- Email: [email protected]
Factors affecting the accuracy of ventricular catheter placement
Journal of Clinical Neuroscience 18 (2011) 485–488
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Factors affecting the accuracy of ventricular catheter placement Kai Rui Wan a,⇑, Jennifer Ah Toy a, Rory Wolfe b, Andrew Danks a,c a
Department of Neurosurgery, Monash Medical Centre, Monash University, Clayton, Victoria, Australia Department of Epidemiology and Preventive Medicine, Monash University, Clayton, Victoria, Australia c Department of Surgery, Monash Medical Centre, Monash University, Clayton, Victoria, Australia b
a r t i c l e
i n f o
Article history: Received 11 June 2010 Accepted 23 June 2010
Keywords: Hydrocephalus Risk factors Shunt placement Ventricular catheter Ventriculoperitoneal shunt
a b s t r a c t Despite technological improvements, ventriculoperitoneal (VP) shunts are still often complicated by malfunction, predominantly with proximal catheter obstruction. There is evidence that accurate placement of the ventricular catheter is significantly related to shunt survival. To identify possible risk factors that might lead to suboptimal shunt placement, we retrospectively reviewed the demographic data and radiological scans of 141 patients who underwent a VP shunt operation from 2005 to 2008 at our institution. We developed and validated a novel scale to assess catheter placement. Almost half (47.9%) of the catheters were ‘‘excellently’’ placed with the entire tip located in the cerebrospinal fluid, and the position of 25% was considered ‘‘good’’. However, 26.8% were less than optimally placed (‘‘poor’’, ‘‘fair’’ or ‘‘moderate’’), with 8.5% (‘‘poor’’) lying entirely outside the ventricular system. Statistical analysis demonstrated that the preoperative size of the ventricles and the age of the patient at shunt insertion were the most important predictors in determining the quality of ventricular catheter placement. Further studies are required to evaluate frameless stereotaxy in optimizing shunt placement in patients with smaller ventricles. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction Ventriculoperitoneal (VP) shunt operations for the treatment of hydrocephalus are one of the most common procedures performed in neurosurgery.1–3 However, shunt survival rates have not increased significantly since its inception in 1952,1,4 despite improvements in both technical equipment4,5 and operative skill.4 About 30% to 50% of shunts fail within the first 12 months,2,6–11 and only 30% to 37% of shunts survive 10 years without revision.6,8 The most common causes of shunt malfunction are obstruction, particularly at the ventricular catheter,2,3,11–21 followed by infection and disconnection.22 Accurate placement of the ventricular catheter11 is one of the most important variables in the longevity of shunt survival.3,15,16,23–25 We retrospectively analyzed 141 patients who underwent a VP shunt operation at Monash Medical Centre, Melbourne, Australia, from 2005 to 2008. Our objective was to determine possible factors that may affect the quality of the ventricular catheter placement. 2. Materials and methods Between 2005 and 2008, we performed a total of 182 VP shunt operations in 141 patients. In all instances, the surgeons used free-
hand techniques employing surface anatomy. Patients whose preoperative and postoperative imaging was unavailable were excluded from the study. Two patients with severe slit ventricles were also excluded because it was not possible to assess shunt placement accurately. Thus, there were 142 eligible procedures in 109 patients. Clinical and demographic data were collected from medical records. The variables recorded included the age of patient at shunt insertion; cause of hydrocephalus; whether the operation was de novo or a shunt revision; and site of the burr hole. The preoperative and postoperative ventricle size were measured both objectively, using the frontal occipital horn width ratio (FOHWR);26–28 and subjectively, dividing into categories of dilation: ‘‘normal’’, ‘‘mild’’, ‘‘moderate’’ or ‘‘severe’’. The intracranial length of the ventricular catheter inserted was also measured. The ventricular catheter location was defined as being in the frontal or occipital horn, atrium of the lateral ventricle; third ventricle, fourth ventricle or intraparenchymal. The total length of the perforated region in the ventricular catheter was measured at 1.5 cm, starting from 0.5 cm from the tip and extending to 2 cm from the tip. These measurements were utilized to formulate a 5-point grading system, based on the length of the tip region that lay within the ventricular cerebrospinal fluid (CSF) (Table 1). 2.1. Statistical methods
⇑ Corresponding author. Postal address: 20 Namly Rise, Singapore 267127, Singapore. Tel.: +65 97854150. E-mail address: [email protected] (K.R. Wan). 0967-5868/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jocn.2010.06.018
The 5-point grading scale outcome (poor/fair/moderate/good/ excellent) was compared between groups of shunts defined by:
486
K.R. Wan et al. / Journal of Clinical Neuroscience 18 (2011) 485–488
Table 1 A 5-point grading system based on the length of the tip within the ventricular cerebrospinal fluid Grading system (5-point)
Length of tip within CSF (cm)
Shunts in series (%)
Excellent Good Moderate Fair Poor
2.0 1.5–<2.0 1.0–<1.5 <1.0 or periventricular Not within the ventricular system
47.9 25.3 15.5 2.8 8.5
Table 3 Effect of age on ventricular catheter placement Age (years)
OR
p value
95% CI
63 >3–18 >18–60 >60
1.0 1.6 1.4 4.2
0.43 0.55 0.01
0.5, 5.1 0.5, 3.7 1.4, 12.4
CI = confidence interval, OR = odds ratio.
CSF = cerebrospinal fluid.
3.3. Placement of burr holes
ventricle size measured subjectively; the FOHWR; site of burr hole placement; age of patient; and primary or revision shunt, with univariate proportional odds cumulative logistic regression models. Multivariate proportional odds cumulative logistic regression models were used to compare shunts defined by length of catheter used for the procedure while adjusting for age of patient, and anatomical approach of surgeon. Inter-rater reliability was assessed using a kappa statistic. Stata 11.0 (StataCorp, College Station, TX, USA) was used for all statistical analyses.
