Posterior Cranial Distraction in Craniosynostosis: A Systematic Review of the Literature

Introduction

Craniosynostosis is the premature fusion of one or more cranial sutures. The incidence of nonsyndromic craniosynostosis is 1/1000 births, while the incidence of syndromic craniosynostosis is 1/2500 births. ¹

Surgery has been the established treatment for craniosynostosis since the late nineteenth century. ² The surgical treatment of craniosynostosis has evolved over the past century, and now includes open remodeling, minimally-invasive suturectomies with spring placement or postoperative helmet therapy, and distraction osteogenesis.

First described in 2009 by White et al., ³ posterior cranial distraction (PCD) affords the ability to progressively expand the posterior cranial fossa, thereby stretching the soft tissue and minimizing the risk of relapse. PCD involves limited epidural dissection, thus allowing shorter surgery and decreased blood loss compared to cranial remodeling. ⁴

Despite its significant advantages, PCD requires a second operation to explant the hardware. In addition, PCD has relatively high complication rates.^5-7 Our goal in this study was to perform a systematic review of the literature, in order to describe the characteristics of patients undergoing PCD, and measure the surgical outcomes of PCD, including gain in intracranial volume and complications. The study aimed to answer two main questions:

What is the average complication rate and the most common complications of PCD?

Methods

Results

Literature Review and Articles Included

As shown in Figure 1, the initial PubMed (using the terms “posterior” AND “distraction” AND “cranial” OR “calvarial”) search yielded 114 articles. After manually excluding articles that were clearly not related to posterior distraction, and articles that were not in the English language, 75 articles remained. After review of the full text of the remaining articles and their references, 18 articles were included in the systematic review. Details of the included articles are shown in Table 1. Two of the articles were published by the same institution,^10,23 but the date ranges of patient recruitment did not overlap.

OPEN IN VIEWER

Figure 1.

Extraction of articles included in the systematic review.

OPEN IN VIEWER

Table 1.

Details of Articles Included in the Systematic Review.

Article	Year of publication	Years of study	Institution	Number of subjects	Number of subjects with syndrome	Mean age (months)	Mean distraction amount (mm)	Mean intracranial volume gain (cm3)	Complication rate (%)
Sakamoto et al. 8	2016	2010–2013	Keio University School of Medicine, Tokyo, Japan	6	4	9.3	18.7	NR	0
Iida et al. 9	2019	2014-2017	Keio University School of Medicine, Tokyo, Japan.	13	9	8.8	NR	NR	0
Komuru et al. 10	2015	NR	Jutendo University, Tokyo, Japan	13	7	21	NR	NR	69.2
Salokorpi et al. 7	2017	2009–2015	Oulu University, Hospital, Oulu, Finland	30	14	34.8	28	275.4	46.7
Kamil et al. 11	2021	NR	Kagoshima University, Japan and Airlangga University, Indonesia	4	2	19.5	22.3	172.5	25
Marji et al. 12	2021	2016-2019	University of Pittsburgh School of Medicine, PA, USA	4	0	10.6	30.3	236	25
Johns et al. 13	2016	2012-2014	Primary Children's Hospital, Salt Lake City, Utah, USA	9	9	NR	NR	NR	22.2
Di Rocco et al. 14	2018	2011-2014	Hôpital Necker-Enfants Malades, Paris, Frances	21	13	8.6	NR	132	14.3
Zhang et al. 15	2018	2009-2016	Children's Hospital of Philadelphia, PA, USA	63	44	36	27.7	NR	25.4
Ong et al. 16	2014	2009-2013	Dell Children's Medical Center of Central Texas, Austin, TX, USA	17	6	51	22.5	NR	23.5
Thatikunta et al. 17	2020	2007-2019	Norton Children's Hospital, Louisville, KY	9	4	27	NR	NR	22.2
Satanin et al. 18	2019	2015–2016	Burdenko Neurosurgery Institute, Moscow, Russia	11	11	22.6	28.7	160.5	63.6
McMillan et al. 19	2017	2006-2015	Birmingham Children's Hospital, United Kingdom	49	33	34.6	24	NR	65.3
Park et al. 20	2022	2012–2020	Asan Medical Center, Seoul, South Korea	11	7	34.6	23.4	261	0
Brandel et al. 21	2018	2013-2017	University of California San Diego, CA, USA	4	3	5.9	29.5	535	NR
Jeon et al. 22	2022	2008–2020	Seoul National University Children's Hospital, Seoul, South Korea	16	14	26.1	25.9	214	68.8
Thomas et al. 23	2014	2007-2012	Oxford Craniofacial Unit, United Kingdom	31	30	20.3	19.3	NR	61.3
Raposo-Amaral et al. 24	2022	2007-2019	SOBRAPAR Hospital, Campinas, Brazil	14	14	NR	19	NR	NR
Overall				325	224	22.1	24.7	248.3

NR, Not Reported.

Surgical Outcomes

The average length of distraction was 24.7 mm, ranging from 18.7 mm to 30.3 mm.^10,11 Eight studies computed the average intracranial volume gain using preoperative and postoperative computed tomography scans.^{8,11,12,14,18,20-22} In those studies, the average volume gain was 253.2 cm³. The average volume gain per mm of distraction was 9.8 cm³.

