ORIGINAL ARTICLE
Lessons Learned in Posterior Cranial Vault Distraction Gregory P.L. Thomas, PhD, FRCS (Plast), Steven A. Wall, FCS(SA) Plast, Jayaratnam Jayamohan, FRCS, Shailendra A. Magdum, FRCS, Peter G. Richards, FRCS Akira Wiberg, MRCS, and David Johnson, DM, FRCS (Plast)
Introduction: Posterior distraction (PD) is rapidly emerging as an important technique to increase the intracranial volume and correct calvarial morphology in patients with severe brachycephaly or turribrachycephaly. Methods: A retrospective review was performed of all 31 patients who underwent PD at the Oxford Craniofacial Unit between 2007 and 2012. Results: Twenty-three patients (74.2%) underwent PD as a primary procedure at a median age of 8 months. Eight patients (25.8%) had PD as a secondary transcranial procedure at a median age of 48 months. Full distraction to 20 mm was achieved in 28 patients (90.3%). Of these, all but 1 demonstrated a significant improvement in morphology, with a resolution of the symptoms and signs of raised intracranial pressure in all proven to have it preoperatively. Unanticipated events occurred in 61.3% of patients, with 19.4% undergoing one or more unplanned procedures. Wound infection (29.0%) and tissue necrosis (22.6%) were the commonest. Cerebrospinal fluid leaks were rarer (6.5%) but prevented full distraction. Nine patients (29.0%) had a consolidation period of less than 30 days without experiencing relapse. In 11 patients who had a later frontoorbital advancement and remodeling, wound closure was tight, resulting in dehiscence in 3 cases (27.3%). Conclusions: Posterior distraction is an effective procedure in the management of severe brachycephaly or turribrachycephaly but has associated risks. Our protocol has evolved with experience to favor a reduced latency period and consolidation phase and the use of 2 distractor devices. Key Words: Posterior distraction, distraction osteogenesis, syndromic craniosynostosis, raised intracranial pressure, Chiari malformation, wound infection, tissue necrosis (J Craniofac Surg 2014;25: 1721–1727)
P
atients with syndromic craniosynostosis manifesting as turribrachycephaly or severe brachycephaly typically have an underdeveloped posterior cranial fossa and are at increased risk for From the Oxford Craniofacial Unit, Oxford University Hospitals NHS Trust, Oxford, UK. Received September 18, 2013. Accepted for publication April 7, 2014. Address correspondence and reprint requests to David Johnson, DM, FRCS, Oxford Craniofacial Unit, Oxford University Hospitals NHS Trust, Headley Way, Oxford OX3 9DU, UK; E-mail: [email protected] The authors report no conflicts of interest. Copyright © 2014 by Mutaz B. Habal, MD ISSN: 1049-2275 DOI: 10.1097/SCS.0000000000000995
having raised intracranial pressure (ICP), hydrocephalus, or a Chiari type 1 malformation.1–4 Early calvarial vault expansion is routinely performed in these patients to protect the rapidly growing brain from such hazards. Traditionally, this was achieved through primary fronto-orbital advancement.5 However, early fronto-orbital advancement is prone to relapse in syndromic cases, necessitating a revision procedure, and does not reliably prevent the later development of raised ICP.6–8 Expansion of the posterior calvarial vault is increasingly favored as an alternative primary procedure because it not only directly addresses the posterior fossa but also has the potential to deliver a far greater increase in intracranial volume.9–11 Posterior release has been practiced for many years, and although often effective, results can be unpredictable, with approximately 30% of patients having an unfavorable outcome.12 This is likely due to the fact that the technique relies on brain expansion to drive the released occipitoparietal bone flap backward. A prolonged supine posture in the immediate postoperative phase can impede expansion and result in early ossification of the craniotomy site.11 The adoption of distraction techniques in the posterior cranium negates this issue because the bone flap is fixed by the distractor device. It is not surprising that posterior calvarial vault expansion by distraction has been widely taken up since it was described by White and colleagues13 in a series of 6 patients in 2009 because it gives the surgical team unparalleled control over the expansion. Furthermore, in posterior distraction (PD), provided that the bone flap is not dissected free of the dura, the opportunities for dural injury and significant hemorrhage are fewer, the bone remains vascularized, and the dead space beneath the bone is eliminated.13,14 This should translate into a decreased operating time, reduced blood loss, and reduced overall morbidity, consequently leading to briefer hospital admissions.14 Posterior distraction does have some disadvantages, not least that it inevitably requires a second surgical procedure at which the distractor devices are removed. Other complications are recognized to occur, particularly wound infection involving the device and necrosis of the tissue overlying the device, leading to its exposure.12,15 The complication profile of the procedure is likely to change as surgeons and units increase in experience and devices along with distraction protocols are modified accordingly. We present a single large consecutive series of 31 patients operated on in our unit between 2007 and December 2012 and describe our evolving approach to PD. In this study, patients underwent 1 of 2 PD techniques. Those with significant turricephaly had PD using the “hinge” technique (described in the “Methods” section), in which the posterior panel was rotated posteriorly and downward during distraction, with the aim of reducing the vertex height in addition to expanding the posterior vault. In the patients in whom expansion of the posterior fossa was the primary goal, the released posterior panel was left unhinged (the “push-back” technique), enabling it to be displaced in line with the distraction vector without rotation, thereby increasing the expansion of the posterior cranial fossa.12
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1721
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
Thomas et al
METHODS A retrospective review was performed of the case notes of all patients who underwent PD under the care of the Oxford Craniofacial Unit between January 2007 and December 2012. Adequate data were available for all 31 cases identified. The operative technique used in this series has been described previously.12 Briefly, in all cases, the calvarium was approached via a coronal zigzag incision. An osteotomy was performed extending from the vertex to below the torcula, circumscribing a posterior panel that remained attached to the underlying dura. Either 1 or 2 KLS Martin distractor devices (KLS Martin Group, Tuttlingen, Germany) were placed across the craniotomy in the temporoparietal region. Arnaud cranio-orbital distractors were used in 29 cases, whereas a modification of this device was used in the remaining 2 cases. In 19 cases, the posterior panel was hinged inferiorly in the midline with either a stainless steel wire or PDS suture in a figure-of-8 configuration (hinge technique). In the remaining 11 cases, the posterior panel was left unhinged, and 2 distractors were fitted to displace the panel directly backward (push-back technique). In 19 cases, initial distraction was started on the operating table, with a mean distraction distance of 2.6 mm (range, 1–6 mm). In the other 12 cases, distraction was started after 1 to 2 days' latency, except in 1 case in which distraction was started on the fourth day after surgery. Distraction then proceeded at approximately 1 mm per day until the devices achieved their maximum distraction distance of
20 mm. Once this was reached, the distractor activation arms were usually trimmed as close to the scalp as feasible to reduce the chance of the arm snagging accidentally. This was performed in the outpatient clinic without recourse to anesthesia. After a consolidation period, the devices were removed under general anesthetic. Serial skull radiographs were used to monitor distraction progress, in addition to standard clinical observations. All patients received broad-spectrum antibiotic prophylaxis (amoxicillin and clavulanic acid) at anesthetic induction and for 5 days after surgery. The patients who presented postoperatively with signs of a wound infection received empirical intravenous or oral antibiotic treatment with regular wound toilet once microbiologic samples had been taken. Antibiotic treatment was modified according to microbiologic culture results, as guided by a specialist microbiologist.
RESULTS In total, 31 patients (21 girls, 10 boys) underwent PD. All but 1 patient had a syndromic condition causing craniosynostosis, confirmed on genetic screening (Table 1). All patients had brachycephaly or progressive turribrachycephaly secondary to bicoronal or multisuture craniosynostosis (Table 2). Twenty-three patients were treated by PD as a primary procedure at a median age of 8 months old (range, 5–39 mo). Three of these patients had raised ICP diagnosed on intraparenchymal
TABLE 1. Demographic Data by Diagnosis
Patient
Diagnosis
Raised ICP/ Chiari Malformation
Apert — Apert — Apert — Apert Chiari Apert — Apert — Apert — Apert Raised ICP Apert — Apert — Apert Chiari Apert Raised ICP Craniofrontonasal — Crouzon Raised ICP Crouzon Raised ICP; Chiari Crouzon Raised ICP Crouzon Raised ICP Hypophosphatasia Raised ICP Meunke — Meunke — Meunke — Meunke — Meunke — Pfeiffer Raised ICP; Chiari Saethre-Chotzen — Saethre-Chotzen — Saethre-Chotzen — TCF12 — TCF12 — Bilambdoid, Chiari sagittal (NS)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Age, mo
No. Distractors: Technique
Distraction Period, d
Consolidation Period, d
Advancement, mm
First Transcranial Operation
Hospital Stay, d
Unplanned Postoperative Event
5 5 5 5 6 7 8 10 14 17 24 43 9 39 52 72 154 7 5 5 6 9 10 45 7 8 12 9 14 16
2: Hinge 1: Hinge 1: Hinge 2: Push-back 1: Hinge 2: Hinge 2: Hinge 1: Hinge 2: Hinge 2: Push-back 2: Push-back 1: Hinge 2: Push-back 1: Hinge 2: Push-back 2: Push-back 2: Push-back 2: Hinge 1: Hinge 2: Push-back 2: Hinge 2: Hinge 2: Push-back 1: Hinge 2: Hinge 2: Hinge 2: Hinge 2: Push-back 2: Hinge 2: Push-back
30 19 30 17 18 19 20 19 29 20 22 23 20 26 19 30 7 20 19 18 21 28 19 19 24 22 24 21 24 21
119 121 83 67 87 23 1 30 27 93 86 71 102 76 23 57 0 0 47 34 77 133 75 77 29 37 67 123 46 13
20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 13 7 20 20 20 20 20 20 17 20 20 20 20 20 20
Yes Yes Yes No Yes Yes Yes Yes Yes Yes No No Yes Yes No No No Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes Yes
8 8 6 30 10 22 5 10 8 11 6 7 8 6 10 13 25 26 8 7 4 11 9 14 8 7 8 4 8 5
— Yes — Yes — Yes Yes Yes Yes — — — Yes Yes — Yes Yes Yes — Yes Yes — Yes Yes — Yes — Yes Yes Yes
NS, no syndromic diagnosis; TCF12, TCF12-related craniosynostosis.