3. Results Table 1 summarises the position of the catheter tip in the 142 shunts in 109 patients. Seventy-five patients with 102 shunts were randomly selected from our patient population and re-assessed to test the classification system for interobserver reliability. Using the 5-point grading system, the two raters (K. Wan, J. Ah Toy) had a 75% exact agreement and a kappa statistic of 0.63 (95% CI: 0.51, 0.75), which represented substantial reliability.
‘‘Frontal’’ and ‘‘parietal’’ burr holes had similar outcomes on the 5-point scale (OR = 1.4, 95% CI: 0.4, 5.1, p value = 0.61) but ‘‘occipital’’ tended towards worse outcomes. The odds ratio of poor placement with an occipital approach as compared to the parietal approach was 0.4 (95% CI: 0.1, 1.1, p value = 0.07). 3.4. Intracranial ventricular catheter length There were similar outcomes for catheters that were 5.5–8 cm in length compared to lengths 65 cm (OR = 1.4, 95% CI: 0.6, 3.0, p value = 0.45) but we observed a trend towards better outcomes when using catheters 8.5–10 cm in length compared with 65 cm (OR = 2.2, 95% CI: 1.0, 5.1, p value = 0.06). These analyses were repeated with adjustment for patient age group and surgeon’s anatomical approach plan. This multivariate analysis confirmed the impression obtained from the univariate analysis. There were again similar outcomes for catheters that were 5.5–8 cm in length compared to lengths 65 cm (multivariate OR = 1.4, 95% CI: 0.6, 3.2, p value = 0.43). There was some evidence of better outcomes being obtained when using catheters >8.5 cm in length compared to 65 cm (OR = 2.7, 95% CI: 1.0, 7.1, p value = 0.05).
3.1. Preoperative ventricle size
3.5. Indications for shunt placement
There was, as expected, better success in catheter placement in patients with larger than normal ventricles. Those patients with a preoperative ventricle size of ‘‘moderate’’ dilation exhibited better success than those with ‘‘normal’’ size (OR = 3.0, 95% CI: 1.3, 7.0, p value = 0.01) and likewise with ‘‘severe’’ dilation versus ‘‘normal’’ (OR = 3.6, 95% CI: 1.4, 8.9, p value = 0.01). There was also evidence of a significant association between a greater category of FOHWR and better outcome (Spearman correlation coefficient = 0.22, p = 0.01). In particular each category had better outcomes than the smallest category (FOHWR 6 0.2) (Table 2).
Ventricle placement outcomes were similar across all causes of hydrocephalus except for normal pressure hydrocephalus (NPH), which had greatly increased odds of better placement (OR 8.4 relative to congenital hydrocephalus; 95% CI 2.1, 32.4, p value = 0.002). This reflected the larger size of the ventricles in the former group.
3.2. Age
3.6. De novo placement compared to shunt revision There was no difference with the accuracy of shunt placement (OR = 0.9, 95% CI 0.5, 1.7, p value = 0.82) when shunts were revised compared to de novo shunt placements. 4. Discussion
There was an association between the (oldest) age group and better outcome, using both grading systems (Table 3).
Table 2 Effect of FOHWR on ventricular catheter placement FOHWR category
OR
p value
95% CI
60.2 0.21–0.30 0.31–0.40 0.41–0.50 >0.5
1.0 3.0 2.8 3.7 5.8
0.02 0.03 0.02 0.03
1.2, 1.1, 1.2, 1.2,
8.0 7.1 11.5 28.3
CI = confidence interval, FOHWR = frontal occipital horn width ratio, OR = odds ratio.
It is a common goal in insertion of a VP shunt to place the ventricular catheter in a sufficient pool of CSF, remote from other structures, as accurate placement of the tip is crucial to shunt survival. However, this goal may be difficult to achieve due to the multiple factors that determine the accuracy of ventricular catheter placement during shunt operations. 4.1. Predictors of optimal shunt placement Ginsberg et al.12 demonstrated that the total resistance CSF drainage and back pressure is inversely proportional to the number of patent holes in the ventricular catheter. Similarly, Tuli et al.11 have shown that ventricular tip location is important in reducing
K.R. Wan et al. / Journal of Clinical Neuroscience 18 (2011) 485–488
shunt malfunction. A ventricular catheter tip surrounded by CSF decreased the risk of shunt failure to one-fifth, whereas a catheter tip touching the brain decreased the risk to one-third, compared with a catheter tip surrounded by brain (residing in a slit ventricle). This gives credence to the notion that if the ventricular catheter can be kept in a pool of CSF, remote from brain structures, the ventricular catheter is less likely to become occluded. Analysis of our data has shown that the preoperative ventricle size is significant in its relation to the accuracy of shunt placement. This concurs with common sense. Caldarelli’s study7 had also demonstrated previously that the number of complications, both mechanical and infective, is directly proportional to the size of the cerebral ventricles. Larger ventricles are more likely to have optimally placed catheters, which decreases the risk of proximal obstruction, and hence increases shunt survival. Age is a known statistically significant predictor of first episode shunt failure,9,29 which could be due to its relationship with the quality of shunt placement, as revealed in our results. Nonetheless, other factors may contribute to this observation: (i) the choroid plexus is relatively smaller in elderly patients;15 and (ii) the reduction of ventricular size post shunting is significantly less, due to smaller centripetal forces to oppose ventricle dilatation.30 In comparison, younger patients have relatively smaller ventricular size at shunt insertion and are at increased risk of infection, thus leading to a higher malfunction rate.9,19,30 Furthermore, there are effects of growth leading to possible late displacement of catheters. There is no consensus among neurosurgeons regarding the best approach for shunt placements and there is conflicting evidence regarding the different methods.14,15,19,20,31,32 The frontal approach has the shortest intracerebral route15 with more consistent anatomical landmarks.15,33,34 However, ventricular catheter shortening and burr hole displacement with growth is more pronounced with this approach.18,19 In addition, there are aesthetic issues due to the need to shave the scalp in the frontal area.32 The atrium is usually the most dilated part of the lateral ventricle and therefore potentially the last part of the chamber to collapse after CSF drainage.20,33 Thus, the parietal approach would be especially useful for patients with congenital or tumoral hydrocephalus, who tend to undergo greater reduction in ventricular size following shunting.30 The occipital and parietal approaches avoid a second scalp incision for tunneling,33 but the catheters tend to migrate with head growth, which increases the chance of obstruction by choroid plexus.32 We have shown similar results in the precision of shunt placement between the frontal and parietal approaches. However, there is evidence of worse outcomes with the occipital approach, which coincides with Lind et al.’s study,25 which demonstrates a smaller margin of error for trajectories.33 There is a general discrepancy in the literature regarding the relationship between the cause of hydrocephalus and shunt survival. Many studies report that there is no relationship between the cause of the hydrocephalus and the incidence of shunt complications,1,10,29 whereas others have shown increased shunt survival in post-meningitis hydrocephalus;1 after intracranial haemorrhage;35 or for NPH.15 The results of this study have not shown any significant difference in the accuracy of catheter placement between the different causes of hydrocephalus, except with NPH, which could be due to the increased ventricle size and age of patients with this condition. Shunt revisions have been associated with a higher complication rate,10 including intraoperative complications17 and have been independently associated with significantly greater odds of proximal obstruction,13,29,36 leading to repeated revisions.17 However, we have not shown any statistically significant difference between the accuracy of placement for a new or a revised shunt. Therefore, this does not appear to be a contributing factor to these problems.
