Abstract

BackgroundA substantial proportion of patients who undergo decompressive craniectomy develop hydrocephalus (HCP), necessitating both cranioplasty (CP) and cerebrospinal fluid (CSF) shunting procedures. There is wide variation in the timing and sequence of these operations. We aim to define the complication rates and types of each treatment algorithm in patients requiring both CP and CSF shunting in an effort to identify the optimal sequence of procedures.

MethodsA systematic literature review and meta-analysis were performed to assess complication rates for preCP shunting, simultaneous CP-shunting, and post-CP shunting.

ResultsThere is a significant difference in complication rates between shunt placement before CP (35% demonstrated complications, confidence interval (CI) 95%, 30.9–39.1, simultaneous ventriculoperitoneal shunt (VPS) and CP (30.8%, CI 95%, 28.4–33.3), and shunt post-CP (24.4%, CI 95%, 20.5–28.2), with the shunt post-CP cohort demonstrating the lowest percentage of all complication subtypes. There is a trend toward decreased odds of complication when CP is performed before VPS.

ConclusionThis information should further strengthen existing recommendations that, whenever possible, CP is performed before VPS to potentially allow for CSF dynamics to normalize. The study population does demonstrate heterogeneity. Therefore, considerations should be made based on the clinical picture and a patient’s course.

Keywords: Complications, Cranioplasty, Posttraumatic hydrocephalus, Shunt

INTRODUCTION

Many patients who undergo decompressive craniectomy (DC) develop hydrocephalus (HCP), necessitating both cranioplasty (CP) and cerebrospinal fluid (CSF) shunting procedures.[ 2 , 16 ] The clinical appearance of HCP varies, arising both before and after CP, complicating the development of routine treatment algorithms or predictive modeling. Historically, HCP was thought to be a consequence of a patient’s initial injury, but recent studies argue that it may develop from alterations in CSF dynamics, which may be attributed to CP.[ 9 ] Uncertainty around the etiology of HCP and the unpredictable clinical presentation following a DC has led to wide variation in the timing and sequence of CP and CSF shunting procedures.

Shunt placement is performed before, after, or simultaneous to CP. Some evidence suggests that staging of shunt placement is associated with similar complication rates and severity of the operation as those with combined shunt placement and CP.[ 11 - 13 , 19 , 28 ] CSF shunting procedures are notoriously known for high complication rates and redo operations. A better understanding of a patient’s outcomes undergoing both surgeries, CSF shunting, and CP, could support the development of a standard approach to clinical management for HCP following DC. In this systematic literature review, we aim to define the types of complications and the complication rates in patients requiring both CP and CSF shunting procedures to understand the sequence of each operation better.

MATERIALS AND METHODS

A systematic literature review was performed in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2009 guidelines [ Figure 1 ]. The PubMed, Cochrane, and Embase databases were queried for the following terms: “cranioplasty” AND (“complication” OR “complications) AND (“VPS” OR “shunt” OR “ventriculoperitoneal shunt”). Our initial search yielded 215 abstracts, which were reduced to 160 after removing duplicates. Of those 160 unique articles, 94 met the criteria for full text review. After the application of inclusion and exclusion criteria, 17 articles were included in the final analysis. Summaries of included studies are outlined in Table 1 . The included studies reported data for 1915 patients and assessed complication rates for pre-CP shunting, simultaneous CP-shunting, and post-CP shunting.

Figure 1:

Illustrating the Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram of our study selection.

Table 1:

Summary characteristics of included studies

Inclusion criteria consisted of original research published between 01/1999 and 05/2020 in PubMed, Embase, and/or Cochrane libraries examining complications in adult patients who underwent CP after a DC. Exclusion criteria included case reports, pediatric patients, studies that examined CSF diversion using devices other than ventriculoperitoneal shunt (VPS), articles not specifying the relative timing of shunting relative to CP, and articles not examining outcome or complications after shunt placement. We extracted data on various parameters, including the number of patients, number of CP procedures, number of VPS procedures, relative timing of VPS procedures in relation to CPs, and complications.

