Baseline characteristics

A total of 191 patients met the inclusion criteria. The mean age of the patients was 38.6 ± 14.3 years, with 83.8% being male. The median follow-up time from decompressive craniectomy to cranioplasty was 14.6 months. The causes of initial injury were primarily road-traffic incidents (40.8%), accidental falls (30.9%), blunt trauma (7.9%), and others (20.4%). Alloplastic materials used for cranioplasties included titanium (83.2%) and PEEK (16.8%). Preoperative hydrocephalus was present in 14.7% of patients, and 4.7% underwent permanent cerebrospinal fluid diversion before cranioplasty. The majority of cranioplasties were performed on unilateral decompressive craniectomy (92.7%). The timing interval was stratified into three groups for comparison: Early CP (less than 3 months), Intermediate CP (3–6 months), and Delayed cranioplasty (more than 6 months). Thus, 27 patients underwent early cranioplasty, 90 during 3–6 months, and 74 were in the group of delayed CP.

During long-term follow-up, the mortality rate was 1.57%, all resulting from non-surgical factors. Overall, 50 patients (26.2%) experienced at least one major complication, including new-onset seizures (18.9%), postoperative hematoma necessitating operation evacuation (3.1%), and implant failure requiring implant explantation (5.8%). The minor complication of subgaleal effusion was found in 36 patients (18.9%) and could be managed conservatively or with drainage. Among these, 24 patients had subgaleal effusion as the only complication without any major complication. All the demographic information is included in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Clinical characteristics of the patients at baseline.

https://doi.org/10.1371/journal.pone.0321870.t001

Overall complications

In univariate analysis, smoking history (OR 2.49, 95% CI = 1.35–4.60, p = < 0.01) was associated with postoperative overall complication. There is a trend indicating a difference in bilateral defect size (OR 3.10, 95% CI = 1.00–9.65, p = 0.051). However, only the smoking history was significantly associated with overall complications in multivariate logistic analysis. No other factors were significantly associated with postoperative overall complications in this multivariate logistic analysis. The logistic regression analysis are displayed in S1 Table.

Major complications

The rate of overall major complications was 26.2% (36 of 191). The proportion of each complication is summarized in Table 2. The most common postoperative major complication was new-onset seizures (26.2% in all patients). In univariate analysis, factors included smoking history, bilateral cranial defect, intermediate and delayed cranioplasty were associated with major complications after CP. In multivariate analysis, smoking history (OR 2.75, 95% CI = 1.25–6.05, p = 0.021), intermediate and delayed CP [(OR = 7.11, 95% CI = 1.44–35.04, p = 0.016), OR = 5.17, 95% CI = 1.01–26.49, p = 0.049)], and titanium mesh CP (OR 3.30, 95% CI = 1.03–10.59, p = 0.045) predicted major complications, see Table 3 and Fig 1. We developed a clinical prediction model for major complications, as shown in S1 Figure. It demonstrated good discriminatory ability, distinguishing patients with and without major complications with an AUC of 0.72 (95% confidence interval [CI]: 0.64–0.80).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Complications following Cranioplasty.

https://doi.org/10.1371/journal.pone.0321870.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Factors contributing to major complications.

https://doi.org/10.1371/journal.pone.0321870.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Forest plot presenting the significant results of the multivariate logistic regression analysis for major complications and new-onset seizures.

PEEK, Polytheretherketone.

https://doi.org/10.1371/journal.pone.0321870.g001

New-onset Seizures

Patients who had seizures before cranioplasty were excluded from the analysis. Multivariate analysis showed a significantly higher rate of new-onset seizures if cranioplasty was performed at 3–6 months after initial craniectomy (OR = 5.46, CI = 1.07–27.95, p = 0.042). Additionally, other significant factors associated with postoperative new-onset seizure were the type of alloplastic material, with a higher rate of seizure when the cranioplasty was performed using titanium mesh (OR = 4.85, 95% CI = 1.01–23.37, p = 0.049), smoking history (OR = 4.69, 95% CI = 1.88–11.71, p = <0.001). Surprisingly, the increasing age is associated with lower rate of seizures even if this factor was not statistically significant in univariate analysis, see Table 3 and Fig 1.

Hematoma evacuation and implant failure

Due to the low incidence rates of hematoma (3.14%, 6 cases) and implant failure (5.76%, 11 cases), reliable analyses cannot be conducted using univariate and multivariate analyses. Consequently, we have presented the distribution characteristics of these cases in the S2 Table. and S3 Table.