487
In designing this study, we considered that surgical techniques using shorter catheters might lead to more accurate catheter placement. There was no evidence to support this postulate. In contrast to our initial hypothesis, longer catheters, those greater than 8.5 cm, appear better placed even after correcting for the patient’s age and method of approach. 4.2. Confounding factors We acknowledge that the variable timing of post-operative radiological scans in our patient series could affect the assessment of shunt placement. However, the time required for ventricles to return to their minimum size after shunting is still poorly defined.30 4.3. Future directions The VP shunts in our study were placed freehand, with a catheter tip malpositioning rate of 26.8% (if we considered the poor, fair and moderate placements as unacceptable); slightly lower than the 30% reported in other studies.34,37 Direct visualization of the catheter is not possible with freehand placement, which is associated with difficulty in gauging optimum catheter length and angle.2,25,31 As a result, accurate catheter placement can be a challenge, especially in patients with small or abnormal ventricular anatomy.17,24 In children, ongoing changes of ventricular size and shape can be substantial, adding to the difficulty. In this study, in the patients with smaller ventricles (FOHWR <0.3), suboptimal positioning occurred in 35% of catheter placements. In the modern era, it may be possible to improve these results. We propose further studies to determine the efficacy of frameless stereotaxy for shunt placement in small/slit ventricles (i.e. FOHWR <0.3) to avoid the morbidity of suboptimally placed catheters.24 Previous trials using stereotactic technique to target small or slit ventricles3,17,36,38,39 have demonstrated accurate placement of all the ventricular catheters with a single pass. None of the trials reported proximal catheter obstruction in the follow-up period, although they each had a small patient population, of less than 17 patients. There are disadvantages to frameless neuronavigation, which include the prolongation of operation time and dependence on expensive technology.17,38 However, considering the high revision rates for patients with small ventricular sizes, this technique may lead to better outcomes, and be cost-effective.38 5. Conclusion Proximal catheter obstruction is the most common complication of insertion of a VP shunt, and accurate placement of the catheter tip is highly important for shunt survival. Among the many factors that may affect optimal ventricular catheter placement, this study has determined that a smaller preoperative ventricular size and younger age are the most important predictors. Suboptimal catheter placement occurred in 27% of patients. In patients with normal-sized ventricles or mildly dilated ventricles (FOHWR <0.3), there was a 35% suboptimal catheter placement rate. In the modern era, we should be able to achieve better results. We propose further studies to evaluate frameless stereotaxy for VP shunt placement in patients with difficult ventricular anatomy (e.g. small or slit ventricles and the pediatric population). Acknowledgements The authors would like to thank Dr Kristy Scandrett, Dr Leon Lai and the neurosurgery team at Monash Medical Centre for all their help and support in this project.