1722
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
TABLE 2. Indication for PD by Technique Operative Technique Indication for Surgery Progressive turribrachycephaly Severe brachycephaly Raised ICP with turribrachycephaly Raised ICP with brachycephaly Chiari malformation with turribrachycephaly Chiari malformation with brachycephaly Chiari + raised ICP with turribrachycephaly Chiari + raised ICP with brachycephaly Total
Hinge
Push-Back
Total (N = 31)
8 6 3 1 0 0 1 0 19
1 4 3 0 1 2 0 1 12
9 (29.0) 10 (32.3) 6 (19.4) 1 (3.2) 1 (3.2) 2 (6.5) 1 (3.2) 1 (3.2)
Values in parentheses are percentages.
pressure monitoring before surgery. A further patient presented with a Chiari malformation. Eight patients had a transcranial procedure before PD (3 patients had fronto-orbital advancement and remodeling [FOAR], 2 had a third ventriculostomy, 1 had FOAR with later monobloc advancement, 1 had posterior calvarial release with later monobloc advancement, and 1 had calvarial expansion with later monobloc advancement and subsequent foramen magnum decompression). Two of these patients had also undergone ventriculoperitoneal (VP) shunting for hydrocephalus. Six patients had raised ICP, 2 of whom also had a Chiari malformation. The remaining 2 patients had a Chiari malformation only. The median age at PD for this group was 48 months old (range, 5–154 months). Two patients with raised ICP had FMD at the same time as PD. During surgery, 6 patients sustained a minor dural injury, which was immediately repaired. Patient 21 experienced a bleed from the inferior margin of the osteotomy in the midline near the torcula during surgery. This was controlled, but the procedure was converted from a hinge to a push-back technique to avoid any further dissection in the region. He had an uneventful recovery. Twenty-five patients required a blood transfusion. The patients younger than 1 year were more likely to require transfusion (18/19 patients) than the older children (5/12 patients) (P = 0.002, Fisher exact test). The mean (SD) transfusion volume of those receiving blood was 270 (60) mL. The mean hospital stay was 10.5 days (range, 4–30 d). The patients were liable to stay longer if they came from a more geographically remote area, had comorbidities, encountered postoperative complications, or had parents who required more supervised practice in activating the distractors. Twenty-eight patients (90%) were distracted to the full 20 mm for a mean of 22 days (range, 17–30 d). Three patients had distraction halted early because of persistent cerebrospinal fluid (CSF) leakage in 2 patients (patients 18 and 25) and threatened wound breakdown in 1 patient (patient 17). Distractors were removed after a mean consolidation period of 59 days (range, 0–133 d). Nine patients (29.0%) had distractors removed within 30 days of completing full distraction, with 2 patients having their devices removed within 24 hours of completion. A clinical and radiologic improvement in head shape was found in all but 1 patient at follow-up (excluding patient 18 in whom PD was abandoned after a significant CSF leak and a posterior-lateral expansion was performed). Patient 13 had a push-back procedure with 2 distractors that initially corrected her occipital flattening. However, during the consolidation phase of 102 days, she was persistently nursed in a supine position. This caused flexion at the patent lambdoid sutures and resulted in a relapse of the deformity,
Lessons From Posterior Distraction
accentuated by widening of the calvarium at the level of the osteotomy, maintained by the in situ distraction devices. A second patient (patient 30) experienced a failure of the activation arm of 1 of her 2 distractors. This left her with a 5-mm discrepancy in the final distraction distance between the left and right osteotomy margins. Although this difference was apparent radiologically, it was not noticeable on clinical examination. Nine patients had proven raised ICP on intraparenchymal pressure monitoring before PD. All patients experienced a resolution of their symptoms and signs after surgery. However, 3 of these patients manifested new signs and symptoms suggestive of raised ICP 2 or more years after distraction. One of them (patient 14) was found to have a normal ICP on intraparenchymal monitoring. Patient 25, a patient with Pfeiffer syndrome, had raised ICP confirmed on monitoring and then underwent VP shunt revision followed by calvarial expansion. The third patient (patient 8) did not have ICP monitoring but underwent FOAR on the basis of clinical and radiologic findings. A further patient (patient 4) with presumed, but not proven, raised ICP before PD was found on monitoring after surgery to have raised ICP and underwent posterior remodeling and expansion 6 months later. Of the 5 patients with a Chiari type 1 malformation, 1 resolved, 1 improved, and 3 remained unchanged on imaging after surgery. One patient (patient 25), with a malformation that did not improve after surgery, underwent PD combined with FMD, whereas the remaining 4 patients were treated by PD only. Patient 4 (previously described), whose malformation was also unaffected by PD, also had an FMD some 2 years later after manifesting symptoms consistent with brain stem compression. Nineteen patients (61.3%) experienced 1 or more unplanned events in the postoperative period, with 7 patients (19.4%) having at least 1 additional unplanned general anesthetic procedure (Tables 3, 4). Wound infection was the commonest complication, occurring in 9 patients (29.0%). Three patients had 1 of 2 distractors removed during the distraction phase for infection, with distraction continued to completion using the remaining device, which was then removed after a consolidation phase. One patient had both distractors removed immediately after the distraction phase. The fifth patient required removal of 1 of 2 distractors 9 days into the consolidation phase. All wound infections settled once the devices were removed. No bone infection occurred. A full distraction of 20 mm was achieved in all patients. Device-related necrosis of scalp tissue, either overlying the mechanism or beneath the activation arm, occurred in 7 patients, requiring the early removal of both devices during the consolidation phase in 4 children. All areas of necrosis were debrided and successfully primarily closed. Cerebrospinal fluid leaks necessitating an early return to the operating theater occurred in 2 patients after PD. Both patients had raised ICP before surgery, and both had had previous transcranial procedures. Patient 25 had a small dural injury repaired at PD, which leaked after distraction. It was successfully rerepaired, and distraction recommenced at a reduced rate with an early termination. Patient 18 was not recognized to have a dural injury during surgery. She developed a brisk CSF leak on day 4 after distraction and, at exploration, was found to have an approximately 3-cm dural tear parallel to the left lateral osteotomy. This was repaired, but distraction was abandoned, and a formal posterior-lateral expansion was performed on day 14. Two other patients had a small dural breach repaired at distractor placement in the setting of raised ICP. Both were distracted without incident. One further patient (patient 16) who underwent secondary PD for raised ICP sustained a minor dural injury at distractor removal. She had a persistent CSF leak that failed to settle with extraventricular
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1723
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
Thomas et al
TABLE 3. Unplanned Events by Patient Patient Postoperative Complications
Early Return to Theater Interval, d
4
Distractor exposure Wound infection
No Yes
— 26
6
Distractor exposure
Yes
52
7
Wound infection
Yes
21
8 9
No Yes
— 56
13 16 17 18
Wound infection Wound infection; skin necrosis beneath both distractor arms Skin necrosis beneath 1 distractor arm Persistent CSF leak after distractor removal Threatened wound breakdown Significant dural tear (persistent CSF leak)
No No No Yes
— — — 7
19
Threatened necrosis of skin over distractors
Yes
20
21
Wound infection
Yes
11
22 24 25 27
Wound infection Distractor exposure Persistent CSF leak; wound infection Distractor exposure
No No Yes Yes
— — 7 59
29
Wound infection
Yes
13
30
Distractor failure
No
—
31
Wound infection
Yes
15
2
Procedure — Remove distractor (1/2) Remove distractors (both) Remove distractors (both) — Remove distractors (both) — — — Remove distractor (1/2); repair dura Remove distractors (both) Remove distractor (1/2) — — Repair dura Remove distractors (both) Remove distractor (1/2) — Remove distractor (1/2)
Final Outcome Healed, no relapse Raised ICP diagnosed 3 mo after surgery: posterior remodeling Healed, no relapse Healed, no relapse Healed, no relapse Healed, no relapse Relapse of occipital flattening during consolidation period EVD then VP shunt, CSF leak resolved Distraction halted at 13 mm Remove remaining distractor and posterior-lateral expansion (day 14) Healed, no relapse Healed, no relapse Healed, Healed, Healed, Healed,
no no no no
relapse relapse relapse relapse
Healed, no relapse 5-mm discrepancy between distractors, not clinically appreciable Second distractor became infected and also removed early; no relapse
EVD, external ventricular drain.
drainage. She was found to have previously undiagnosed hydrocephalus and ceased to leak after VP shunting. Patient age at PD, technique of PD (hinge versus push-back), or whether PD was the primary or the secondary transcranial procedure did not have a statistically significant effect on the incidence of any complication in this series. Eleven patients, who had PD as a primary procedure, proceeded to FOAR at a mean of 25 months later (range, 5–50 mo). The procedure was performed through the same zigzag coronal incision used at PD. The remodeled orbital bar and the frontal complex were advanced a mean of 15.5 mm (range, 8–20 mm). In 6
patients, the coronal wound was found to be abnormally tight at closure, requiring greater undermining of the posterior galeal flap than would normally be performed in a routine FOAR. The coronal incision was usually closed with absorbable suture material. However, 2 patients required the use of a nonabsorbable monofilament suture to achieve satisfactory closure. Of the remaining 4 patients with a tight wound closed with absorbable sutures, 2 experienced a dehiscence that required a general anesthetic debridement and reclosure, with 1 patient having a full-thickness skin graft to the site. A further patient had a patchy superficial necrosis of the scalp along the wound margin that healed by secondary intent.