Subgaleal effusion

Subgaleal effusion-epidural or subdural effusion with clinical signs and symptoms of increased intracranial pressure with or without midline shift on postoperative CT scan, was a minor complication which could be managed conservatively. Univariate analysis revealed that hydrocephalus shunts before CP (OR = 6.09, 95% CI = 1.55–23.97, p = 0.010) was associated with a higher risk, while hydrocephalus after DC showed a trending difference (OR = 2.39, 95% CI = 0.98–5.83, p = 0.057). Other factors did not reach statistical significance. In multivariate analysis, only female gender predicted a higher rate of postoperative subgaleal effusion, see Table 4.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Factors contributing to subgaleal effusion (minor complications).

https://doi.org/10.1371/journal.pone.0321870.t004

Risk factors for Overall complications, Major complications, and related factors

Although some studies have shown that age [10,19] is associated with a higher risk of complications, Our study found that younger age is a risk factor for new-onset seizures. This may be due to the retrospective nature of our study, which introduces a certain degree of selection bias, combined with an insufficient sample size, both of which may contribute to these results. Additionally, due to the limited sample size, this study only included key factors in the regression model analysis and did not account for confounding factors such as the presence of underlying comorbidities in patients. These unaccounted factors might also influence the analysis results.

Smoking has been reported as a risk factor in previous study [8,19]. In our study, both univariate and multivariate analysis revealed that smoking was related to major complications and new-onset seizures. Likewise, according to our results, smoking is the only independent risk factor for overall complications (S1 Table). Smoking adversely impacts wound healing through various mechanisms. Nicotine triggers vasoconstriction, which reduces blood flow and oxygen delivery to tissues—both essential for would healing. Additionally, smoking impairs fibroblast function and collagen synthesis, crucial components of tissue repair. These factors contribute to an increased incidence of surgical site infections and implant failure in smokers. Moreover, smoking is associated with systemic inflammation and oxidative stress, which can heighten neuronal excitability and increase the risk of seizures[33,34]. Overall, the neurosurgeons should recognize these relevant factors and take active measures to prevent the potential complications, such as implementing smoking cessation programs.

To data, the optimal interval between craniectomy and cranioplasty remains unclear. Most studies rely on two-group comparison and categorize cranioplasty timing into “early” and “late”, with 3 or 6 months as the chosen time points. Some recent studies have shown that early cranioplasty could improve neurological function [22,20], whereas others suggest early cranioplasty is related to higher complication rates [24,25]. Meanwhile, other studies reported late cranioplasty is associated with higher complications rates [26], and some showed no difference in complication rate between early and late groups [27,35]. Our study classifies time intervals as early CP (less than 3 months), Intermediate CP (3–6 months), and Delayed cranioplasty (more than 6 months) according to a recent consensus [30]. The results indicate that cranioplasty performed more than 3 months after the initial surgery is a risk factor for postoperative major complications and new-onset seizures, consistent with previous reports [7]. The majority of our patients underwent cranioplasty more than 3 months after the initial surgery, and none of the patients who had cranioplasty within 3 months developed implant failure. This finding aligns with previous reports [36], while work of some researchers shows opposite trend [37]. In the past, some recommended waiting at least one year or waiting as long as possible to allow sufficient recovery for patients [36]. However, our study found that late cranioplasty was clearly associated with a high rate of complications, indicating that early cranioplasty is beneficial in reducing the occurrence of major complications. More studies are needed to determine the effect of timing on neurological function outcome [38].

Cranioplasty has undergone many advancements in the search for ideal materials to improve outcomes. Although autologous bone was always regarded as gold standard in skull reconstruction, bone flap resorption is a common complication [28,39]. Some recent studies have reported that autologous bone grafts appear to carry an increased reoperation rate [16,29,40]. Zaed et al. conducted a systematic review focusing on the pediatric population, comparing autologous and heterologous outcome. Their findings revealed that heterologous materials are superior to autologous bone for CR in children [41]. Also, Signorelli et al. conducted a comprehensive analysis of factors leading to bone flap resorption after decompressive craniectomy. Their study concluded that bone flap fragmentation, TBI etiology, and younger age increase the risk of bone flap resorption. This raises the question of whether alloplastic materials should be recommended for these patient population [42]. Various methods of alloplastic cranioplasty have been developed, each with different characteristics [13,14,43]. In this study, titanium mesh and PEEK were the implanted materials. Titanium mesh showed a higher incidence of overall major complications, particularly seizures. A possible explanation for this is titanium’s high thermal conductivity, which allows environment temperature fluctuations to be transmitted to the brain. These may influence neuronal homeostasis, thereby increasing risk of seizures. Also, we found the incidence of implant failure and hematoma were higher in titanium CP [(90.9% vs. 9.1%), (83.3% vs. 16.7%), respectively], as shown in S3 Table. However, considering the limited sample in the two groups, no logistic regression analysis was done. Our previous study also demonstrated that cranioplasty with PEEK had advantages in brain function improvement and fewer complications [13]. Therefore, we consider PEEK cranioplasty to be more ideal for cranial reconstruction. But is PEEK necessarily the best choice? Although PEEK CP has various advantages, our previous studies have also shown that PEEK CP requires longer surgical time and costs nearly four times as much as titanium. Therefore, it is necessary to consider the cost-effectiveness and the patients’ financial situation. More research data are needed in the future [44, 45].