488
K.R. Wan et al. / Journal of Clinical Neuroscience 18 (2011) 485–488
References 1. Ferguson SD, Michael N, Frim DM. Observations regarding failure of cerebrospinal fluid shunts early after implantation. Neurosurg Focus 2007;22:E7. 2. Browd SR, Gottfried ON, Kestle JR. Failure of cerebrospinal fluid shunts: part I: obstruction and mechanical failure. Pediatr Neurol 2006;34:83–92. 3. Kim YB, Lee JW, Lee KS, et al. Image-guided placement of ventricular shunt catheter. J Clin Neurosci 2006;13:50–4. 4. Stein SC, Guo WS. Have we made progress in preventing shunt failure? A critical analysis. J Neurosurg Pediatr 2008;1:40–7. 5. Meier U, Konig A, Miethke C. Predictors of outcome in patients with normalpressure hydrocephalus. Eur Neurol 2004;51:59. 6. Bergsneider M, Egnor MR, Johnston M, et al. What we don’t (but should) know about hydrocephalus. J Neurosurg 2006;104:157–9. 7. Caldarelli M, Di Rocco C, La Marca F. Shunt complications in the first postoperative year in children with meningomyelocele. Childs Nerv Syst 1996;12:748–54. 8. Tuli S, Drake J, Lawless J, et al. Risk factors for repeated cerebrospinal shunt failures in pediatric patients with hydrocephalus. J Neurosurg 2000;92: 31–8. 9. Wu Y, Green N, Wrensch M, et al. Ventriculoperitoneal shunt complications in California: 1990–2000. Congress Neurol Surg 2007;61:557–63. 10. Griebel R, Khan M, Tan L. CSF shunt complications: an analysis of contributory factors. Childs Nerv Syst 1985;1:77–80. 11. Tuli S, O’Hayon B, Drake J, et al. Change in ventricular size and effect of ventricular catheter placement in pediatric patients with shunted hydrocephalus. Neurosurgery 1999;45:1329. 12. Ginsberg HJ, Sum A, Drake JM, et al. Ventriculoperitoneal shunt flow dependency on the number of patent holes in a ventricular catheter. Pediatr Neurosurg 2000;33:7. 13. Puca A, Anile C, Maira G, et al. Cerebrospinal fluid shunting for hydrocephalus in the adult: factors related to shunt revision. Neurosurgery 1991;29:822–6. 14. Choudhury AR. CT measurement of the ventricular catheter length in shunt surgery. Br J Neurosurg 1993;7:541–4. 15. Dickerman RD, McConathy WJ, Morgan J, et al. Failure rate of frontal versus parietal approaches for proximal catheter placement in ventriculoperitoneal shunts: revisited. J Clin Neurosci 2005;12:781–3. 16. Kang JK, Lee IW. Long-term follow-up of shunting therapy. Childs Nerv Syst 1999;15:711–7. 17. Gil Z, Siomin V, Beni-Adani L, et al. Ventricular catheter placement in children with hydrocephalus and small ventricles: the use of a frameless neuronavigation system. Childs Nerv Syst 2002;18:26–9. 18. Nakahara K, Shimizu S, Utsuki S, et al. Shortening of ventricular shunt catheter associated with cranial growth: effect of the frontal and pareito-occipital access route on long term shunt patency. Childs Nerv Syst 2008;25:91–4. 19. Sainte-Rose C, Piatt JH, Pierre-Kahn A, et al. Mechanical complications in shunts. Pediatr Neurosurg 1991;17:2–9. 20. Sainte-Rose C. Shunt obstruction: a preventable complication? Pediatr Neurosurg 1993;19:156–64.
21. Collins P, Hockley A, Woollam DH. Surface ultrastructure of tissues occluding ventricular catheters. J Neurosurg 1978;48:609. 22. Lazareff J, Peacock W, Holly L, et al. Multiple shunt failures: an analysis of relevant factors. Childs Nerv Syst 1998;14:271–5. 23. Torrez-Corzo J, Rodriguez-Della V, Chalita-Williams JC, et al. Endoscopic management of brainstem injury due to ventriculoperitoneal shunt placement. Childs Nerv Syst 2009;25:627–30. 24. Azeem SS, Origitano TC. Ventricular catheter placement with a frameless neuronavigational system: a 1-year experience. Neurosurgery 2007;60:243–8. 25. Lind CRP, Tsai AMC, Lind CJ, et al. Ventricular catheter placement accuracy in non-stereotactic shunt surgery for hydrocephalus. J Clin Neurosci 2009;16:918–20. 26. Kulkarni AV, Drake JM, Armstrong DC, et al. Measurement of ventricular size: reliability of the frontal and occipital horn ratio compared to subjective assessment. Pediatr Neurosurg 1999;31:65–70. 27. Jamous M, Sood S, Kumar R, et al. Frontal and occipital horn width ratio for the evaluation of small and asymmetrical ventricles. Pediatr Neurosurg 2003;39:17–21. 28. O’Hayon BB, Drake JM, Ossip MG, et al. Frontal and occipital horn ratio: a linear estimate of ventricular size for multiple imaging modalities in pediatric hydrocephalus. Pediatr Neurosurg 1998;29:245–9. 29. McGirt MJ, Leveque JC, Wellons 3rd JC, et al. Cerebrospinal fluid shunt survival and etiology of failures: a seven-year institutional experience. Pediatr Neurol 2002;36:248–55. 30. Cadoso ER, Del Bigio MR. Age-related changes of cerebral ventricular size. Acta Neurochir 1989;97:135–8. 31. Lind CRP, Correia JA, Law AJJ, et al. A survey of surgical techniques for catheterising the cerebral lateral ventricles. J Clin Neurosci 2008;15:886–90. 32. Bierbrauer KS, Storrs BB, McLone DG, et al. A prospective randomized study of shunt function and infections as a function of shunt placement. Pediatr Neurosurg 1991;16:287–91. 33. Lind CR, Tsai AM, Law AJ, et al. Ventricular catheter trajectories from traditional shunt approaches: a morphometric study in adults with hydrocephalus. J Neurosurg 2008;108:930–3. 34. Strowitzi M, Komenda Y, Eymann R, et al. Accuracy of ultrasound-guided puncture of the ventricular system. Childs Nerv Syst 2008;24:65–9. 35. Hoh B, Lang SS, Ortiz MV, et al. Lower incidence of reoperation with longer shunt survival with adult ventriculoperitoneal shunts placed for hemorrhagerelated hydrocephalus. Neurosurgery 2008;63:70–5. 36. Woodworth GF, McGirt MJ, Elfert P, et al. Frameless stereotactic ventricular shunt placement for idiopathic intracranial hypertension, stereotactic and functional. Stereotact Funct Neurosurg 2005;83:12–6. 37. Missori P, Artizzu S, Salvati M. Immediate postoperative CT to assess the correct positioning of a ventricular catheter. Br J Neurosurg 2000;14:44. 38. Tulipan N, Lavin PJM, Copeland M. Stereotactic ventriculoperitoneal shunt for idiopathic intracranial hypertension: technical note. Neurosurgery 1998;43:175. 39. Woerdeman PA, Williems PW, Han KS, et al. Framless stereotactic placement of ventriculoperitoneal shunts in undersized ventricles: a simple modification to free-hand procedures. Br J Neurosurg 2005;19:484–7.