TABLE 4. Summary of Unplanned Postoperative Events Complication
Total
Wound infection
(29.0) (22.6) (3.2) (9.7) (3.2) 21 (61.3)
9 Distractor-induced tissue necrosis*†‡ 7 Threatened wound breakdown 1 CSF leak§ 3 Distractor failure 1 Total complications Total patients (N = 31) 19
Early Return to Theater
Early Removal of 1 or Both Distractors
5 4 0 2 0
(16.1) (12.9) (0.0) (6.5) (0.0) 11 11 (35.5)
5 4 0 1 0
(16.1) (12.9) (0.0) (3.2) (0.0) 10 10 (32.3)
Unplanned Further General Anaesthetic Procedure 4 0 0 3§ 0
(12.9) (0.0) (0.0) (9.7) (0.0) 7 7 (19.4)
Values in parentheses are percentages. *Distractor-induced tissue necrosis includes distractor exposure cases. †Distractor-induced tissue necrosis includes threatened skin necrosis (1 case). ‡One patient (patient 13) who had distractor-induced tissue necrosis also experienced a relapse of her deformity after distraction, during the consolidation phase. §Includes patient 16, who had a CSF leak after distractor removal.
1724
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
DISCUSSION Since 2009, several series of patients with turribrachycephaly and syndromic craniosynostosis treated by PD have been published.12–14,16–19 By virtue of the fact that PD is a relatively novel technique, most of these series are small, with only 2 containing more than 10 patients treated by PD.16,19 Differences between studies with regard to surgical technique, device type, and distraction protocol and the inevitable variation among patients within series make it difficult to draw useful conclusions concerning management approaches when studies are small. By contrast, the consecutive series of 31 patients we present here is the largest managed by a single unit in the literature to date. Previous studies that have simulated, estimated through modeling, or directly computed the increase in intracranial volume from PD indicate that, for 20 mm of distraction, which most of our patients experienced, the intracranial volume will have increased by approximately 20%.9,11,16,17 Unremarkably, intracranial volume increases with greater distraction distance, but the final distraction length is not the sole determinant of volume increase. As Nowinski and colleagues11 have pointed out, the position of the inferior osteotomy is also an important factor, although the benefit of increased volume from a lower osteotomy site has to be balanced with the increased risk for significant hemorrhage from venous sinuses within the posterior fossa. The patients did not routinely undergo three-dimensional computed tomographic imaging after PD in our series, to minimize their radiation exposure.20 As a result, the increase in intracranial volume has not been calculated. Plain lateral skull radiographs were used routinely to objectively assess the advancement gap during and after distraction. These were repeated in follow-up when necessary. The patients experienced an increase in skull length largely equivalent to their final advancement distance, with a corresponding subjective improvement in morphology. However, we believe that simple morphometric measurements do not adequately describe the complex calvarial deformity of most patients with syndromic craniosynostosis and hence are of little use in objectively assessing the outcomes of PD. The use of nonirradiating modalities to assess volumetric changes, such as three-dimensional laser scanning, hold promise but are currently expensive, are time consuming, and require operator technical expertise to be reliable. Furthermore, the maximum advancement achievable from PD, 20 mm in this series, is limited by device design and tissue factors. Increasing skull length by such a distance in children with severe turribrachycephaly is usually not sufficient to fully correct their deformity. Rather, the primary goals of PD are to redirect calvarial growth along the fronto-occipital plane, to provide sufficient intracranial space to defer FOAR surgery, and to prevent raised ICP from developing or to treat it, if present. In our series, secondary FOAR was deferred to a mean of 2 years after surgery. All patients with proven raised ICP preoperatively had no clinical symptoms or signs of intracranial hypertension after distraction. However, PD does not prevent a rise in ICP in perpetuity as evidenced by the 2 patients who underwent further surgery for new-onset raised ICP more than 2 years after PD. In addition, 1 further patient with Apert syndrome, severe brachycephaly, and a Chiari malformation (patient 4) was found to have raised ICP within 6 months of PD, although his situation was complicated by earlier hydrocephalus necessitating a third ventriculostomy. In this study, 4 patients with raised ICP had minor dural injures that were identified and repaired during device placement. Distraction in the presence of raised ICP and a dural tear potentially risk causing a persistent CSF leak because CSF pressure with distraction-induced tension across the wound might prevent dural healing. Interestingly, the 1 patient who developed a persistent CSF leak was the only one to be distracted on the table (by 5 mm). He
Lessons From Posterior Distraction
and 2 of the remaining 3 patients began distraction per protocol the following day. The final patient waited 4 days before distraction was started, on account of her dural injury. With hindsight, it is our opinion that patients who have intracranial hypertension and a dural injury should not be distracted on the table. They would also benefit from a longer latency period before commencing distraction. We encountered 1 delayed CSF leak associated with a significant dural tear that required distraction to be abandoned, in a patient with raised ICP who was not identified to have a dural injury during device placement (patient 18). She had also been distracted by 5 mm on the table. Cerebrospinal fluid leaks not attributable to recognized dural injury have occurred with a similar incidence in other series.13,16 Goldstein and colleagues16 hypothesized that the 2 leaks in their series were caused by “scratches” from screw tips. As a result, they changed to self-tapping screws and began to place an absorbable gelatine sponge beneath the screws to protect the dura. We have always routinely placed Spongostan (Ferrosan A/S, Soeborg, Denmark) beneath the bone at the distractor sites because of similar concerns.12 Recent technological advances that eliminate the need for screw fixation of devices, such as ultrasound-activated pin osteosynthesis (SonicWeld Rx, KLS Martin Group, Tuttlingen, Germany), could prevent this complication entirely.21 The latency period of 24 to 48 hours used in this series is briefer than the 72-hour to 5-day period used by others.13,14,16,19,22 Three- to 5-day latency periods are routinely used elsewhere in the skeleton and are derived from Ilizarov and Ledyaev's23 original work and that of others more recently (summarized by McCarthy et al,24 2001). The length of the latency period is an attempt to balance the advantages of a greater callus mass, gained through a longer latency period, with the increasing risk for premature consolidation that delaying distraction entails. The well-documented dural capacity for osteoinduction in infants and younger children strongly suggests that distraction of callus is of secondary importance in calvarial osteogenesis in PD.25–27 We observed consistently reliable ossification of the distraction defect despite the reduced latency period (as exampled by Fig. 1). We can identify little benefit in using a latency period designed for optimal callus formation in long bones in the distraction of the calvarium. We do support an extended latency period in circumstances of dural injury, discussed previously, or where PD is the second transcranial procedure and the wound closure is tight, with a consequent risk for wound dehiscence with early distraction. In this series, 2 surgical approaches were used, the hinge and push-back techniques. As we have discussed elsewhere, the hinge technique facilitates the rotation of the posterior panel backward and downward, dropping the height of the vertex.12 It is our preferred approach to correcting turribrachycephaly. We favor the push-back technique in cases in which expansion of the posterior cranial fossa is an important indication for surgery, as in the presence of a Chiari malformation, or in primarily brachycephalic deformities in which cranial height is already normal. Effective PD undoubtedly has the capacity to improve established Chiari malformations, as we and others have found.14,28 Even in the 3 patients in whom the malformation did not improve, in this series, there has been no radiologic deterioration during follow-up (range, 2–4.5 y), although 1 child with a complex condition later had an FMD for suspected brain stem compression. The number of distractor devices used in PD varies between series, ranging from 4 to a single device per patient.12–14,16,17 The use of more than 2 distractors is probably unnecessary because it raises the cost and the complexity of the procedure and increases the risk for developing conflicting distraction vectors and other complications.11 Initially, we favored using only 1 distractor in the hinge technique, to minimize the aforementioned risks. However, since 2010, we have routinely used 2 devices in both procedures. Although we have shown symmetric distraction with a single
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1725
Thomas et al
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
FIGURE 1. A, Preoperative three-dimensional computed tomography of patient 19, who had proven raised ICP. B, Day 18 after insertion of distractors, 18 mm of distraction. The distractors were removed immediately at the end of distraction, day 20. C, Sixty days after device removal. D, One hundred fifty-eight days after device removal. The distraction osseous defect continues to ossify without relapse.
device in the hinge technique, 2 devices give finer control over the distraction, helping to ensure symmetry. Further, 2 devices allow for redundancy. In the 1 case of distractor failure and 3 instances in which 1 distractor was removed during the distraction phase, the remaining device enabled distraction to be continued to completion. Wound infections occurred in 29% of the patients in this series, with 16% returning to the operating theater earlier than anticipated. Although this rate was higher than that reported by other authors in PD, it is similar to that observed in distraction procedures in other areas of the craniofacial skeleton.15,29 Importantly, wound infections did not prevent full distraction being achieved or lead to serious sequelae, such as osteomyelitis. Undoubtedly, the increased wound infection rate we observed was due to the presence of the distractor devices because we have not seen similar rates in other posterior cranial procedures, despite an identical antibiotic prophylaxis regimen. Although the PD series currently in the literature do not discuss antibiotic prophylaxis, the regimen used in this study is similar to that reported in other transcranial craniofacial operation series30 and exceeds in duration the 48 hours of prophylaxis that Ahmad and colleagues have used in frontofacial monobloc distraction advancement procedures.31 Tissue hypoxia and frank necrosis of the scalp caused by distractor devices are likely to be important factors in the pathogenesis of infection in PD. We observed tissue necrosis to occur both beneath the activation arms and also over the hub covering the terminal end of the central rod, buried beneath the scalp. Because of the spherical nature of the cranium and the linear vector of distraction, the scalp becomes increasingly tented over the high points of the distractor, principally the hub and the central rod (Fig. 2). Likewise, scalp tissue beneath the activation arm is increasingly compressed, which occurred in this study despite the use of flexible activation arms. This highlights the likely importance of both device design and placement. It may also explain the differences in infection rates between our series and others because only 1 other group published in the literature has used the same device.11,32
1726
We believe that low-profile devices that are adapted to the convex contour of the posterior cranium will significantly reduce both wound infection and device-induced tissue necrosis rates. Modification of devices to use soft flexible components, wherever practicable, is also likely to be of benefit. We plan to use such a device, currently being designed, in the future. Early distractor removal with a reduced or absent consolidation period was not associated with relapse of the deformity in this series (Fig. 1). This indicates that new bone formation is not the principle mechanism by which the distraction gap is maintained in PD. Instead, distraction allows the previously constricted intracranial contents to expand behind the distracted posterior panel and splint it in its new position.12,32 Indeed, the only significant relapse we encountered occurred in a patient who had a consolidation period of 102 days. The prolonged presence of the devices may even have contributed to the relapse because the patient experienced scalp necrosis beneath the activation arms of both distractors and was therefore nursed supine, to protect these areas. Furthermore, 6 of the 7 episodes of scalp necrosis we observed occurred during the consolidation phase. In these cases, early distractor removal may well have successfully salvaged the compromised tissue, as it did in patient 19. We therefore advocate the early removal of devices, at 1 month after the termination of distraction, because we can find no advantage in the traditional consolidation period of 2 to 3 months, whereas the disadvantages of a prolonged consolidation period in terms of the risk for tissue necrosis, wound infection, and parental anxiety are quite apparent. A clear advantage of PD over posterior advancement techniques is that it permits closure of the scalp without tension.13,14,22 During distraction, the overlying soft tissues are expanded to ultimately accommodate a greater cranial volume than they could have done in a single-stage procedure. Although the scalp expands sufficiently to encompass the distracted calvarium, it remains relatively tight after distraction. In more than half of the patients who underwent FOAR after PD, the scalp was found to be unusually tight, despite a mean interval of more than 2 years between PD and FOAR. The overall reoperation rate in this group was 18%, entirely due to wound dehiscence. We now anticipate this problem and have modified our wound closure for FOAR after PD. The posterior galeal flap is raised as far backward as the origin of occipitalis, and the bicoronal incision is then closed with continuous nonabsorbable
FIGURE 2. Tenting of the scalp over the central rod of a fully extended distractor (black arrows), with exposure of the prominent distractor hub (patient 2).