Complications were stratified by severity into minor, major, and serious [ Table 2 ]. Minor complications were those that did not require immediate attention by medical staff. Minor complications included sunken skin flap, transitory neurologic impairment, bleeding, sunken bone flap, and extradural collection. Major complications were those that required immediate neurosurgical intervention or led to death. These included intracranial hemorrhage, epidural hematoma, reoperation, and death. Serious complications were those that would likely develop into life-threatening or debilitating morbidities without surgical and/or medical intervention. Serious complications included infection, hygroma, seizure, bone flap resorption/necrosis, and shunt failure/dysfunction. Due to the heterogeneity of study types, the shunted patients of each pooled group were compared with the rest of the pooled patients in each cohort. All patients and complications were not included in the initial study’s report. Patients included were stratified and pooled when possible.

Table 2:

Complications reported per study

Complication rates and mixed effect odd ratios (ORs) were calculated with meta-analysis to explore the odds of a single complication being associated with shunting for all studies included. Secondarily, the distinct types of complications in each of the three groups were examined. First, cumulative and weighted complication rates were established. Then, contingency tables were created to compare OR between shunt and nonshunt (comparison) groups. Then, these ORs were entered into a spreadsheet and performed a single-staged random effect meta-analysis with inverse variance testing. All calculations were performed using SPSS version 26 (IBM, Armonk, NY). The forest plot of OR was created using Microsoft Excel 2017 (Microsoft, Redmond, WA) using an established spreadsheet by Neyeloff et al.[ 14 ] Because the natural logarithm was used in place of complication rates, the variance was subject to error due to the presence of negative values. The summary values and effect size were not affected by this factor. All zero values used to calculate the OR underwent the Haldane–Anscombe correction. Chi-squared tests were used to examine the relationship between types of complications and the type of procedure.

Using Cochran’s Q statistic and the I² measure, the statistical heterogeneity was assessed among the study populations. Publication bias and overall quality of the studies in relation to their use in this meta-analysis were accounted for using the Newcastle–Ottawa Scale, a three category, 9-point scale assessing cohort selection, comparability, and outcome. Studies including more than one of the groups/procedures of interest were evaluated independently on the NOS scale for each group. In describing the community, the average representative group was considered the same as for other adult trauma cohorts who received CP. Exposed groups were considered as one of the three shunt-CP procedures. For comparability, the principal factor to control for was VPS and its timing. We selected a follow-up of 90% or greater as necessary for the final outcomes star. In addition, we have indicated whether the studies under investigation were directly comparing shunt timing relative to CP, indirectly through investigation of either VPS or CP, or reported relevant results incidentally in an unrelated publication.

RESULTS

Summary of studies

Of the 17 studies that were included in the final analysis, 10 studies (1241 patients) reported data on the shunt-prior-to-CP sequence, 8 studies (844 patients) reported on the combined procedure, and 6 studies (851 patients) evaluated the CP-prior-to-shunt sequence. The majority of studies were retrospective (15), with 2 observational cohort studies included as well. The reported range of mean ages was between 33 and 61 years old, with 984 males (51.4%), 712 females (37.2%), and 219 unstratified patients (11.4%). The most common reasons for DC were trauma, followed by stroke. The size of the DC was reported by 9 studies, ranging from 80 cm[ 2 ] to 205.7 cm[ 2 ]. All studies reporting shunt type used programmable shunt devices. Time from DC to CP was reported in all but 3 studies.

Complication rates

Complication rates were compared between pooled cohorts of patients receiving both procedures and patients who did not have a CSF shunting procedure, i.e., underwent CP alone [ Figure 2 ]. Absolute complication rates were highest in the CP-prior-to-shunt group at 35% (confidence interval [CI] 95%, 30.9–39.1), but the complication rate among studies included in this group was high overall, with a 37.6% complication in the CP-only group. In contrast, the shunt-prior-to-CP group had the lowest complication rate among patients undergoing both procedures at 24.4% (CI 95%, 20.5–28.2), and this was higher than its CP-only control group (13.8%). Similarly, the group of patients undergoing both procedures simultaneously had a complication rate of 30.8% (CI 95%, 28.4–33.3) versus 10% in CP-only controls.