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical Study
Factors affecting the accuracy of ventricular catheter placement Kai Rui Wan a,⇑, Jennifer Ah Toy a, Rory Wolfe b, Andrew Danks a,c a
Department of Neurosurgery, Monash Medical Centre, Monash University, Clayton, Victoria, Australia Department of Epidemiology and Preventive Medicine, Monash University, Clayton, Victoria, Australia c Department of Surgery, Monash Medical Centre, Monash University, Clayton, Victoria, Australia b
a r t i c l e
i n f o
Article history: Received 11 June 2010 Accepted 23 June 2010
Keywords: Hydrocephalus Risk factors Shunt placement Ventricular catheter Ventriculoperitoneal shunt
a b s t r a c t Despite technological improvements, ventriculoperitoneal (VP) shunts are still often complicated by malfunction, predominantly with proximal catheter obstruction. There is evidence that accurate placement of the ventricular catheter is significantly related to shunt survival. To identify possible risk factors that might lead to suboptimal shunt placement, we retrospectively reviewed the demographic data and radiological scans of 141 patients who underwent a VP shunt operation from 2005 to 2008 at our institution. We developed and validated a novel scale to assess catheter placement. Almost half (47.9%) of the catheters were ‘‘excellently’’ placed with the entire tip located in the cerebrospinal fluid, and the position of 25% was considered ‘‘good’’. However, 26.8% were less than optimally placed (‘‘poor’’, ‘‘fair’’ or ‘‘moderate’’), with 8.5% (‘‘poor’’) lying entirely outside the ventricular system. Statistical analysis demonstrated that the preoperative size of the ventricles and the age of the patient at shunt insertion were the most important predictors in determining the quality of ventricular catheter placement. Further studies are required to evaluate frameless stereotaxy in optimizing shunt placement in patients with smaller ventricles. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction Ventriculoperitoneal (VP) shunt operations for the treatment of hydrocephalus are one of the most common procedures performed in neurosurgery.1–3 However, shunt survival rates have not increased significantly since its inception in 1952,1,4 despite improvements in both technical equipment4,5 and operative skill.4 About 30% to 50% of shunts fail within the first 12 months,2,6–11 and only 30% to 37% of shunts survive 10 years without revision.6,8 The most common causes of shunt malfunction are obstruction, particularly at the ventricular catheter,2,3,11–21 followed by infection and disconnection.22 Accurate placement of the ventricular catheter11 is one of the most important variables in the longevity of shunt survival.3,15,16,23–25 We retrospectively analyzed 141 patients who underwent a VP shunt operation at Monash Medical Centre, Melbourne, Australia, from 2005 to 2008. Our objective was to determine possible factors that may affect the quality of the ventricular catheter placement. 2. Materials and methods Between 2005 and 2008, we performed a total of 182 VP shunt operations in 141 patients. In all instances, the surgeons used free-
hand techniques employing surface anatomy. Patients whose preoperative and postoperative imaging was unavailable were excluded from the study. Two patients with severe slit ventricles were also excluded because it was not possible to assess shunt placement accurately. Thus, there were 142 eligible procedures in 109 patients. Clinical and demographic data were collected from medical records. The variables recorded included the age of patient at shunt insertion; cause of hydrocephalus; whether the operation was de novo or a shunt revision; and site of the burr hole. The preoperative and postoperative ventricle size were measured both objectively, using the frontal occipital horn width ratio (FOHWR);26–28 and subjectively, dividing into categories of dilation: ‘‘normal’’, ‘‘mild’’, ‘‘moderate’’ or ‘‘severe’’. The intracranial length of the ventricular catheter inserted was also measured. The ventricular catheter location was defined as being in the frontal or occipital horn, atrium of the lateral ventricle; third ventricle, fourth ventricle or intraparenchymal. The total length of the perforated region in the ventricular catheter was measured at 1.5 cm, starting from 0.5 cm from the tip and extending to 2 cm from the tip. These measurements were utilized to formulate a 5-point grading system, based on the length of the tip region that lay within the ventricular cerebrospinal fluid (CSF) (Table 1). 2.1. Statistical methods
⇑ Corresponding author. Postal address: 20 Namly Rise, Singapore 267127, Singapore. Tel.: +65 97854150. E-mail address: [email protected] (K.R. Wan). 0967-5868/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jocn.2010.06.018
The 5-point grading scale outcome (poor/fair/moderate/good/ excellent) was compared between groups of shunts defined by:
486
K.R. Wan et al. / Journal of Clinical Neuroscience 18 (2011) 485–488
Table 1 A 5-point grading system based on the length of the tip within the ventricular cerebrospinal fluid Grading system (5-point)
Length of tip within CSF (cm)
Shunts in series (%)
Excellent Good Moderate Fair Poor
2.0 1.5–<2.0 1.0–<1.5 <1.0 or periventricular Not within the ventricular system
47.9 25.3 15.5 2.8 8.5
Table 3 Effect of age on ventricular catheter placement Age (years)
OR
p value
95% CI
63 >3–18 >18–60 >60
1.0 1.6 1.4 4.2
0.43 0.55 0.01
0.5, 5.1 0.5, 3.7 1.4, 12.4
CI = confidence interval, OR = odds ratio.
CSF = cerebrospinal fluid.
3.3. Placement of burr holes
ventricle size measured subjectively; the FOHWR; site of burr hole placement; age of patient; and primary or revision shunt, with univariate proportional odds cumulative logistic regression models. Multivariate proportional odds cumulative logistic regression models were used to compare shunts defined by length of catheter used for the procedure while adjusting for age of patient, and anatomical approach of surgeon. Inter-rater reliability was assessed using a kappa statistic. Stata 11.0 (StataCorp, College Station, TX, USA) was used for all statistical analyses.
