T h e aspirating effect is considered J. Neurosurg. / Volume 42 / February, 1975

250- 350 300- 600 350 - 700 100- 150 250- 300 200- 300 250- 450 400- 600 250- 500 400- 550 400-600 125 150100200200-

200 230 150 250 300

m o r e i m p o r t a n t than the siphoning effect as long as the patient is r e c u m b e n t or is experiencing vigorous r e s p i r a t o r y m o v e m e n t , as in r e s p i r a t o r y distress or crying. This study enables one to a s s u m e that C S F will flow at the t i m e of inspiration, and that the difference between I V P and R A P , or I V P and I P P is 197

H. Yamada, M. Tajima and M. Nagaya VFP mmH20

300 200 lO0 0 -lO0

.,,~.h.v,.%-C'~X,\'~'f tyq,~.%',"~'e ~%,..r,,,,~,.,,x,

RESPIRATORY PATTERN

TIMER {SEC)

Crying

Quiet

~1111~ill

llllt

illllililI~lllll

III

lllm~lit

Fio. 3. Case 12. Intraventricular pressure recording in a 74-da~c-old infant with a VA shunt. The pressure is negative when the patient is quiet; it elevates during crying, but to no more than 200 mm H20. VFP mmH20

8O0

" :::

,

0

Quiet TIMER (SEC)

Crying

Quiet

I!!I)I!I!I!I!!!I!I[!!!!I!!!!!!!X!t!!IllI[)I!!I!!II!I!I!!!!]!~!!!![!!!!!!!!!!}!

!

Flo. 4. Case 8. Intraventricular pressure recording in a 29-day-old infant with a VP shunt. The pressure is 90 mm H20 when patient is quiet, and rises to 400 to 600 mm H20 with patient crying. greater during inspiration than during expiration. RAP and IPP are known to be higher than IVP during expiration. Furthermore, the difference between IVP and RAP is greater than that between IVP and IPP; this suggests that during expiration more CSF will flow in a VA than in a VP shunt. Finally, the aspirating effect of the right atrium during laborious respiration will reduce the lateral ventricle to a negative pressure following a VA shunting procedure. VA and VP shunts are widely recognized to significantly lower intraventricular pressure when the patient is erect. In .most hydrocephalic infants so treated, the brain mantle will become progressively thicker and the ventricular system smaller; by the time the patients are able to sit or stand up, the complications due to the siphoning effect will be a less important factor. But in cases of normal pressure hydrocephalus or chronic hydrocephalus with large lateral ventricles and a thin brain mantle, the intraventricular or lumbar pressure should be checked 198

periodically after the operation; if the pressure is negative or too low, the valve should be converted to one of higher pressure. As the infant develops, conversion from high to low or low to high pressure valves may be necessary. Recently we have been routinely performing VP shunts for infantile hydrocephalus. The postoperative IVP seems to be satisfactory; a lower pressure valve may be indicated in patients who are expected to be bedridden for a long time. When other congenital anomalies such as holoprosencephaly or hydranencephaly are associated with hydrocephalus the brain mantle may stay thin; but in these states the patients are usually bedridden and the siphoning effect is not as important as in a normally developing brain. We have observed no serious complications from the use of VP shunts in such infants. The development of a subdural hematoma after the insertion of a VA shunt for relief of hydrocephalus has been only infrequently reported. Becker and Nulsen 2 reported seven J. Neurosurg. / Volume 42 / February, 1975

Effect of respiratory m o v e m e n t on CSF pressure in s h u n t s cases of subdural hematoma in a series of 140 patients. In 175 consecutive ventriculocaval shunts, Illingworth 9 found eight cases of subdural hematoma in a series of 24 cases of normal pressure hydrocephalus. Greitz, et al., 7 reported three cases with angiographic evidence of subdural collections after VA shunts for normal pressure hydrocephalus and postulated that the collections were subdural CSF. We have seen six cases of subdural hematoma or effusion after VA shunt for chronic infantile hydrocephalus. Rayport and Reiss ~7 considered that normal ventricular pressure was sufficiently negative to obviate any significant further reduction of CSF pressure after VA shunting. Subdural hematoma seems to occur more commonly after VA than after VP shunts, although VA shunts are performed more often. A rise in jugular pressure is generally known to accompany a rise in CSF pressure; atrial (venous) pressure is also closely related to CSF pressure. Conception of atrial or intraperitoneal pressure during vigorous respiratory movement such as crying or straining is poor, s although these possible pressure changes are important. Natelson and Molnar x' reported three adult patients in whom the atrial catheter coiled on itself after a VA shunt. Pressures in the jugular and subclavian veins and right atrium were measured during respiration and coughing. A dramatic rise and fall in atrial pressure was recorded during a brief period of respiratory movement. Pressure is known to drop at the time of inspiration, but it has not been widely recognized that pressure falls to - 1 0 0 to -200 mm H20 during crying or straining. Resistance at the VP shunt termination varies with posture, site of shunt termination, obesity, adherence of adjacent tissues, and the rate of CSF absorption. Intraabdominal pressure also varies with posture or site; it ranges from 0 to 150 mm H20 in the supine position and is known to be negative in the subphragmatic portion of the abdomen during respiratory movement? 2'15'1s In our study the termination of the intrapefitoneal tube was in the lower abdomen; the average pressure while patients were quiet was 0 to 180 mm H20, and never fell below 0 mm H20 even while patients were crying. In early infancy the respiratory effect and peritoneal capacity to absorb CSF are important factors, particularly when deciding whether to J. Neurosurg. / Volume 42 / February, 1975