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
monofilament suture, rather than the absorbable sutures we routinely use for scalp closure in other circumstances. An appealing alternative approach to avoiding a tight coronal wound at FOAR is to advance the fronto-orbital construct via distraction only once the incision has been closed.33 We plan to pursue this approach in some of the patients presented here in the future. In conclusion, in this large consecutive series of 31 patients treated by PD, unanticipated events occurred in 61.3% of the patients, with 32.2% experiencing an earlier return to the operating theater than planned and 19.4% undergoing 3 or more general anesthetic procedures. Wound infection (29.0%) and tissue necrosis caused by the distractor (22.6%) were the commonest complications, but all patients who endured these were fully distracted and had no long-term sequelae. Cerebrospinal fluid leaks were rarer (6.5%) after distractor placement but had a more significant impact with early cessation of distraction in 1 case and abandonment of PD entirely in the other. Distraction was stopped early in only 1 other patient who had a threatened wound dehiscence. In all but 1 patient who completed distraction, there was an improvement in cranial morphology and resolution of clinical indicators of raised ICP, where present. In the patients with a Chiari malformation, 40% improved after surgery. Twenty-nine percent of the patients had distractors removed within 30 days of completing full distraction. None of these patients experienced a relapse. Our 5-year experience of PD has led us to make important changes to our distraction protocol. We advocate short latency periods of 48 hours or less, in the absence of a tight wound or recognized dural tear associated with raised ICP. The use of 2 devices allows for redundancy; thus, where 1 device needs to be removed early, full distraction can still be achieved. An abbreviated consolidation period does not result in relapse and, if anything, is likely to reduce complication rates. Posterior distraction is an important advance in the treatment of patients with severe brachycephaly or turribrachycephaly, but it is not without risks and forms only part of an integrated management approach to this sometimes challenging group of patients.
REFERENCES 1. Collmann H, Sorensen N, Krauss J. Hydrocephalus in craniosynostosis: a review. Childs Nerv Syst 2005;21:902–912 2. Marucci DD, Dunaway DJ, Jones BM, et al. Raised intracranial pressure in Apert syndrome. Plast Reconstr Surg 2008;122:1162–1168 3. Thompson DN, Harkness W, Jones BM, et al. Aetiology of herniation of the hindbrain in craniosynostosis. An investigation incorporating intracranial pressure monitoring and magnetic resonance imaging. Pediatr Neurosurg 1997;26:288–295 4. Cinalli G, Spennato P, Sainte-Rose C, et al. Chiari malformation in craniosynostosis. Childs Nerv Syst 2005;21:889–901 5. Marchac D, Renier D, Broumand S. Timing of treatment for craniosynostosis and facio-craniosynostosis: a 20-year experience. Br J Plast Surg 1994;47:211–222 6. Wall SA, Goldin JH, Hockley AD, et al. Fronto-orbital re-operation in craniosynostosis. Br J Plast Surg 1994;47:180–184 7. Wong GB, Kakulis EG, Mulliken JB. Analysis of fronto-orbital advancement for Apert, Crouzon, Pfeiffer, and Saethre-Chotzen syndromes. Plast Reconstr Surg 2000;105:2314–2323 8. Woods RH, Ul-Haq E, Wilkie AO, et al. Reoperation for intracranial hypertension in TWIST1-confirmed Saethre-Chotzen syndrome: a 15-year review. Plast Reconstr Surg 2009;123:1801–1810 9. Choi M, Flores RL, Havlik RJ. Volumetric analysis of anterior versus posterior cranial vault expansion in patients with syndromic craniosynostosis. J Craniofac Surg 2012;23:455–458 10. Sgouros S, Goldin JH, Hockley AD, et al. Posterior skull surgery in craniosynostosis. Childs Nerv Syst 1996;12:727–733
Lessons From Posterior Distraction
11. Nowinski D, Di Rocco F, Renier D, et al. Posterior cranial vault expansion in the treatment of craniosynostosis. Comparison of current techniques. Childs Nerv Syst 2012;28:1537–1544 12. Wiberg A, Magdum S, Richards PG, et al. Posterior calvarial distraction in craniosynostosis—an evolving technique. J Craniomaxillofac Surg 2012;40:799–806 13. White N, Evans M, Dover MS, et al. Posterior calvarial vault expansion using distraction osteogenesis. Childs Nerv Syst 2009;25:231–236 14. Steinbacher DM, Skirpan J, Puchala J, et al. Expansion of the posterior cranial vault using distraction osteogenesis. Plast Reconstr Surg 2011;127:792–801 15. Yonehara Y, Hirabayashi S, Sugawara Y, et al. Complications associated with gradual cranial vault distraction osteogenesis for the treatment of craniofacial synostosis. J Craniofac Surg 2003;14:526–528 16. Goldstein JA, Paliga JT, Wink JD, et al. A craniometric analysis of posterior cranial vault distraction osteogenesis. Plast Reconstr Surg 2013;131:1367–1375 17. Serlo WS, Ylikontiola LP, Lahdesluoma N, et al. Posterior cranial vault distraction osteogenesis in craniosynostosis: estimated increases in intracranial volume. Childs Nerv Syst 2011;27:627–633 18. Taylor JA, Derderian CA, Bartlett SP, et al. Perioperative morbidity in posterior cranial vault expansion: distraction osteogenesis versus conventional osteotomy. Plast Reconstr Surg 2012;129:674e–680e 19. Ylikontiola LP, Sandor GK, Salokorpi N, et al. Experience with craniosynostosis treatment using posterior cranial vault distraction osteogenesis. Ann Maxillofac Surg 2012;2:4–7 20. Mathews JD, Forsythe AV, Brady Z, et al. Cancer risk in 680 000 people exposed to computed tomography scans in childhood or adolescence: data linkage study of 11 million Australians. BMJ 2013;346:f2360 21. Arnaud E, Renier D. Pediatric craniofacial osteosynthesis and distraction using an ultrasonic-assisted pinned resorbable system: a prospective report with a minimum 30 months’ follow-up. J Craniofac Surg 2009;20:2081–2086 22. Derderian CA, Bastidas N, Bartlett SP. Posterior cranial vault expansion using distraction osteogenesis. Childs Nerv Syst 2012;28:1551–1556 23. Ilizarov GA, Ledyaev VI. The replacement of long tubular bone defects by lengthening distraction osteotomy of one of the fragments. 1969. Clin Orthop Relat Res 1992;280:7–10 24. McCarthy JG, Stelnicki EJ, Mehrara BJ, et al. Distraction osteogenesis of the craniofacial skeleton. Plast Reconstr Surg 2001;107:1812–1827 25. Greenwald JA, Mehrara BJ, Spector JA, et al. Biomolecular mechanisms of calvarial bone induction: immature versus mature dura mater. Plast Reconstr Surg 2000;105:1382–1392 26. Greenwald JA, Mehrara BJ, Spector JA, et al. Immature versus mature dura mater: II. Differential expression of genes important to calvarial reossification. Plast Reconstr Surg 2000;106:630–638 27. Yang JY, Yang WG. Large scalp and skull defect in aplasia cutis congenita. Br J Plast Surg 2000;53:619–622 28. Ahmad F, Evans M, White N, et al. Amelioration of Chiari type 1 malformation and syringomyelia following posterior calvarial distraction in Crouzon’s syndrome—a case report. Childs Nerv Syst 2014;30:177–179 29. Dunaway DJ, Britto JA, Abela C, et al. Complications of frontofacial advancement. Childs Nerv Syst 2012;28:1571–1576 30. Fearon JA, Yu J, Bartlett SP, et al. Infections in craniofacial surgery: a combined report of 567 procedures from two centers. Plast Reconstr Surg 1997;100:862–868 31. Ahmad F, Cobb AR, Mills C, et al. Frontofacial monobloc distraction in the very young: a review of 12 consecutive cases. Plast Reconstr Surg 2012;129:488e–497e 32. Nowinski D, Saiepour D, Leikola J, et al. Posterior cranial vault expansion performed with rapid distraction and time-reduced consolidation in infants with syndromic craniosynostosis. Childs Nerv Syst 2011;27:1999–2003 33. Komuro Y, Shimizu A, Ueda A, et al. Whole cranial vault expansion by continual occipital and fronto-orbital distraction in syndromic craniosynostosis. J Craniofac Surg 2011;22:269–272
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1727
Lessons Learned in Posterior Cranial Vault Distraction Gregory P.L. Thomas, PhD, FRCS (Plast), Steven A. Wall, FCS(SA) Plast, Jayaratnam Jayamohan, FRCS, Shailendra A. Magdum, FRCS, Peter G. Richards, FRCS Akira Wiberg, MRCS, and David Johnson, DM, FRCS (Plast)
Introduction: Posterior distraction (PD) is rapidly emerging as an important technique to increase the intracranial volume and correct calvarial morphology in patients with severe brachycephaly or turribrachycephaly. Methods: A retrospective review was performed of all 31 patients who underwent PD at the Oxford Craniofacial Unit between 2007 and 2012. Results: Twenty-three patients (74.2%) underwent PD as a primary procedure at a median age of 8 months. Eight patients (25.8%) had PD as a secondary transcranial procedure at a median age of 48 months. Full distraction to 20 mm was achieved in 28 patients (90.3%). Of these, all but 1 demonstrated a significant improvement in morphology, with a resolution of the symptoms and signs of raised intracranial pressure in all proven to have it preoperatively. Unanticipated events occurred in 61.3% of patients, with 19.4% undergoing one or more unplanned procedures. Wound infection (29.0%) and tissue necrosis (22.6%) were the commonest. Cerebrospinal fluid leaks were rarer (6.5%) but prevented full distraction. Nine patients (29.0%) had a consolidation period of less than 30 days without experiencing relapse. In 11 patients who had a later frontoorbital advancement and remodeling, wound closure was tight, resulting in dehiscence in 3 cases (27.3%). Conclusions: Posterior distraction is an effective procedure in the management of severe brachycephaly or turribrachycephaly but has associated risks. Our protocol has evolved with experience to favor a reduced latency period and consolidation phase and the use of 2 distractor devices. Key Words: Posterior distraction, distraction osteogenesis, syndromic craniosynostosis, raised intracranial pressure, Chiari malformation, wound infection, tissue necrosis (J Craniofac Surg 2014;25: 1721–1727)