3. Results Table 1 summarises the position of the catheter tip in the 142 shunts in 109 patients. Seventy-five patients with 102 shunts were randomly selected from our patient population and re-assessed to test the classification system for interobserver reliability. Using the 5-point grading system, the two raters (K. Wan, J. Ah Toy) had a 75% exact agreement and a kappa statistic of 0.63 (95% CI: 0.51, 0.75), which represented substantial reliability.
‘‘Frontal’’ and ‘‘parietal’’ burr holes had similar outcomes on the 5-point scale (OR = 1.4, 95% CI: 0.4, 5.1, p value = 0.61) but ‘‘occipital’’ tended towards worse outcomes. The odds ratio of poor placement with an occipital approach as compared to the parietal approach was 0.4 (95% CI: 0.1, 1.1, p value = 0.07). 3.4. Intracranial ventricular catheter length There were similar outcomes for catheters that were 5.5–8 cm in length compared to lengths 65 cm (OR = 1.4, 95% CI: 0.6, 3.0, p value = 0.45) but we observed a trend towards better outcomes when using catheters 8.5–10 cm in length compared with 65 cm (OR = 2.2, 95% CI: 1.0, 5.1, p value = 0.06). These analyses were repeated with adjustment for patient age group and surgeon’s anatomical approach plan. This multivariate analysis confirmed the impression obtained from the univariate analysis. There were again similar outcomes for catheters that were 5.5–8 cm in length compared to lengths 65 cm (multivariate OR = 1.4, 95% CI: 0.6, 3.2, p value = 0.43). There was some evidence of better outcomes being obtained when using catheters >8.5 cm in length compared to 65 cm (OR = 2.7, 95% CI: 1.0, 7.1, p value = 0.05).
3.1. Preoperative ventricle size
3.5. Indications for shunt placement
There was, as expected, better success in catheter placement in patients with larger than normal ventricles. Those patients with a preoperative ventricle size of ‘‘moderate’’ dilation exhibited better success than those with ‘‘normal’’ size (OR = 3.0, 95% CI: 1.3, 7.0, p value = 0.01) and likewise with ‘‘severe’’ dilation versus ‘‘normal’’ (OR = 3.6, 95% CI: 1.4, 8.9, p value = 0.01). There was also evidence of a significant association between a greater category of FOHWR and better outcome (Spearman correlation coefficient = 0.22, p = 0.01). In particular each category had better outcomes than the smallest category (FOHWR 6 0.2) (Table 2).
Ventricle placement outcomes were similar across all causes of hydrocephalus except for normal pressure hydrocephalus (NPH), which had greatly increased odds of better placement (OR 8.4 relative to congenital hydrocephalus; 95% CI 2.1, 32.4, p value = 0.002). This reflected the larger size of the ventricles in the former group.
3.2. Age
3.6. De novo placement compared to shunt revision There was no difference with the accuracy of shunt placement (OR = 0.9, 95% CI 0.5, 1.7, p value = 0.82) when shunts were revised compared to de novo shunt placements. 4. Discussion
There was an association between the (oldest) age group and better outcome, using both grading systems (Table 3).
Table 2 Effect of FOHWR on ventricular catheter placement FOHWR category
OR
p value
95% CI
60.2 0.21–0.30 0.31–0.40 0.41–0.50 >0.5
1.0 3.0 2.8 3.7 5.8
0.02 0.03 0.02 0.03
1.2, 1.1, 1.2, 1.2,
8.0 7.1 11.5 28.3
CI = confidence interval, FOHWR = frontal occipital horn width ratio, OR = odds ratio.
It is a common goal in insertion of a VP shunt to place the ventricular catheter in a sufficient pool of CSF, remote from other structures, as accurate placement of the tip is crucial to shunt survival. However, this goal may be difficult to achieve due to the multiple factors that determine the accuracy of ventricular catheter placement during shunt operations. 4.1. Predictors of optimal shunt placement Ginsberg et al.12 demonstrated that the total resistance CSF drainage and back pressure is inversely proportional to the number of patent holes in the ventricular catheter. Similarly, Tuli et al.11 have shown that ventricular tip location is important in reducing
K.R. Wan et al. / Journal of Clinical Neuroscience 18 (2011) 485–488
shunt malfunction. A ventricular catheter tip surrounded by CSF decreased the risk of shunt failure to one-fifth, whereas a catheter tip touching the brain decreased the risk to one-third, compared with a catheter tip surrounded by brain (residing in a slit ventricle). This gives credence to the notion that if the ventricular catheter can be kept in a pool of CSF, remote from brain structures, the ventricular catheter is less likely to become occluded. Analysis of our data has shown that the preoperative ventricle size is significant in its relation to the accuracy of shunt placement. This concurs with common sense. Caldarelli’s study7 had also demonstrated previously that the number of complications, both mechanical and infective, is directly proportional to the size of the cerebral ventricles. Larger ventricles are more likely to have optimally placed catheters, which decreases the risk of proximal obstruction, and hence increases shunt survival. Age is a known statistically significant predictor of first episode shunt failure,9,29 which could be due to its relationship with the quality of shunt placement, as revealed in our results. Nonetheless, other factors may contribute to this observation: (i) the choroid plexus is relatively smaller in elderly patients;15 and (ii) the reduction of ventricular size post shunting is significantly less, due to smaller centripetal forces to oppose ventricle dilatation.30 In comparison, younger patients have relatively smaller ventricular size at shunt insertion and are at increased risk of infection, thus leading to a higher malfunction rate.9,19,30 Furthermore, there are effects of growth leading to possible late displacement of catheters. There is no consensus among neurosurgeons regarding the best approach for shunt placements and there is conflicting evidence regarding the different methods.14,15,19,20,31,32 The frontal approach has the shortest intracerebral route15 with more consistent anatomical landmarks.15,33,34 However, ventricular catheter shortening and burr hole displacement with growth is more pronounced with this approach.18,19 In addition, there are aesthetic issues due to the need to shave the scalp in the frontal area.32 The atrium is usually the most dilated part of the lateral ventricle and therefore potentially the last part of the chamber to collapse after CSF drainage.20,33 Thus, the parietal approach would be especially useful for patients with congenital or tumoral hydrocephalus, who tend to undergo greater reduction in ventricular size following shunting.30 The occipital and parietal approaches avoid a second scalp incision for tunneling,33 but the catheters tend to migrate with head growth, which increases the chance of obstruction by choroid plexus.32 We have shown similar results in the precision of shunt placement between the frontal and parietal approaches. However, there is evidence of worse outcomes with the occipital approach, which coincides with Lind et al.’s study,25 which demonstrates a smaller margin of error for trajectories.33 There is a general discrepancy in the literature regarding the relationship between the cause of hydrocephalus and shunt survival. Many studies report that there is no relationship between the cause of the hydrocephalus and the incidence of shunt complications,1,10,29 whereas others have shown increased shunt survival in post-meningitis hydrocephalus;1 after intracranial haemorrhage;35 or for NPH.15 The results of this study have not shown any significant difference in the accuracy of catheter placement between the different causes of hydrocephalus, except with NPH, which could be due to the increased ventricle size and age of patients with this condition. Shunt revisions have been associated with a higher complication rate,10 including intraoperative complications17 and have been independently associated with significantly greater odds of proximal obstruction,13,29,36 leading to repeated revisions.17 However, we have not shown any statistically significant difference between the accuracy of placement for a new or a revised shunt. Therefore, this does not appear to be a contributing factor to these problems.