Figure 2:

The average complication rates weighted by the number of patients who had the procedure in each study and compared against a pooled comparison cohort who did not have the procedure. VPS: Ventriculoperitoneal shunt, CP:Cranioplasty

When analyzed by complication type, CP-prior-to-shunt demonstrated the lowest percentage of all complication subtypes, minor, major, and serious [P < 0.001, Figure 3 ].

Figure 3:

Illustration of the percentage of complication subtypes relative to the total recorded complications across all studies reviewed. *P < 0.01, **P < 0.001. CP: Cranioplasty

Complication odds by procedure sequence and study heterogeneity

Odds of complication varied widely in studies examining shunt-prior-to-CP, with odds ranging from 0.129 to 11. Aggregating these studies through meta-analysis demonstrated a trend toward increased risk of complication with this sequence of procedures that did not reach statistical significance [OR 1.89, 95% CI 0.36, 5.62; Figure 4a ]. The analysis revealed an I² value of −26.1%, implying moderate heterogeneity among these studies.

Figure 4:

(a) Graph and table illustrating the odds ratio of patient complication shunt placement before cranioplasty. (b) Illustrates this for the simultaneous shunt cranioplasty combined procedure and (c) for the cranioplasty first procedure patients. CI: Confidence interval

Meta-analysis of studies reporting on simultaneous procedures similarly demonstrated a nonsignificant increase in odds of complication. Odds of complication ranged from 0.409 to 11.8, with a weighted summary OR of 2.42 [95% CI 0.993, 5.92; Figure 4b]. The analysis revealed an I² value of −13.4% based on random effects, indicating low heterogeneity among the studies.

In stark contrast to their overall complication rate above, pooled analysis of patients receiving CP -prior-to-shunt showed a trend toward a lower likelihood of complication, though this result also did not achieve statistical significance. These studies comprised a weighted OR of 0.560 (CI 0.209, 1.49), with odds among individual studies ranging from 0.19 to 10. The analysis revealed an I² value of 1.56% based on random effects, indicating low heterogeneity among the studies [ Figure 4c ].

DISCUSSION

Patients requiring DC frequently develop HCP comorbid to their need for CP during convalescence. The safest timing and sequence of CSF shunting and CP – both operations with high complication rates – are controversial. Further, the relationship between HCP and craniectomy, particularly whether CP could potentially preempt a need for shunting, is gaining increased attention. Therefore, we performed a systematic review and meta-analysis of studies reporting patient outcomes for VPS placement before, during, or after CP to define better the relative risk of different orders of these operations.

Statistical analysis favored no particular order of the two operations. Odds of complication were notably higher for simultaneous operations and shunt-prior-to-CP, but both failed to achieve statistical significance. The trend toward negative odds of complication for CP -prior-to-shunt, however, was a notable contrast. While this also failed to achieve statistical significance, this group of studies also demonstrated lower proportions of minor, major, and serious complications compared to other procedure sequences. The CP-before-shunt group had the highest absolute rate of complications because this group included a relatively high number of patients who developed complications. However, the likelihood of complications (rate per patient) was lower because the group was smaller, and the overall proportion of patients who experienced complications was less compared to the other groups. In essence, while more complications occurred overall in this group, each individual in the group was still less likely to experience a complication than individuals in other groups. The lack of statistical significance may be due to study heterogeneity, but this was low to moderate when analyzed. Alternatively, even though meta-analysis, the study could be underpowered, though each study group contained nearly 800 or more patients, also suggesting a potentially small effect size.