P
atients with syndromic craniosynostosis manifesting as turribrachycephaly or severe brachycephaly typically have an underdeveloped posterior cranial fossa and are at increased risk for From the Oxford Craniofacial Unit, Oxford University Hospitals NHS Trust, Oxford, UK. Received September 18, 2013. Accepted for publication April 7, 2014. Address correspondence and reprint requests to David Johnson, DM, FRCS, Oxford Craniofacial Unit, Oxford University Hospitals NHS Trust, Headley Way, Oxford OX3 9DU, UK; E-mail: [email protected] The authors report no conflicts of interest. Copyright © 2014 by Mutaz B. Habal, MD ISSN: 1049-2275 DOI: 10.1097/SCS.0000000000000995
having raised intracranial pressure (ICP), hydrocephalus, or a Chiari type 1 malformation.1–4 Early calvarial vault expansion is routinely performed in these patients to protect the rapidly growing brain from such hazards. Traditionally, this was achieved through primary fronto-orbital advancement.5 However, early fronto-orbital advancement is prone to relapse in syndromic cases, necessitating a revision procedure, and does not reliably prevent the later development of raised ICP.6–8 Expansion of the posterior calvarial vault is increasingly favored as an alternative primary procedure because it not only directly addresses the posterior fossa but also has the potential to deliver a far greater increase in intracranial volume.9–11 Posterior release has been practiced for many years, and although often effective, results can be unpredictable, with approximately 30% of patients having an unfavorable outcome.12 This is likely due to the fact that the technique relies on brain expansion to drive the released occipitoparietal bone flap backward. A prolonged supine posture in the immediate postoperative phase can impede expansion and result in early ossification of the craniotomy site.11 The adoption of distraction techniques in the posterior cranium negates this issue because the bone flap is fixed by the distractor device. It is not surprising that posterior calvarial vault expansion by distraction has been widely taken up since it was described by White and colleagues13 in a series of 6 patients in 2009 because it gives the surgical team unparalleled control over the expansion. Furthermore, in posterior distraction (PD), provided that the bone flap is not dissected free of the dura, the opportunities for dural injury and significant hemorrhage are fewer, the bone remains vascularized, and the dead space beneath the bone is eliminated.13,14 This should translate into a decreased operating time, reduced blood loss, and reduced overall morbidity, consequently leading to briefer hospital admissions.14 Posterior distraction does have some disadvantages, not least that it inevitably requires a second surgical procedure at which the distractor devices are removed. Other complications are recognized to occur, particularly wound infection involving the device and necrosis of the tissue overlying the device, leading to its exposure.12,15 The complication profile of the procedure is likely to change as surgeons and units increase in experience and devices along with distraction protocols are modified accordingly. We present a single large consecutive series of 31 patients operated on in our unit between 2007 and December 2012 and describe our evolving approach to PD. In this study, patients underwent 1 of 2 PD techniques. Those with significant turricephaly had PD using the “hinge” technique (described in the “Methods” section), in which the posterior panel was rotated posteriorly and downward during distraction, with the aim of reducing the vertex height in addition to expanding the posterior vault. In the patients in whom expansion of the posterior fossa was the primary goal, the released posterior panel was left unhinged (the “push-back” technique), enabling it to be displaced in line with the distraction vector without rotation, thereby increasing the expansion of the posterior cranial fossa.12
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1721
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
Thomas et al
METHODS A retrospective review was performed of the case notes of all patients who underwent PD under the care of the Oxford Craniofacial Unit between January 2007 and December 2012. Adequate data were available for all 31 cases identified. The operative technique used in this series has been described previously.12 Briefly, in all cases, the calvarium was approached via a coronal zigzag incision. An osteotomy was performed extending from the vertex to below the torcula, circumscribing a posterior panel that remained attached to the underlying dura. Either 1 or 2 KLS Martin distractor devices (KLS Martin Group, Tuttlingen, Germany) were placed across the craniotomy in the temporoparietal region. Arnaud cranio-orbital distractors were used in 29 cases, whereas a modification of this device was used in the remaining 2 cases. In 19 cases, the posterior panel was hinged inferiorly in the midline with either a stainless steel wire or PDS suture in a figure-of-8 configuration (hinge technique). In the remaining 11 cases, the posterior panel was left unhinged, and 2 distractors were fitted to displace the panel directly backward (push-back technique). In 19 cases, initial distraction was started on the operating table, with a mean distraction distance of 2.6 mm (range, 1–6 mm). In the other 12 cases, distraction was started after 1 to 2 days' latency, except in 1 case in which distraction was started on the fourth day after surgery. Distraction then proceeded at approximately 1 mm per day until the devices achieved their maximum distraction distance of
20 mm. Once this was reached, the distractor activation arms were usually trimmed as close to the scalp as feasible to reduce the chance of the arm snagging accidentally. This was performed in the outpatient clinic without recourse to anesthesia. After a consolidation period, the devices were removed under general anesthetic. Serial skull radiographs were used to monitor distraction progress, in addition to standard clinical observations. All patients received broad-spectrum antibiotic prophylaxis (amoxicillin and clavulanic acid) at anesthetic induction and for 5 days after surgery. The patients who presented postoperatively with signs of a wound infection received empirical intravenous or oral antibiotic treatment with regular wound toilet once microbiologic samples had been taken. Antibiotic treatment was modified according to microbiologic culture results, as guided by a specialist microbiologist.
RESULTS In total, 31 patients (21 girls, 10 boys) underwent PD. All but 1 patient had a syndromic condition causing craniosynostosis, confirmed on genetic screening (Table 1). All patients had brachycephaly or progressive turribrachycephaly secondary to bicoronal or multisuture craniosynostosis (Table 2). Twenty-three patients were treated by PD as a primary procedure at a median age of 8 months old (range, 5–39 mo). Three of these patients had raised ICP diagnosed on intraparenchymal
TABLE 1. Demographic Data by Diagnosis
Patient
Diagnosis
Raised ICP/ Chiari Malformation
Apert — Apert — Apert — Apert Chiari Apert — Apert — Apert — Apert Raised ICP Apert — Apert — Apert Chiari Apert Raised ICP Craniofrontonasal — Crouzon Raised ICP Crouzon Raised ICP; Chiari Crouzon Raised ICP Crouzon Raised ICP Hypophosphatasia Raised ICP Meunke — Meunke — Meunke — Meunke — Meunke — Pfeiffer Raised ICP; Chiari Saethre-Chotzen — Saethre-Chotzen — Saethre-Chotzen — TCF12 — TCF12 — Bilambdoid, Chiari sagittal (NS)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Age, mo
No. Distractors: Technique
Distraction Period, d
Consolidation Period, d
Advancement, mm
First Transcranial Operation
Hospital Stay, d
Unplanned Postoperative Event
5 5 5 5 6 7 8 10 14 17 24 43 9 39 52 72 154 7 5 5 6 9 10 45 7 8 12 9 14 16
2: Hinge 1: Hinge 1: Hinge 2: Push-back 1: Hinge 2: Hinge 2: Hinge 1: Hinge 2: Hinge 2: Push-back 2: Push-back 1: Hinge 2: Push-back 1: Hinge 2: Push-back 2: Push-back 2: Push-back 2: Hinge 1: Hinge 2: Push-back 2: Hinge 2: Hinge 2: Push-back 1: Hinge 2: Hinge 2: Hinge 2: Hinge 2: Push-back 2: Hinge 2: Push-back
30 19 30 17 18 19 20 19 29 20 22 23 20 26 19 30 7 20 19 18 21 28 19 19 24 22 24 21 24 21
119 121 83 67 87 23 1 30 27 93 86 71 102 76 23 57 0 0 47 34 77 133 75 77 29 37 67 123 46 13
20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 13 7 20 20 20 20 20 20 17 20 20 20 20 20 20
Yes Yes Yes No Yes Yes Yes Yes Yes Yes No No Yes Yes No No No Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes Yes
8 8 6 30 10 22 5 10 8 11 6 7 8 6 10 13 25 26 8 7 4 11 9 14 8 7 8 4 8 5
— Yes — Yes — Yes Yes Yes Yes — — — Yes Yes — Yes Yes Yes — Yes Yes — Yes Yes — Yes — Yes Yes Yes
NS, no syndromic diagnosis; TCF12, TCF12-related craniosynostosis.