487
In designing this study, we considered that surgical techniques using shorter catheters might lead to more accurate catheter placement. There was no evidence to support this postulate. In contrast to our initial hypothesis, longer catheters, those greater than 8.5 cm, appear better placed even after correcting for the patient’s age and method of approach. 4.2. Confounding factors We acknowledge that the variable timing of post-operative radiological scans in our patient series could affect the assessment of shunt placement. However, the time required for ventricles to return to their minimum size after shunting is still poorly defined.30 4.3. Future directions The VP shunts in our study were placed freehand, with a catheter tip malpositioning rate of 26.8% (if we considered the poor, fair and moderate placements as unacceptable); slightly lower than the 30% reported in other studies.34,37 Direct visualization of the catheter is not possible with freehand placement, which is associated with difficulty in gauging optimum catheter length and angle.2,25,31 As a result, accurate catheter placement can be a challenge, especially in patients with small or abnormal ventricular anatomy.17,24 In children, ongoing changes of ventricular size and shape can be substantial, adding to the difficulty. In this study, in the patients with smaller ventricles (FOHWR <0.3), suboptimal positioning occurred in 35% of catheter placements. In the modern era, it may be possible to improve these results. We propose further studies to determine the efficacy of frameless stereotaxy for shunt placement in small/slit ventricles (i.e. FOHWR <0.3) to avoid the morbidity of suboptimally placed catheters.24 Previous trials using stereotactic technique to target small or slit ventricles3,17,36,38,39 have demonstrated accurate placement of all the ventricular catheters with a single pass. None of the trials reported proximal catheter obstruction in the follow-up period, although they each had a small patient population, of less than 17 patients. There are disadvantages to frameless neuronavigation, which include the prolongation of operation time and dependence on expensive technology.17,38 However, considering the high revision rates for patients with small ventricular sizes, this technique may lead to better outcomes, and be cost-effective.38 5. Conclusion Proximal catheter obstruction is the most common complication of insertion of a VP shunt, and accurate placement of the catheter tip is highly important for shunt survival. Among the many factors that may affect optimal ventricular catheter placement, this study has determined that a smaller preoperative ventricular size and younger age are the most important predictors. Suboptimal catheter placement occurred in 27% of patients. In patients with normal-sized ventricles or mildly dilated ventricles (FOHWR <0.3), there was a 35% suboptimal catheter placement rate. In the modern era, we should be able to achieve better results. We propose further studies to evaluate frameless stereotaxy for VP shunt placement in patients with difficult ventricular anatomy (e.g. small or slit ventricles and the pediatric population). Acknowledgements The authors would like to thank Dr Kristy Scandrett, Dr Leon Lai and the neurosurgery team at Monash Medical Centre for all their help and support in this project.