Even so, there is growing evidence to support the CP-prior-to-shunt approach. A possible reason for the reduction in complications when VPS is performed after CP is the return to a more equilibrized CSF state. It is well known that the removal of the bone flap during DC alters CSF dynamics.[ 15 ] A 1984 experiment by Fodstad studied fluid dynamics in 40 patients who underwent CSF infusion testing before and after CP, leading to the conclusion that craniectomy initiates a siphon phenomenon, causing midline shift and disruption in cerebral blood flow.[ 3 ] This mechanism was demonstrated in a cat experiment in which craniectomy altered the pressure/volume relationships in the CNS and was reversed by CP.[ 21 ] A small retrospective review by Oh et al. found that patients undergoing VPS insertion after CP tended to have better outcomes when compared to patients undergoing CP after VPS, though their findings did not reach statistical significance.[ 15 ] In addition, Goedemans et al. found that half of their patients with symptomatic HCP before CP experienced resolution of HCP after undergoing CP.[ 6 ] Collectively these studies provide a breadth of evidence that suggests CP -prior-to-shunt could help restore CSF dynamics, potentially avoiding permanent CSF diversion with shunt placement. Along these lines, it has been suggested that patients who appear to have HCP before CP should undergo temporary diversion through lumbar or external ventricular drains as an alternative to permanent shunting to allow for CSF dynamics to equilibrate and time to determine if permanent CSF diversion is necessary.[ 24 ] The findings from this study go some way toward supporting that management algorithm by indicating a potentially lower likelihood of complication with the CP -first approach. Even so, the need for an approach to CSF diversion must be individualized to each patient as it stands.

In addition, much of the current literature on shunt and CP timing has focused on the combined versus staged approach. While several studies have highlighted the additive risk of two separate operations,[ 5 , 11 , 12 , 23 , 26 ] other retrospective reviews suggest that a combined VPS and CP approach may have even more complications.[ 8 , 20 , 25 ] This result was exaggerated in patients with brain herniation through the craniectomy site or “brain bulging.”[ 8 ] Temporary CSF diversion was proposed as a method to alleviate a “bulging brain” to allow for CP and then VPS if needed, avoiding the complications associated with a combined procedure. Temporary measures may lead to increased length of stay in the hospital and overall cost, but they protect the patient from additional operations and their associated complications. Multiple retrospective reviews reached similar conclusions implicating the presence of a VPS at the time of CP as correlative with significantly increased complications.[ 17 , 24 , 30 ] Importantly, Zhang et al.[ 29 ] performed a meta-analysis of 515 patients to determine the safety of staged versus simultaneous CP and VPS procedures in patients needing both after DC. They similarly found trends toward lower complications in staged surgery but ultimately concluded that it is prudent to individualize decision-making based on each patient’s condition.

This review is limited by the heterogeneity of the studies available for meta-analysis and the low number of studies explicitly examining the VPS and CP procedures. The diverse nature of the DC and CP procedures, including various confounding factors such as the patient’s overall health, injury severity, and size of DC or CP, hinders the ability to control for complications. The existing literature lacks significant data to standardize these variables, making it difficult to draw definitive conclusions. Furthermore, the majority of studies in this field are retrospective and lack control groups. The varying exclusion criteria, bone flap size removal, injury severity, and HCP diagnosis criteria across studies further limit the study’s applicability. Preclinical modeling, as well as prospective clinical studies are needed to determine if CP before shunting truly does exhibit the lowest odds of complication or prevents patients from needing permanent CSF shunt procedures. A large-scale data registry will lend support to stronger conclusions.

CONCLUSION

This systematic literature review and meta-analysis provide an estimated average complication rate for each of the three possible sequences for patients requiring both CSF shunting and CP. The pooled analysis demonstrates no clear winner between procedure sequences but suggests a trend of lower risk of complication with CP before shunt placement. Treatment decisions should continue to be individualized by the patient.

Ethical approval

Institutional Review Board approval is not required.