Some studies have found bilateral defect site to be associated with cranioplasty complications, such as seizure and death [31,46]. In our study, we also found that patients with bilateral cranioplasty had a higher rate of major complications. However, the statistical difference was not present in the multivariate analysis. We hypothesize multiple operations, longer incisions, thin soft tissue, and involvement of the skull base contribute to these complications, as observed in previous studies [47,48].

Seizures that follow cranioplasty was the most common complication in our study, with incidence of 18.9% (36 cases). This may be due to the fact that the included patients underwent decompressive craniectomy for TBI, as some studies have shown that TBI before cranioplasty increased the risk of seizures [8,21]. Additionally, the large majority of patients had cranioplasty performed more than 3 months after the initial surgery, which may also be another reason, as a longer interval between decompressive and cranioplasty increases the likelihood of postoperative seizures [7]. Other contributing factors might include the use of bipolar coagulator during the operation and scar-related epilepsy.

Risk factors for subgaleal effusion and related factors

Subgaleal effusion, while common, has received less attention in similar publications. Some authors speculated it as delayed allergic reactions particularly in PEEK cranioplasty [17]. We found no significant difference in the incidence of subgaleal effusion across different materials (Table 2). Our study suggests that there are associations between hydrocephalus shunts before CP, and postoperative subgaleal effusion; however, this difference was only observed in the univariate analysis and did not reach statistical significance in the multivariate analysis. Yao [49] also reported that a history of ventriculoperitoneal shunt was an independent risk factor for overall complications including subgaleal effusion. The disturbance of cerebrospinal fluid circulation could be the possible reason. Planning cranioplasty in conjunction with hydrocephalus remains a significant challenge. Our findings align with a recent consensus emphasizing the need for standardized diagnostic and surgical protocols for both cranioplasty and post-traumatic hydrocephalus. The recommended management sequence prioritizes cranioplasty as the primary intervention [50]. Additionally, cranioplasty performed within 3 months was identified as a risk factor. Therefore, the adjustment of drainage speed could be important. We also identified sex as a significant variable and found that females had a higher risk of subgeleal effusion, which is consistent with findings of Chang [10]. However, laboratory examinations of the effusion were not performed, and more data are needed in the future to elucidate the mechanisms and outcomes of this complication.

Limitations and strengths of this study

This study has several limitations. First, as a retrospective study, it is subject to possible selection bias. Subsequent analyses will employ Propensity score matching (PSM), which has been widely used to reduce confounding biases in observational studies. Additionally, the study did not include enough patient-specific and surgery-specific risk factors, such as diabetes, hypertension, and dura tearing during the cranioplasty surgery, and it lacked stratification of some factors. Furthermore, most patients underwent cranioplasty more than three months after the initial surgery, and the small sample for “early” cranioplasty was relatively small. Moreover, the classification of subgaleal effusion as a minor complication may not fully reflect its potential severity or its progression to major complications, such as seizures. This categorization could lead to an underestimation of its clinical impact. Future research should focus on stratifying subgaleal effusion based on severity and clinical management requirements to provide a more comprehensive understanding of its impact.

One of the major strengths is its multi-center study design, which allows for a larger and more diverse patient population, thereby improving the robustness of the results. Additionally, by including only patients with TBI, we reduced the heterogeneity of the pathologies of decompressive craniectomy. The large sample size and long-term follow-up allow for robust statistical analysis. Moreover, we analyzed the timing and various alloplastic materials used in cranioplasty, which are two contentious points in cranioplasty. Although this study did not focus on the relationship between the timing of cranioplasty and neurological functions, the effect of timing on neurological outcomes is currently an important area of research. Further prospective studies are required to explore those findings.