terminate the shunt in the right atrium or peritoneal cavity. The intraperitoneal CSF absorption capacity seemed to be sufficient in our patients. Our study suggests that in infants it is easier to control IVP by a VP shunt since IPP is less significantly affected by vigorous respiratory movement than RAP. Intraventricular occlusion of VA shunt catheters occurred more often than with VP catheters; on the other hand, obstruction of the properly placed distal tube was less frequent than in the VA group. There are still many disagreements regarding the ideal treatment of infantile hydrocephalus. We believe it advisable to use a higher pressure valve or some other antinegative pressure method in association with VA shunt procedures. The hazards of low intracranial pressures and frequent elongation of the distal tube in VA shunts have influenced our decision to prefer the use of VP shunts in the treatment of infantile hydrocephalus.

References 1. Andersson H: Craniosynostosis as a complication after operation for hydrocephalus. Acta Paediatr Scand 55:192-196, 1966 2. Becker CP, Nulsen FE: Control of hydrocephalus by valve-regulated venous shunt: avoidance of complications in prolonged shunt maintenance. J Neurosurg 28:215-226, 1968 3. Foltz EL, Shurtleff DB: Conversion of communicating hydrocephalus to stenosis or occlusion of the aqueduct during ventricular shunt. J Neurosurg 24:520-529, 1966 4. Forrest DM, Cooper DGW: Complication of ventriculo-atrial shunts. A review of 455 cases. J Neurnsurg 29:506-512, 1968 5. Fox JL, McCullough DC, Green RC: Effect of cerebrospinal fluid shunts on intracranial pressure and on cerebrospinal fluid dynamics. II. A new technique of pressure measurements, results, and concepts. J Neurol Neurosnrg Psychiatry 36:302-312, 1973 6. Fox JL, Portnoy HD, Shulte RR: Cerebrospinal fluid shunts: an experimental evaluation of flow rates and pressure values in the anti-siphon valve. Snrg Neuroi 1:299-302, 1973 7. Greitz TVB, Grepe AOL, Kalmer MSF et al: Pre- and postoperative evaluation of cerebral blood flow in low-pressure hydrocephalus. J Neurosurg 31:644-651, 1969 199

H. Yamada, M. Tajima and M. Nagaya 8. Hamilton WF, Woodbury RA, Harper HT Jr: Arterial, cerebrospinal and venous pressures in man during cough and strain. Am J Physiol 141:42-50, 1944 9. Illingworth RD: Subdural haematoma after the treatment of chronic hydrocephalus by ventriculocaval shunts. J Neurol Neurosurg Psychiatry 33:95-99, 1970 10. Kaufman B, Weiss MH, Young HF et al: Effects of prolonged cerebrospinal fluid shunting on the skull and brain. J Neurosurg 38:288-297, 1973 11. Kloss JL: Craniosynostosis secondary to ventricular shunt. Am J Dis Child 116:315-317, 1968 12. Lam CR: Intra-abdominal pressure. A critical review and an experimental study. Arch Surg 39:1006-1015, 1939 13. McCullough DC, Fox JL, Curl FD et al: Effect of CSF shunts on intracranial pressure and CSF dynamics, in Harbert JC (ed): Cisternography and Hydrocephalus. Springfield, Ill, Charles C Thomas, 1972, pp. 335-342 14. Natelson SE, Molnar W: Malfunction of ventriculoatrial shunts caused by the circulatory dynamics of coughing. J Neurosurg 36: 283-286, 1972 15. Overholt RH: Intraperitoneal pressure. Arch Surg 22:691-703, 1931 16. Portnoy HD, Shulte RR, Fox JL et al: Antisiphon and reversible occlusion valves for

200

17.

18. 19.

20. 21.

shunting in hydrocephalus and preventing post-shunt subdural hematomas. J Neurosurg 38:729-738, 1973 Rayport M, Reiss J: Hydrodynamic properties of certain shunt assemblies for the treatment of hydrocephalus. Part 1. Report of a case of communicating hydrocephalus with increased cerebrospinal fluid production treated by duplication of shunting device. Part 2. Pressure-flow characteristics of the SpitzHolter, Pudenz-Hyer, and Cordis-Hakim shunt systems. J Neurosurg 30:455-467, 1969 Rushmer RF: The nature of intraperitoneal and intrarectal pressures. Amer J Physiol 147:242-249, 1946 Samuelson S, Long DM, Chou SN: Subdural hematoma as a complication of shunting procedures for normal pressure hydrocephalus. J Neurosurg 37:548-_551, 1972 Shulman K (ed): Workshopin Hydrocephalus. Philadelphia, University of Pennsylvania, 1965 Waga S, Hayashi K, Otsubo K et al: Bilateral chronic subdural hematomas developing after ventriculo-atrial shunt for "normal pressure hydrocephalus." A case report. Brain Nerve (Tokyo) 22:325-329, 1970

Address reprint requests to: Hiroshi Yamada, M.D., 90 Nishisakura-machi, Minami-ku, Nagoya, Japan.

J. Neurosurg. / Volume 42 / February, 1975