Sixteen studies representing 307 subjects detailed postoperative complications.^8-21,23,25 The overall complication rate in those studies was 32.2% (Figure 2). The most common complication was surgical-site infection (10.7%), followed by hardware failure/fracture/dislodgement (8.5%), delayed wound healing (6.2%), persistent cerebrospinal fluid leak (5.5%), significant hemorrhage (0.7%), meningitis (0.3%) and death (0.3%). The two cases of significant hemorrhage included one postoperative epidural hematoma, ¹⁶ and one case of severe intraoperative hemorrhage that required abandoning the procedure and returning to complete in a several days. ⁹ The death was due to postoperative respiratory failure in a patient with Crouzon syndrome. ¹³

OPEN IN VIEWER

Figure 2.

Complications of posterior cranial distraction.

Four studies included 35 patients with pre-existing cerebellar tonsillar herniation.^13,20,21,25 After posterior cranial distraction, tonsillar herniation improved radiologically in 26 patients (74.3%).

Two studies included 40 patients with pre-existing signs of increased intracranial pressure.^9,20 Those studies found that 37 patients (92.5%) had improvement in their intracranial hypertension, based on clinical, ophthalmologic or radiologic criteria.

Discussion

Posterior cranial distraction has gained significant popularity in the management of multisuture and syndromic craniosynostosis, due to its ability to significantly increase intracranial volume. On average, in the included studies, PCD increased intracranial volume by 253.2 cm³, which represents an increase in intracranial volume of approximately 25%. ¹² Several studies have demonstrated that PCD is more effective at increasing intracranial volume than fronto-orbital advancement,^26,27 which only provides approximately 8% of volume increase. ¹⁸ In the 40 patients with pre-existing signs of intracranial hypertension, 92.5% improved after PCD. These advantages, in addition to the fact that PCD has been found to improve turribrachycephaly ²⁸ and fronto-orbital contour,^24,29 have changed the paradigm for the treatment of multisuture and syndromic craniosynostosis,^26,30 allowing the delay of fronto-orbital advancement until an older age when relapse is less likely. ³⁰

In addition to increasing intracranial volume and attenuating intracranial hypertension, posterior cranial distraction has the potential to treat cerebellar tonsillar herniation, with improvement in up to 75% of patients, as noted above, thereby potentially obviating the need for posterior fossa decompression.^31,32 In addition, Lin et al. found that patients who underwent PCD were significantly less likely than those who underwent posterior cranial remodeling to develop new-onset cerebellar tonsillar descent postoperatively. ³³ Despite the benefits of PCD, there remain a significant proportion of patients whose tonsillar herniation does not improve postoperatively. It remains difficult to predict when PCD will be insufficient at preventing or improving tonsillar herniation. Therefore, where there is significant concern by the neurologic surgeon (significant tonsillar descent, very restricted foramen magnum), in-continuity posterior fossa decompression can usually be safely performed concurrently with PCD. ³⁴

CSF leakage remains a relatively common complication of PCD, occurring in approximately 5.9% of cases. The dural laceration and subsequent CSF leak are believed to be due to the screw tips lacerating the dura during the activation phase of distraction. Various strategies have been employed to decrease this risk, including using resorbable material in the epidural space to act as a barrier between the dura and the screws,^5,35 using self-drilling screws, ¹⁷ using the shortest screws possible, ⁶ and using resorbable pins instead of titanium screws. ¹⁴ No comparative study has been published to establish the effectiveness of any of these maneuvers. Overall, we found that PCD has a complication rate of 32.2%, which is higher than calvarial vault remodeling (5-11%), ³⁶ spring-assisted cranioplasty (7.4%) ³⁷ and minimally-invasive suturectomy and helmet therapy (3.5%). ³⁸

Another important variable with PCD is the craniocaudal level of the occipital osteotomy. This osteotomy may be performed cranial or caudal to the level of the torcula. Some authors describe only performing supratorcular osteotomies.^26,28,33 Other choose the level of the osteotomy based on preoperative imaging. ²⁰ It is unclear whether an infratorcular osteotomy is more effective at improving tonsillar descent, ²⁰ or increasing intracranial volume, ¹² than supratorcular osteotomy. Zapatero et al. found that patients with a supratorcular osteotomy had a higher risk of developing an occipital step-off deformity. ³⁹ Several authors have attempted to minimize that deformity by performing a supratorcular osteotomy and adding barrel staves to the bone caudal to the osteotomy,^13,28,40 although evidence suggests that these barrel stave osteotomies do not help improve the step-off deformity.^17,34 In addition, infratorcular osteotomies may minimize the risk of injury to the torcula and bleeding,^3,9 although there is limited data to support that.

Since this is a systematic review of retrospective studies, this study has similar limitations to the underlying studies, namely small number of patients and lack of randomization. In addition, certain factors, such as the craniocaudal level of the osteotomy (supratorcular versus infratorcular), were only analyzed by a small number of studies, thereby making extraction of useful data difficult. The volumetric measurements may suffer from heterogeneity due to lack of standard interval between surgery and follow-up CT scan, ranging from a few months to over a year. Our study focused on posterior cranial distraction, and did not include other types of cranial distraction, such as anterior distraction for coronal craniosynostosis, and spring-assisted cranioplasty.

Conclusion

Posterior cranial distraction is a powerful technique in the initial treatment of multisuture or syndromic craniosynostosis, although the complication rates are significantly higher than traditional remodeling techniques. Despite its relatively high complication rates, the procedure allows significant increase in intracranial volume, decreasing intracranial pressure, and potentially alleviating cerebellar tonsillar descent. Further studies are needed to better evaluate the optimal craniocaudal level of the occipital osteotomy, and techniques to minimize the risk of CSF leakage.