Declaration of patient consent

Patient’s consent is not required as there are no patients in this study.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Use of artificial intelligence (AI)-assisted technology for manuscript preparation

The authors confirm that there was no use of artificial intelligence (AI)-assisted technology for assisting in the writing or editing of the manuscript and no images were manipulated using AI.

Disclaimer

The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Journal or its management. The information contained in this article should not be considered to be medical advice; patients should consult their own physicians for advice as to their specific medical needs.

References

1. Carvi Y Nievas MN, Höllerhage HG. Early combined cranioplasty and programmable shunt in patients with skull bone defects and CSF-circulation disorders. Neurol Res. 2006. 28: 139-44

2. Fattahian R, Bagheri S, Sadeghi M. Development of posttraumatic hydrocephalus requiring ventriculoperitoneal shunt after decompressive craniectomy for traumatic brain injury: A systematic review and meta-analysis of retrospective studies. Med Arch. 2018. 72: 214

3. Fodstad H, Love JA, Ekstedt J, Fridén H, Liliequist B. Effect of cranioplasty on cerebrospinal fluid hydrodynamics in patients with the syndrome of the trephined. Acta Neurochir (Wien). 1984. 70: 21-30

4. Fotakopoulos G, Tsianaka E, Siasios G, Vagkopoulos K, Fountas K. Posttraumatic hydrocephalus after decompressive craniectomy in 126 patients with severe traumatic brain injury. J Neurol Surg Part Cent Eur Neurosurg. 2016. 77: 88-92

5. Gill JH, Choi HH, Lee SH, Jang KM, Nam TK, Park YS. Comparison of postoperative complications between simultaneous and staged surgery in cranioplasty and ventriculoperitoneal shunt placement after decompressive craniectomy. Korean J Neurotrauma. 2021. 17: 100-7

6. Goedemans T, Verbaan D, van der Veer O, Bot M, Post R, Hoogmoed J. Complications in cranioplasty after decompressive craniectomy: Timing of the intervention. J Neurol. 2020. 267: 1312-20

7. Güresir E, Vatter H, Schuss P, Oszvald A, Raabe A, Seifert V. Rapid closure technique in decompressive craniectomy. J Neurosurg. 2011. 114: 954-60

8. Heo J, Park SQ, Cho SJ, Chang JC, Park HK. Evaluation of simultaneous cranioplasty and ventriculoperitoneal shunt procedures. J Neurosurg. 2014. 121: 313-8

9. Iaccarino C, Kolias AG, Roumy LG, Fountas K, Adeleye AO. Cranioplasty following decompressive craniectomy. Front Neurol. 2020. 10: 1357

10. Lin BJ, Chou KN, Feng SW, Ju DT, Ma HI, Hueng DY. Role of post-decompressive hydrocephalus in patients with malignant cerebral infarction. Turk Neurosurg. 2015. 25: 742-8

11. Lin HY, Lin KC, Tsai CC, Wan D. Simultaneous or staged operation? Timing of cranioplasty and ventriculoperitoneal shunt after decompressive craniectomy. Formos J Surg. 2019. 52: 122

12. Meyer RM, Morton RP, Abecassis IJ, Barber JK, Emerson SN, Nerva JD. Risk of complications with simultaneous cranioplasty and placement of ventriculoperitoneal shunt. World Neurosurg. 2017. 107: 830-3

13. Nasi D, Dobran M. Can early cranioplasty reduce the incidence of hydrocephalus after decompressive craniectomy? A meta-analysis. Surg Neurol Int. 2020. 11: 94

14. Neyeloff JL, Fuchs SC, Moreira LB. Meta-analyses and Forest plots using a microsoft excel spreadsheet: Step-by-step guide focusing on descriptive data analysis. BMC Res Notes. 2012. 5: 52

15. Oh CH, Park CO, Hyun DK, Park HC, Yoon SH. Comparative study of outcomes between shunting after cranioplasty and in cranioplasty after shunting in large concave flaccid cranial defect with hydrocephalus. J Korean Neurosurg Soc. 2008. 44: 211-6