488
K.R. Wan et al. / Journal of Clinical Neuroscience 18 (2011) 485–488
References 1. Ferguson SD, Michael N, Frim DM. Observations regarding failure of cerebrospinal fluid shunts early after implantation. Neurosurg Focus 2007;22:E7. 2. Browd SR, Gottfried ON, Kestle JR. Failure of cerebrospinal fluid shunts: part I: obstruction and mechanical failure. Pediatr Neurol 2006;34:83–92. 3. Kim YB, Lee JW, Lee KS, et al. Image-guided placement of ventricular shunt catheter. J Clin Neurosci 2006;13:50–4. 4. Stein SC, Guo WS. Have we made progress in preventing shunt failure? A critical analysis. J Neurosurg Pediatr 2008;1:40–7. 5. Meier U, Konig A, Miethke C. Predictors of outcome in patients with normalpressure hydrocephalus. Eur Neurol 2004;51:59. 6. Bergsneider M, Egnor MR, Johnston M, et al. What we don’t (but should) know about hydrocephalus. J Neurosurg 2006;104:157–9. 7. Caldarelli M, Di Rocco C, La Marca F. Shunt complications in the first postoperative year in children with meningomyelocele. Childs Nerv Syst 1996;12:748–54. 8. Tuli S, Drake J, Lawless J, et al. Risk factors for repeated cerebrospinal shunt failures in pediatric patients with hydrocephalus. J Neurosurg 2000;92: 31–8. 9. Wu Y, Green N, Wrensch M, et al. Ventriculoperitoneal shunt complications in California: 1990–2000. Congress Neurol Surg 2007;61:557–63. 10. Griebel R, Khan M, Tan L. CSF shunt complications: an analysis of contributory factors. Childs Nerv Syst 1985;1:77–80. 11. Tuli S, O’Hayon B, Drake J, et al. Change in ventricular size and effect of ventricular catheter placement in pediatric patients with shunted hydrocephalus. Neurosurgery 1999;45:1329. 12. Ginsberg HJ, Sum A, Drake JM, et al. Ventriculoperitoneal shunt flow dependency on the number of patent holes in a ventricular catheter. Pediatr Neurosurg 2000;33:7. 13. Puca A, Anile C, Maira G, et al. Cerebrospinal fluid shunting for hydrocephalus in the adult: factors related to shunt revision. Neurosurgery 1991;29:822–6. 14. Choudhury AR. CT measurement of the ventricular catheter length in shunt surgery. Br J Neurosurg 1993;7:541–4. 15. Dickerman RD, McConathy WJ, Morgan J, et al. Failure rate of frontal versus parietal approaches for proximal catheter placement in ventriculoperitoneal shunts: revisited. J Clin Neurosci 2005;12:781–3. 16. Kang JK, Lee IW. Long-term follow-up of shunting therapy. Childs Nerv Syst 1999;15:711–7. 17. Gil Z, Siomin V, Beni-Adani L, et al. Ventricular catheter placement in children with hydrocephalus and small ventricles: the use of a frameless neuronavigation system. Childs Nerv Syst 2002;18:26–9. 18. Nakahara K, Shimizu S, Utsuki S, et al. Shortening of ventricular shunt catheter associated with cranial growth: effect of the frontal and pareito-occipital access route on long term shunt patency. Childs Nerv Syst 2008;25:91–4. 19. Sainte-Rose C, Piatt JH, Pierre-Kahn A, et al. Mechanical complications in shunts. Pediatr Neurosurg 1991;17:2–9. 20. Sainte-Rose C. Shunt obstruction: a preventable complication? Pediatr Neurosurg 1993;19:156–64.
21. Collins P, Hockley A, Woollam DH. Surface ultrastructure of tissues occluding ventricular catheters. J Neurosurg 1978;48:609. 22. Lazareff J, Peacock W, Holly L, et al. Multiple shunt failures: an analysis of relevant factors. Childs Nerv Syst 1998;14:271–5. 23. Torrez-Corzo J, Rodriguez-Della V, Chalita-Williams JC, et al. Endoscopic management of brainstem injury due to ventriculoperitoneal shunt placement. Childs Nerv Syst 2009;25:627–30. 24. Azeem SS, Origitano TC. Ventricular catheter placement with a frameless neuronavigational system: a 1-year experience. Neurosurgery 2007;60:243–8. 25. Lind CRP, Tsai AMC, Lind CJ, et al. Ventricular catheter placement accuracy in non-stereotactic shunt surgery for hydrocephalus. J Clin Neurosci 2009;16:918–20. 26. Kulkarni AV, Drake JM, Armstrong DC, et al. Measurement of ventricular size: reliability of the frontal and occipital horn ratio compared to subjective assessment. Pediatr Neurosurg 1999;31:65–70. 27. Jamous M, Sood S, Kumar R, et al. Frontal and occipital horn width ratio for the evaluation of small and asymmetrical ventricles. Pediatr Neurosurg 2003;39:17–21. 28. O’Hayon BB, Drake JM, Ossip MG, et al. Frontal and occipital horn ratio: a linear estimate of ventricular size for multiple imaging modalities in pediatric hydrocephalus. Pediatr Neurosurg 1998;29:245–9. 29. McGirt MJ, Leveque JC, Wellons 3rd JC, et al. Cerebrospinal fluid shunt survival and etiology of failures: a seven-year institutional experience. Pediatr Neurol 2002;36:248–55. 30. Cadoso ER, Del Bigio MR. Age-related changes of cerebral ventricular size. Acta Neurochir 1989;97:135–8. 31. Lind CRP, Correia JA, Law AJJ, et al. A survey of surgical techniques for catheterising the cerebral lateral ventricles. J Clin Neurosci 2008;15:886–90. 32. Bierbrauer KS, Storrs BB, McLone DG, et al. A prospective randomized study of shunt function and infections as a function of shunt placement. Pediatr Neurosurg 1991;16:287–91. 33. Lind CR, Tsai AM, Law AJ, et al. Ventricular catheter trajectories from traditional shunt approaches: a morphometric study in adults with hydrocephalus. J Neurosurg 2008;108:930–3. 34. Strowitzi M, Komenda Y, Eymann R, et al. Accuracy of ultrasound-guided puncture of the ventricular system. Childs Nerv Syst 2008;24:65–9. 35. Hoh B, Lang SS, Ortiz MV, et al. Lower incidence of reoperation with longer shunt survival with adult ventriculoperitoneal shunts placed for hemorrhagerelated hydrocephalus. Neurosurgery 2008;63:70–5. 36. Woodworth GF, McGirt MJ, Elfert P, et al. Frameless stereotactic ventricular shunt placement for idiopathic intracranial hypertension, stereotactic and functional. Stereotact Funct Neurosurg 2005;83:12–6. 37. Missori P, Artizzu S, Salvati M. Immediate postoperative CT to assess the correct positioning of a ventricular catheter. Br J Neurosurg 2000;14:44. 38. Tulipan N, Lavin PJM, Copeland M. Stereotactic ventriculoperitoneal shunt for idiopathic intracranial hypertension: technical note. Neurosurgery 1998;43:175. 39. Woerdeman PA, Williems PW, Han KS, et al. Framless stereotactic placement of ventriculoperitoneal shunts in undersized ventricles: a simple modification to free-hand procedures. Br J Neurosurg 2005;19:484–7.