Subcutaneous bone flap storage after emergency craniectomy: cost-effectiveness and rate of resorption

OBJECTIVEDecompressive craniectomy is used for uncontrolled intracranial pressure in traumatic brain injury and malignant hemispheric stroke. Subcutaneous preservation of the autologous bone flap in the abdomen is a simple, portable technique but has largely been abandoned due to perceived concerns of resorption. The authors sought to characterize their experience with subcutaneous preservation of the bone flap and cranioplasty.METHODSThe authors performed a retrospective single-institution review of subcutaneous preservation of the autologous bone flap after decompressive craniectomy from 2005 to 2015. The primary outcome was clinically significant bone resorption, defined as requiring a complete mesh implant at the time of cranioplasty, or delayed revision. The outcome also combined cases with any minor bone resorption to determine predictors of this outcome. Logistic regression modeling was used to determine the risk factors for predicting resorption. A cost comparison analysis was also used via the 2-sided t-test to compare the cost of cranioplasty using an autologous bone flap with standard custom implant costs.RESULTSA total of 193 patients with craniectomy were identified, 108 of whom received a cranioplasty. The mean time to cranioplasty was 104.31 days. Severe resorption occurred in 10 cases (9.26%): 4 were clinically significant (2 early and 2 late) and 6 demonstrated type II (severe) necrosis on CT, but did not require revision. Early resorption of any kind (mild or severe) occurred in 28 (25.93%) of 108 cases. Of the 108 patients, 26 (24.07%) required supplemental cranioplasty material. Late resorption of any kind (mild or severe) occurred in 6 (5.88%) of 102 cases. Of these, a clinically noticeable but nonoperative deformity was noted in 4 (3.92%) and minor (type I) necrosis on CT in 37 (37%) of 100. Bivariate analysis identified fragmentation of bone (OR 3.90, 95% CI 1.03-14.8), shunt-dependent hydrocephalus (OR 7.97, 95% CI 1.57-40.46), and presence of post-cranioplasty drain (OR 9.39, 95% CI 1.14-1000) to be significant risk factors for bone resorption. A binary logistic regression optimized using Fisher's scoring determined the optimal multivariable combination of factors. Fragmentation of bone (OR 5.84, 95% CI 1.38-28.78), diabetes (OR 7.61, 95% CI 1.37-44.56), and shunt-dependent hydrocephalus (OR 9.35, 95% CI 1.64-56.21) were found to be most predictive of resorption, with a C value of 0.78. Infections occurred in the subcutaneous pocket in 5 (2.60%) of the 193 cases and after cranioplasty in 10 (9.26%) of the 108 who underwent cranioplasty. The average cost of cranioplasty with autologous bone was $2156.28 ± $1144.60 (n = 15), and of a custom implant was $35,118.60 ± $2067.51 (3 different sizes; p < 0.0001).CONCLUSIONSCraniectomy with autologous bone cranioplasty using subcutaneous pocket storage is safe and compares favorably to cryopreservation in terms of resorption and favorably to a custom synthetic implant in terms of cost. While randomized data are required to definitively prove the superiority of one method, subcutaneous preservation has enough practical advantages with low risk to warrant routine use for most patients.