16. Pachatouridis D, Alexiou GA, Zigouris A, Michos E, Drosos D, Fotakopoulos G. Management of hydrocephalus after decompressive craniectomy. Turk Neurosurg. 2014. 24: 855-8

17. Piedra MP, Ragel BT, Dogan A, Coppa ND, Delashaw JB. Timing of cranioplasty after decompressive craniectomy for ischemic or hemorrhagic stroke: Clinical article. J Neurosurg. 2013. 118: 109-14

18. Pierson M, Birinyi PV, Bhimireddy S, Coppens JR. Analysis of decompressive craniectomies with subsequent cranioplasties in the presence of collagen matrix dural substitute and polytetrafluoroethylene as an adhesion preventative material. World Neurosurg. 2016. 86: 153-60

19. Rashidi A, Neumann J, Adolf D, Sandalcioglu IE, Luchtmann M. An investigation of factors associated with the development of postoperative bone flap infection following decompressive craniectomy and subsequent cranioplasty. Clin Neurol Neurosurg. 2019. 186: 105509

20. Schuss P, Borger V, Güresir Á, Vatter H, Güresir E. Cranioplasty and ventriculoperitoneal shunt placement after decompressive craniectomy: Staged surgery is associated with fewer postoperative complications. World Neurosurg. 2015. 84: 1051-4

21. Shapiro K, Fried A, Takei F, Kohn I. Effect of the skull and dura on neural axis pressure-volume relationships and CSF hydrodynamics. J Neurosurg. 1985. 63: 76-81

22. Still M, Kane A, Roux A, Zanello M, Dezamis E, Parraga E. Independent factors affecting postoperative complication rates after custom-made porous hydroxyapatite cranioplasty: A single-center review of 109 cases. World Neurosurg. 2018. 114: e1232-44

23. Ting CW, Lu CH, Lan CM, Lee TH, Hsu SW, Su TM. Simultaneous cranioplasty and ventriculoperitoneal shunt placement in patients with traumatic brain injury undergoing unilateral decompressive craniectomy. J Clin Neurosci. 2020. 79: 45-50

24. Tsang AC, Hui VK, Lui WM, Leung GK. Complications of post-craniectomy cranioplasty: Risk factor analysis and implications for treatment planning. J Clin Neurosci. 2015. 22: 834-7

25. von der Brelie C, Stojanovski I, Meier U, Lemcke J. Open traumatic brain injury is a strong predictor for aseptic bone necrosis after cranioplasty surgery: A retrospective analysis of 219 patients. J Neurol Surg Part Cent Eur Neurosurg. 2016. 77: 19-24

26. Yang XF, Wang H, Wen L, Huang X, Li G, Gong JB. The safety of simultaneous cranioplasty and shunt implantation. Brain Inj. 2017. 31: 1651-5

27. Yeap MC, Chen CC, Liu ZH, Hsieh PC, Lee CC, Liu YT. Postcranioplasty seizures following decompressive craniectomy and seizure prophylaxis: A retrospective analysis at a single institution. J Neurosurg. 2018. 131: 936-40

28. Yu S, Choi HJ, Lee JH, Ha M, Kim BC. A Reappraisal of the necessity of a ventriculoperitoneal shunt after decompressive craniectomy in traumatic brain injury. J Trauma Inj. 2020. 33: 236-41

29. Zhang J, Deng X, Yuan Q, Fu P, Wang M, Wu G. Staged or simultaneous operations for ventriculoperitoneal shunt and cranioplasty: Evidence from a meta-analysis. CNS Neurosci Ther. 2023. 29: 3136-49

30. Zheng WJ, Li LM, Hu ZH, Liao W, Lin QC, Zhu YH. Complications in staged late titanium cranioplasty and ventriculoperitoneal shunting for patients with traumatic brain injury. World Neurosurg. 2019. 127: e1166-71