1722
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
TABLE 2. Indication for PD by Technique Operative Technique Indication for Surgery Progressive turribrachycephaly Severe brachycephaly Raised ICP with turribrachycephaly Raised ICP with brachycephaly Chiari malformation with turribrachycephaly Chiari malformation with brachycephaly Chiari + raised ICP with turribrachycephaly Chiari + raised ICP with brachycephaly Total
Hinge
Push-Back
Total (N = 31)
8 6 3 1 0 0 1 0 19
1 4 3 0 1 2 0 1 12
9 (29.0) 10 (32.3) 6 (19.4) 1 (3.2) 1 (3.2) 2 (6.5) 1 (3.2) 1 (3.2)
Values in parentheses are percentages.
pressure monitoring before surgery. A further patient presented with a Chiari malformation. Eight patients had a transcranial procedure before PD (3 patients had fronto-orbital advancement and remodeling [FOAR], 2 had a third ventriculostomy, 1 had FOAR with later monobloc advancement, 1 had posterior calvarial release with later monobloc advancement, and 1 had calvarial expansion with later monobloc advancement and subsequent foramen magnum decompression). Two of these patients had also undergone ventriculoperitoneal (VP) shunting for hydrocephalus. Six patients had raised ICP, 2 of whom also had a Chiari malformation. The remaining 2 patients had a Chiari malformation only. The median age at PD for this group was 48 months old (range, 5–154 months). Two patients with raised ICP had FMD at the same time as PD. During surgery, 6 patients sustained a minor dural injury, which was immediately repaired. Patient 21 experienced a bleed from the inferior margin of the osteotomy in the midline near the torcula during surgery. This was controlled, but the procedure was converted from a hinge to a push-back technique to avoid any further dissection in the region. He had an uneventful recovery. Twenty-five patients required a blood transfusion. The patients younger than 1 year were more likely to require transfusion (18/19 patients) than the older children (5/12 patients) (P = 0.002, Fisher exact test). The mean (SD) transfusion volume of those receiving blood was 270 (60) mL. The mean hospital stay was 10.5 days (range, 4–30 d). The patients were liable to stay longer if they came from a more geographically remote area, had comorbidities, encountered postoperative complications, or had parents who required more supervised practice in activating the distractors. Twenty-eight patients (90%) were distracted to the full 20 mm for a mean of 22 days (range, 17–30 d). Three patients had distraction halted early because of persistent cerebrospinal fluid (CSF) leakage in 2 patients (patients 18 and 25) and threatened wound breakdown in 1 patient (patient 17). Distractors were removed after a mean consolidation period of 59 days (range, 0–133 d). Nine patients (29.0%) had distractors removed within 30 days of completing full distraction, with 2 patients having their devices removed within 24 hours of completion. A clinical and radiologic improvement in head shape was found in all but 1 patient at follow-up (excluding patient 18 in whom PD was abandoned after a significant CSF leak and a posterior-lateral expansion was performed). Patient 13 had a push-back procedure with 2 distractors that initially corrected her occipital flattening. However, during the consolidation phase of 102 days, she was persistently nursed in a supine position. This caused flexion at the patent lambdoid sutures and resulted in a relapse of the deformity,
Lessons From Posterior Distraction
accentuated by widening of the calvarium at the level of the osteotomy, maintained by the in situ distraction devices. A second patient (patient 30) experienced a failure of the activation arm of 1 of her 2 distractors. This left her with a 5-mm discrepancy in the final distraction distance between the left and right osteotomy margins. Although this difference was apparent radiologically, it was not noticeable on clinical examination. Nine patients had proven raised ICP on intraparenchymal pressure monitoring before PD. All patients experienced a resolution of their symptoms and signs after surgery. However, 3 of these patients manifested new signs and symptoms suggestive of raised ICP 2 or more years after distraction. One of them (patient 14) was found to have a normal ICP on intraparenchymal monitoring. Patient 25, a patient with Pfeiffer syndrome, had raised ICP confirmed on monitoring and then underwent VP shunt revision followed by calvarial expansion. The third patient (patient 8) did not have ICP monitoring but underwent FOAR on the basis of clinical and radiologic findings. A further patient (patient 4) with presumed, but not proven, raised ICP before PD was found on monitoring after surgery to have raised ICP and underwent posterior remodeling and expansion 6 months later. Of the 5 patients with a Chiari type 1 malformation, 1 resolved, 1 improved, and 3 remained unchanged on imaging after surgery. One patient (patient 25), with a malformation that did not improve after surgery, underwent PD combined with FMD, whereas the remaining 4 patients were treated by PD only. Patient 4 (previously described), whose malformation was also unaffected by PD, also had an FMD some 2 years later after manifesting symptoms consistent with brain stem compression. Nineteen patients (61.3%) experienced 1 or more unplanned events in the postoperative period, with 7 patients (19.4%) having at least 1 additional unplanned general anesthetic procedure (Tables 3, 4). Wound infection was the commonest complication, occurring in 9 patients (29.0%). Three patients had 1 of 2 distractors removed during the distraction phase for infection, with distraction continued to completion using the remaining device, which was then removed after a consolidation phase. One patient had both distractors removed immediately after the distraction phase. The fifth patient required removal of 1 of 2 distractors 9 days into the consolidation phase. All wound infections settled once the devices were removed. No bone infection occurred. A full distraction of 20 mm was achieved in all patients. Device-related necrosis of scalp tissue, either overlying the mechanism or beneath the activation arm, occurred in 7 patients, requiring the early removal of both devices during the consolidation phase in 4 children. All areas of necrosis were debrided and successfully primarily closed. Cerebrospinal fluid leaks necessitating an early return to the operating theater occurred in 2 patients after PD. Both patients had raised ICP before surgery, and both had had previous transcranial procedures. Patient 25 had a small dural injury repaired at PD, which leaked after distraction. It was successfully rerepaired, and distraction recommenced at a reduced rate with an early termination. Patient 18 was not recognized to have a dural injury during surgery. She developed a brisk CSF leak on day 4 after distraction and, at exploration, was found to have an approximately 3-cm dural tear parallel to the left lateral osteotomy. This was repaired, but distraction was abandoned, and a formal posterior-lateral expansion was performed on day 14. Two other patients had a small dural breach repaired at distractor placement in the setting of raised ICP. Both were distracted without incident. One further patient (patient 16) who underwent secondary PD for raised ICP sustained a minor dural injury at distractor removal. She had a persistent CSF leak that failed to settle with extraventricular
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1723
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
Thomas et al
TABLE 3. Unplanned Events by Patient Patient Postoperative Complications
Early Return to Theater Interval, d
4
Distractor exposure Wound infection
No Yes
— 26
6
Distractor exposure
Yes
52
7
Wound infection
Yes
21
8 9
No Yes
— 56
13 16 17 18
Wound infection Wound infection; skin necrosis beneath both distractor arms Skin necrosis beneath 1 distractor arm Persistent CSF leak after distractor removal Threatened wound breakdown Significant dural tear (persistent CSF leak)
No No No Yes
— — — 7
19
Threatened necrosis of skin over distractors
Yes
20
21
Wound infection
Yes
11
22 24 25 27
Wound infection Distractor exposure Persistent CSF leak; wound infection Distractor exposure
No No Yes Yes
— — 7 59
29
Wound infection
Yes
13
30
Distractor failure
No
—
31
Wound infection
Yes
15
2
Procedure — Remove distractor (1/2) Remove distractors (both) Remove distractors (both) — Remove distractors (both) — — — Remove distractor (1/2); repair dura Remove distractors (both) Remove distractor (1/2) — — Repair dura Remove distractors (both) Remove distractor (1/2) — Remove distractor (1/2)
Final Outcome Healed, no relapse Raised ICP diagnosed 3 mo after surgery: posterior remodeling Healed, no relapse Healed, no relapse Healed, no relapse Healed, no relapse Relapse of occipital flattening during consolidation period EVD then VP shunt, CSF leak resolved Distraction halted at 13 mm Remove remaining distractor and posterior-lateral expansion (day 14) Healed, no relapse Healed, no relapse Healed, Healed, Healed, Healed,
no no no no
relapse relapse relapse relapse
Healed, no relapse 5-mm discrepancy between distractors, not clinically appreciable Second distractor became infected and also removed early; no relapse
EVD, external ventricular drain.
drainage. She was found to have previously undiagnosed hydrocephalus and ceased to leak after VP shunting. Patient age at PD, technique of PD (hinge versus push-back), or whether PD was the primary or the secondary transcranial procedure did not have a statistically significant effect on the incidence of any complication in this series. Eleven patients, who had PD as a primary procedure, proceeded to FOAR at a mean of 25 months later (range, 5–50 mo). The procedure was performed through the same zigzag coronal incision used at PD. The remodeled orbital bar and the frontal complex were advanced a mean of 15.5 mm (range, 8–20 mm). In 6
patients, the coronal wound was found to be abnormally tight at closure, requiring greater undermining of the posterior galeal flap than would normally be performed in a routine FOAR. The coronal incision was usually closed with absorbable suture material. However, 2 patients required the use of a nonabsorbable monofilament suture to achieve satisfactory closure. Of the remaining 4 patients with a tight wound closed with absorbable sutures, 2 experienced a dehiscence that required a general anesthetic debridement and reclosure, with 1 patient having a full-thickness skin graft to the site. A further patient had a patchy superficial necrosis of the scalp along the wound margin that healed by secondary intent.
TABLE 4. Summary of Unplanned Postoperative Events Complication
Total
Wound infection
(29.0) (22.6) (3.2) (9.7) (3.2) 21 (61.3)
9 Distractor-induced tissue necrosis*†‡ 7 Threatened wound breakdown 1 CSF leak§ 3 Distractor failure 1 Total complications Total patients (N = 31) 19
Early Return to Theater
Early Removal of 1 or Both Distractors
5 4 0 2 0
(16.1) (12.9) (0.0) (6.5) (0.0) 11 11 (35.5)
5 4 0 1 0
(16.1) (12.9) (0.0) (3.2) (0.0) 10 10 (32.3)
Unplanned Further General Anaesthetic Procedure 4 0 0 3§ 0
(12.9) (0.0) (0.0) (9.7) (0.0) 7 7 (19.4)
Values in parentheses are percentages. *Distractor-induced tissue necrosis includes distractor exposure cases. †Distractor-induced tissue necrosis includes threatened skin necrosis (1 case). ‡One patient (patient 13) who had distractor-induced tissue necrosis also experienced a relapse of her deformity after distraction, during the consolidation phase. §Includes patient 16, who had a CSF leak after distractor removal.
1724
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
DISCUSSION Since 2009, several series of patients with turribrachycephaly and syndromic craniosynostosis treated by PD have been published.12–14,16–19 By virtue of the fact that PD is a relatively novel technique, most of these series are small, with only 2 containing more than 10 patients treated by PD.16,19 Differences between studies with regard to surgical technique, device type, and distraction protocol and the inevitable variation among patients within series make it difficult to draw useful conclusions concerning management approaches when studies are small. By contrast, the consecutive series of 31 patients we present here is the largest managed by a single unit in the literature to date. Previous studies that have simulated, estimated through modeling, or directly computed the increase in intracranial volume from PD indicate that, for 20 mm of distraction, which most of our patients experienced, the intracranial volume will have increased by approximately 20%.9,11,16,17 Unremarkably, intracranial volume increases with greater distraction distance, but the final distraction length is not the sole determinant of volume increase. As Nowinski and colleagues11 have pointed out, the position of the inferior osteotomy is also an important factor, although the benefit of increased volume from a lower osteotomy site has to be balanced with the increased risk for significant hemorrhage from venous sinuses within the posterior fossa. The patients did not routinely undergo three-dimensional computed tomographic imaging after PD in our series, to minimize their radiation exposure.20 As a result, the increase in intracranial volume has not been calculated. Plain lateral skull radiographs were used routinely to objectively assess the advancement gap during and after distraction. These were repeated in follow-up when necessary. The patients experienced an increase in skull length largely equivalent to their final advancement distance, with a corresponding subjective improvement in morphology. However, we believe that simple morphometric measurements do not adequately describe the complex calvarial deformity of most patients with syndromic craniosynostosis and hence are of little use in objectively assessing the outcomes of PD. The use of nonirradiating modalities to assess volumetric changes, such as three-dimensional laser scanning, hold promise but are currently expensive, are time consuming, and require operator technical expertise to be reliable. Furthermore, the maximum advancement achievable from PD, 20 mm in this series, is limited by device design and tissue factors. Increasing skull length by such a distance in children with severe turribrachycephaly is usually not sufficient to fully correct their deformity. Rather, the primary goals of PD are to redirect calvarial growth along the fronto-occipital plane, to provide sufficient intracranial space to defer FOAR surgery, and to prevent raised ICP from developing or to treat it, if present. In our series, secondary FOAR was deferred to a mean of 2 years after surgery. All patients with proven raised ICP preoperatively had no clinical symptoms or signs of intracranial hypertension after distraction. However, PD does not prevent a rise in ICP in perpetuity as evidenced by the 2 patients who underwent further surgery for new-onset raised ICP more than 2 years after PD. In addition, 1 further patient with Apert syndrome, severe brachycephaly, and a Chiari malformation (patient 4) was found to have raised ICP within 6 months of PD, although his situation was complicated by earlier hydrocephalus necessitating a third ventriculostomy. In this study, 4 patients with raised ICP had minor dural injures that were identified and repaired during device placement. Distraction in the presence of raised ICP and a dural tear potentially risk causing a persistent CSF leak because CSF pressure with distraction-induced tension across the wound might prevent dural healing. Interestingly, the 1 patient who developed a persistent CSF leak was the only one to be distracted on the table (by 5 mm). He
Lessons From Posterior Distraction
and 2 of the remaining 3 patients began distraction per protocol the following day. The final patient waited 4 days before distraction was started, on account of her dural injury. With hindsight, it is our opinion that patients who have intracranial hypertension and a dural injury should not be distracted on the table. They would also benefit from a longer latency period before commencing distraction. We encountered 1 delayed CSF leak associated with a significant dural tear that required distraction to be abandoned, in a patient with raised ICP who was not identified to have a dural injury during device placement (patient 18). She had also been distracted by 5 mm on the table. Cerebrospinal fluid leaks not attributable to recognized dural injury have occurred with a similar incidence in other series.13,16 Goldstein and colleagues16 hypothesized that the 2 leaks in their series were caused by “scratches” from screw tips. As a result, they changed to self-tapping screws and began to place an absorbable gelatine sponge beneath the screws to protect the dura. We have always routinely placed Spongostan (Ferrosan A/S, Soeborg, Denmark) beneath the bone at the distractor sites because of similar concerns.12 Recent technological advances that eliminate the need for screw fixation of devices, such as ultrasound-activated pin osteosynthesis (SonicWeld Rx, KLS Martin Group, Tuttlingen, Germany), could prevent this complication entirely.21 The latency period of 24 to 48 hours used in this series is briefer than the 72-hour to 5-day period used by others.13,14,16,19,22 Three- to 5-day latency periods are routinely used elsewhere in the skeleton and are derived from Ilizarov and Ledyaev's23 original work and that of others more recently (summarized by McCarthy et al,24 2001). The length of the latency period is an attempt to balance the advantages of a greater callus mass, gained through a longer latency period, with the increasing risk for premature consolidation that delaying distraction entails. The well-documented dural capacity for osteoinduction in infants and younger children strongly suggests that distraction of callus is of secondary importance in calvarial osteogenesis in PD.25–27 We observed consistently reliable ossification of the distraction defect despite the reduced latency period (as exampled by Fig. 1). We can identify little benefit in using a latency period designed for optimal callus formation in long bones in the distraction of the calvarium. We do support an extended latency period in circumstances of dural injury, discussed previously, or where PD is the second transcranial procedure and the wound closure is tight, with a consequent risk for wound dehiscence with early distraction. In this series, 2 surgical approaches were used, the hinge and push-back techniques. As we have discussed elsewhere, the hinge technique facilitates the rotation of the posterior panel backward and downward, dropping the height of the vertex.12 It is our preferred approach to correcting turribrachycephaly. We favor the push-back technique in cases in which expansion of the posterior cranial fossa is an important indication for surgery, as in the presence of a Chiari malformation, or in primarily brachycephalic deformities in which cranial height is already normal. Effective PD undoubtedly has the capacity to improve established Chiari malformations, as we and others have found.14,28 Even in the 3 patients in whom the malformation did not improve, in this series, there has been no radiologic deterioration during follow-up (range, 2–4.5 y), although 1 child with a complex condition later had an FMD for suspected brain stem compression. The number of distractor devices used in PD varies between series, ranging from 4 to a single device per patient.12–14,16,17 The use of more than 2 distractors is probably unnecessary because it raises the cost and the complexity of the procedure and increases the risk for developing conflicting distraction vectors and other complications.11 Initially, we favored using only 1 distractor in the hinge technique, to minimize the aforementioned risks. However, since 2010, we have routinely used 2 devices in both procedures. Although we have shown symmetric distraction with a single
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1725
Thomas et al
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
FIGURE 1. A, Preoperative three-dimensional computed tomography of patient 19, who had proven raised ICP. B, Day 18 after insertion of distractors, 18 mm of distraction. The distractors were removed immediately at the end of distraction, day 20. C, Sixty days after device removal. D, One hundred fifty-eight days after device removal. The distraction osseous defect continues to ossify without relapse.
device in the hinge technique, 2 devices give finer control over the distraction, helping to ensure symmetry. Further, 2 devices allow for redundancy. In the 1 case of distractor failure and 3 instances in which 1 distractor was removed during the distraction phase, the remaining device enabled distraction to be continued to completion. Wound infections occurred in 29% of the patients in this series, with 16% returning to the operating theater earlier than anticipated. Although this rate was higher than that reported by other authors in PD, it is similar to that observed in distraction procedures in other areas of the craniofacial skeleton.15,29 Importantly, wound infections did not prevent full distraction being achieved or lead to serious sequelae, such as osteomyelitis. Undoubtedly, the increased wound infection rate we observed was due to the presence of the distractor devices because we have not seen similar rates in other posterior cranial procedures, despite an identical antibiotic prophylaxis regimen. Although the PD series currently in the literature do not discuss antibiotic prophylaxis, the regimen used in this study is similar to that reported in other transcranial craniofacial operation series30 and exceeds in duration the 48 hours of prophylaxis that Ahmad and colleagues have used in frontofacial monobloc distraction advancement procedures.31 Tissue hypoxia and frank necrosis of the scalp caused by distractor devices are likely to be important factors in the pathogenesis of infection in PD. We observed tissue necrosis to occur both beneath the activation arms and also over the hub covering the terminal end of the central rod, buried beneath the scalp. Because of the spherical nature of the cranium and the linear vector of distraction, the scalp becomes increasingly tented over the high points of the distractor, principally the hub and the central rod (Fig. 2). Likewise, scalp tissue beneath the activation arm is increasingly compressed, which occurred in this study despite the use of flexible activation arms. This highlights the likely importance of both device design and placement. It may also explain the differences in infection rates between our series and others because only 1 other group published in the literature has used the same device.11,32
1726
We believe that low-profile devices that are adapted to the convex contour of the posterior cranium will significantly reduce both wound infection and device-induced tissue necrosis rates. Modification of devices to use soft flexible components, wherever practicable, is also likely to be of benefit. We plan to use such a device, currently being designed, in the future. Early distractor removal with a reduced or absent consolidation period was not associated with relapse of the deformity in this series (Fig. 1). This indicates that new bone formation is not the principle mechanism by which the distraction gap is maintained in PD. Instead, distraction allows the previously constricted intracranial contents to expand behind the distracted posterior panel and splint it in its new position.12,32 Indeed, the only significant relapse we encountered occurred in a patient who had a consolidation period of 102 days. The prolonged presence of the devices may even have contributed to the relapse because the patient experienced scalp necrosis beneath the activation arms of both distractors and was therefore nursed supine, to protect these areas. Furthermore, 6 of the 7 episodes of scalp necrosis we observed occurred during the consolidation phase. In these cases, early distractor removal may well have successfully salvaged the compromised tissue, as it did in patient 19. We therefore advocate the early removal of devices, at 1 month after the termination of distraction, because we can find no advantage in the traditional consolidation period of 2 to 3 months, whereas the disadvantages of a prolonged consolidation period in terms of the risk for tissue necrosis, wound infection, and parental anxiety are quite apparent. A clear advantage of PD over posterior advancement techniques is that it permits closure of the scalp without tension.13,14,22 During distraction, the overlying soft tissues are expanded to ultimately accommodate a greater cranial volume than they could have done in a single-stage procedure. Although the scalp expands sufficiently to encompass the distracted calvarium, it remains relatively tight after distraction. In more than half of the patients who underwent FOAR after PD, the scalp was found to be unusually tight, despite a mean interval of more than 2 years between PD and FOAR. The overall reoperation rate in this group was 18%, entirely due to wound dehiscence. We now anticipate this problem and have modified our wound closure for FOAR after PD. The posterior galeal flap is raised as far backward as the origin of occipitalis, and the bicoronal incision is then closed with continuous nonabsorbable
FIGURE 2. Tenting of the scalp over the central rod of a fully extended distractor (black arrows), with exposure of the prominent distractor hub (patient 2).
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
The Journal of Craniofacial Surgery • Volume 25, Number 5, September 2014
monofilament suture, rather than the absorbable sutures we routinely use for scalp closure in other circumstances. An appealing alternative approach to avoiding a tight coronal wound at FOAR is to advance the fronto-orbital construct via distraction only once the incision has been closed.33 We plan to pursue this approach in some of the patients presented here in the future. In conclusion, in this large consecutive series of 31 patients treated by PD, unanticipated events occurred in 61.3% of the patients, with 32.2% experiencing an earlier return to the operating theater than planned and 19.4% undergoing 3 or more general anesthetic procedures. Wound infection (29.0%) and tissue necrosis caused by the distractor (22.6%) were the commonest complications, but all patients who endured these were fully distracted and had no long-term sequelae. Cerebrospinal fluid leaks were rarer (6.5%) after distractor placement but had a more significant impact with early cessation of distraction in 1 case and abandonment of PD entirely in the other. Distraction was stopped early in only 1 other patient who had a threatened wound dehiscence. In all but 1 patient who completed distraction, there was an improvement in cranial morphology and resolution of clinical indicators of raised ICP, where present. In the patients with a Chiari malformation, 40% improved after surgery. Twenty-nine percent of the patients had distractors removed within 30 days of completing full distraction. None of these patients experienced a relapse. Our 5-year experience of PD has led us to make important changes to our distraction protocol. We advocate short latency periods of 48 hours or less, in the absence of a tight wound or recognized dural tear associated with raised ICP. The use of 2 devices allows for redundancy; thus, where 1 device needs to be removed early, full distraction can still be achieved. An abbreviated consolidation period does not result in relapse and, if anything, is likely to reduce complication rates. Posterior distraction is an important advance in the treatment of patients with severe brachycephaly or turribrachycephaly, but it is not without risks and forms only part of an integrated management approach to this sometimes challenging group of patients.
REFERENCES 1. Collmann H, Sorensen N, Krauss J. Hydrocephalus in craniosynostosis: a review. Childs Nerv Syst 2005;21:902–912 2. Marucci DD, Dunaway DJ, Jones BM, et al. Raised intracranial pressure in Apert syndrome. Plast Reconstr Surg 2008;122:1162–1168 3. Thompson DN, Harkness W, Jones BM, et al. Aetiology of herniation of the hindbrain in craniosynostosis. An investigation incorporating intracranial pressure monitoring and magnetic resonance imaging. Pediatr Neurosurg 1997;26:288–295 4. Cinalli G, Spennato P, Sainte-Rose C, et al. Chiari malformation in craniosynostosis. Childs Nerv Syst 2005;21:889–901 5. Marchac D, Renier D, Broumand S. Timing of treatment for craniosynostosis and facio-craniosynostosis: a 20-year experience. Br J Plast Surg 1994;47:211–222 6. Wall SA, Goldin JH, Hockley AD, et al. Fronto-orbital re-operation in craniosynostosis. Br J Plast Surg 1994;47:180–184 7. Wong GB, Kakulis EG, Mulliken JB. Analysis of fronto-orbital advancement for Apert, Crouzon, Pfeiffer, and Saethre-Chotzen syndromes. Plast Reconstr Surg 2000;105:2314–2323 8. Woods RH, Ul-Haq E, Wilkie AO, et al. Reoperation for intracranial hypertension in TWIST1-confirmed Saethre-Chotzen syndrome: a 15-year review. Plast Reconstr Surg 2009;123:1801–1810 9. Choi M, Flores RL, Havlik RJ. Volumetric analysis of anterior versus posterior cranial vault expansion in patients with syndromic craniosynostosis. J Craniofac Surg 2012;23:455–458 10. Sgouros S, Goldin JH, Hockley AD, et al. Posterior skull surgery in craniosynostosis. Childs Nerv Syst 1996;12:727–733
Lessons From Posterior Distraction
11. Nowinski D, Di Rocco F, Renier D, et al. Posterior cranial vault expansion in the treatment of craniosynostosis. Comparison of current techniques. Childs Nerv Syst 2012;28:1537–1544 12. Wiberg A, Magdum S, Richards PG, et al. Posterior calvarial distraction in craniosynostosis—an evolving technique. J Craniomaxillofac Surg 2012;40:799–806 13. White N, Evans M, Dover MS, et al. Posterior calvarial vault expansion using distraction osteogenesis. Childs Nerv Syst 2009;25:231–236 14. Steinbacher DM, Skirpan J, Puchala J, et al. Expansion of the posterior cranial vault using distraction osteogenesis. Plast Reconstr Surg 2011;127:792–801 15. Yonehara Y, Hirabayashi S, Sugawara Y, et al. Complications associated with gradual cranial vault distraction osteogenesis for the treatment of craniofacial synostosis. J Craniofac Surg 2003;14:526–528 16. Goldstein JA, Paliga JT, Wink JD, et al. A craniometric analysis of posterior cranial vault distraction osteogenesis. Plast Reconstr Surg 2013;131:1367–1375 17. Serlo WS, Ylikontiola LP, Lahdesluoma N, et al. Posterior cranial vault distraction osteogenesis in craniosynostosis: estimated increases in intracranial volume. Childs Nerv Syst 2011;27:627–633 18. Taylor JA, Derderian CA, Bartlett SP, et al. Perioperative morbidity in posterior cranial vault expansion: distraction osteogenesis versus conventional osteotomy. Plast Reconstr Surg 2012;129:674e–680e 19. Ylikontiola LP, Sandor GK, Salokorpi N, et al. Experience with craniosynostosis treatment using posterior cranial vault distraction osteogenesis. Ann Maxillofac Surg 2012;2:4–7 20. Mathews JD, Forsythe AV, Brady Z, et al. Cancer risk in 680 000 people exposed to computed tomography scans in childhood or adolescence: data linkage study of 11 million Australians. BMJ 2013;346:f2360 21. Arnaud E, Renier D. Pediatric craniofacial osteosynthesis and distraction using an ultrasonic-assisted pinned resorbable system: a prospective report with a minimum 30 months’ follow-up. J Craniofac Surg 2009;20:2081–2086 22. Derderian CA, Bastidas N, Bartlett SP. Posterior cranial vault expansion using distraction osteogenesis. Childs Nerv Syst 2012;28:1551–1556 23. Ilizarov GA, Ledyaev VI. The replacement of long tubular bone defects by lengthening distraction osteotomy of one of the fragments. 1969. Clin Orthop Relat Res 1992;280:7–10 24. McCarthy JG, Stelnicki EJ, Mehrara BJ, et al. Distraction osteogenesis of the craniofacial skeleton. Plast Reconstr Surg 2001;107:1812–1827 25. Greenwald JA, Mehrara BJ, Spector JA, et al. Biomolecular mechanisms of calvarial bone induction: immature versus mature dura mater. Plast Reconstr Surg 2000;105:1382–1392 26. Greenwald JA, Mehrara BJ, Spector JA, et al. Immature versus mature dura mater: II. Differential expression of genes important to calvarial reossification. Plast Reconstr Surg 2000;106:630–638 27. Yang JY, Yang WG. Large scalp and skull defect in aplasia cutis congenita. Br J Plast Surg 2000;53:619–622 28. Ahmad F, Evans M, White N, et al. Amelioration of Chiari type 1 malformation and syringomyelia following posterior calvarial distraction in Crouzon’s syndrome—a case report. Childs Nerv Syst 2014;30:177–179 29. Dunaway DJ, Britto JA, Abela C, et al. Complications of frontofacial advancement. Childs Nerv Syst 2012;28:1571–1576 30. Fearon JA, Yu J, Bartlett SP, et al. Infections in craniofacial surgery: a combined report of 567 procedures from two centers. Plast Reconstr Surg 1997;100:862–868 31. Ahmad F, Cobb AR, Mills C, et al. Frontofacial monobloc distraction in the very young: a review of 12 consecutive cases. Plast Reconstr Surg 2012;129:488e–497e 32. Nowinski D, Saiepour D, Leikola J, et al. Posterior cranial vault expansion performed with rapid distraction and time-reduced consolidation in infants with syndromic craniosynostosis. Childs Nerv Syst 2011;27:1999–2003 33. Komuro Y, Shimizu A, Ueda A, et al. Whole cranial vault expansion by continual occipital and fronto-orbital distraction in syndromic craniosynostosis. J Craniofac Surg 2011;22:269–272
© 2014 Mutaz B. Habal, MD
Copyright © 2014 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.
1727
