Surgery for Nonsyndromic Single-Suture Craniosynostosis

The types of craniosynostosis may be described in either of the following ways:

• According to the clinical deformity itself often described by its shape (eg, trigonocephaly)
• According to the sutures affected (eg, metopic synostosis)

The former descriptive approach was developed by Virchow, who was influenced by then-prevailing anthropologic concepts; the latter was developed by Ingraham and Matson and was based mainly on radiologic evidence of fused sutures. There is considerable correspondence between the two nomenclatures, which is better appreciated in single-suture forms of craniosynostosis.

The rules for growth with regard to the fused sutures are indicated by the arrows in the illustrations below. Closure of a single suture not only causes restriction of growth perpendicular to the fused suture but also causes compensatory growth at adjacent sutures.[3] If the adjacent suture is parallel to the fused suture, the compensatory growth occurs equally in both directions. If the suture is perpendicular to the fused suture, the compensatory growth occurs away from the fused suture.

The following are four common types of craniosynostosis (in order of frequency):

• Scaphocephaly corresponds to sagittal synostosis (see the first image below) and is derived from the Greek words skaphos ("boat") and kephale ("head")
• Plagiocephaly corresponds to unilateral coronal synostosis (see the second image below) and is derived from the Greek words plagios ("oblique or sloping") and kephale
• Trigonocephaly corresponds to metopic synostosis (see the third image below) and is derived from the Greek words  trigonos ("triangular") and  kephale
• Posterior plagiocephaly corresponds to lambdoid synostosis (see the fourth image below)

The etiology of craniosynostosis has not been fully elucidated, but genetic defects are increasingly being recognized. No inheritance pattern has been identified for nonsyndromic forms of craniosynostosis, though a familial occurrence has been observed in the vast minority of patients. First-degree relatives are affected most commonly in cases of metopic craniosynostosis, followed by sagittal, lambdoid, and, rarely, coronal craniosynostosis.[4]  In familial cases, variable vertical and horizontal penetrance has been observed.

In 2010, a retrospective genetic study of craniosynostosis showed mutations as the cause in 37.5% of bilateral coronal, 17.5% of unilateral coronal, and 11% of multiple-suture craniosynostosis. The mutations were absent in all patients with nonsyndromic metopic, sagittal, and lambdoid craniosynostosis.[5]

The overall prevalence of craniosynostosis has been estimated to be 1 case in 2000 live births. Of the SSCs, the most common, in order of decreasing incidence, are sagittal, unilateral coronal, metopic, and lambdoid synostosis. 

Surgery improves the cranial deformity in most patients with SSC. However, it is not uncommon to perceive some residual skull-shape abnormality. In a minority of patients, the deformity recurs after a few years and necessitates reoperation. The majority of patients with SSC are treated primarily from a progressive deformity standpoint and do not develop elevated ICP or other neurological sequelae.

Some studies of children who have SSC have suggested that although there does not appear to be an increased incidence of intellectual disability, craniosynostosis may be associated with an increased incidence of subtle learning disabilities.[6, 7]  As many as half of all SSC patients may be found to have neurodevelopmental issues.[5, 8]  Nevertheless, the cause of these subtle learning disabilities has not been fully elucidated.

Although it has sometimes been assumed that increased ICP plays a role, only a minority of infants with SSC are found to have increased ICP; yet a higher percentage have learning disabilities. Altered brain morphology has also been proposed as an explanation, but studies have not shown worsened cognitive outcomes in children with more severe deformities.[9]  Advanced imaging (diffusion tensor imaging [DTI]) has also suggested persistent altered brain connectivity in children with treated sagittal craniosynostosis.[10]

Junn et al retrospectively reviewed long-term neurocognitive outcomes in 204 pediatric patients (mean age at surgery, 9.0 ± 12.2 mo; mean age at testing,10.9 ± 4.0 y) who underwent surgical correction of nonsyndromic SSC (139 sagittal, 39 metopic, 22 unicoronal, 4 lambdoid).[11]  Patients with metopic synostosis had lower scores in verbal IQ, full-scale IQ, visuomotor integration, visual perception, and motor control after surgical correction than patients with sagittal synostosis did. Patients with unicoronal had lower visuomotor integration and visual perception scores.

Lynn et al assessed neurodevelopmental outcomes in 66 patients with nonsyndromic SSC who were treated by means of cranial vault remodeling and were tested pre- and postoperatively with the Bayley Scales of Infant and Toddler Development (Third Edition) (BSID-III).[12] There were no significant changes between preoperative and postoperative neurodevelopmental functioning. The authors suggested that language and motor development are modestly delayed in such patients and that these delays are present before and after cranial vault remodeling. They did not find cranial vault remodeling to have a significant impact on the neurodevelopmental trajectory.

Regardless of the timing or type of surgery performed, there is an increased incidence of mild neuropsychological deficits.

The abnormal head shape is often noticeable at birth and leads to a referral to a specialist in cranial deformity. Typically, a pediatric neurosurgeon or a craniofacial plastic surgeon is able to determine the diagnosis on physical examination before a confirmatory imaging study is performed.

Many infants referred by pediatricians to neurosurgeons and craniofacial plastic surgeons for abnormalities in head shapes do not have craniosynostosis and instead have positional (or deformational) plagiocephaly. Positional plagiocephaly most often masquerades as either lambdoid or unilateral coronal craniosynostosis.

The most common type of positional plagiocephaly is characterized by unilateral flattening in the occipital region (mimicking lambdoid craniosynostosis). The severity of the flattening can vary but often leads to a parallelogram shape of the skull as viewed from above, with anterior displacement of the ipsilateral ear, forehead, and cheek. This forehead asymmetry mimics unilateral coronal craniosynostosis.

Careful clinical evaluation usually suffices to distinguish craniosynostosis from positional plagiocephaly. Lambdoid craniosynostosis causes a parallelogram-shaped skull as viewed from behind, in which the ear ipsilateral to the flattening is inferiorly deflected and the contralateral parietal bone experiences compensatory superior growth. Lambdoid synostosis consistently presents with a mastoid bulge or bossing on the side of the fusion which is not present in children with deformational plagiocephaly.

Although both positional plagiocephaly and unilateral coronal craniosynostosis are associated with forehead asymmetry, the nature of the asymmetry is quite different. Whereas in positional plagiocephaly the forehead and orbital rim contralateral to the flattened occiput are normal, the forehead and orbital rim in unilateral coronal craniosynostosis are recessed and the contralateral frontal bone is overgrown.  

Patients undergoing surgical treatment have their coagulation profiles, hemoglobin, and electrolyte levels checked to ensure that these levels are within the reference range so that they can be submitted to general anesthesia.

Cranial remodeling operations are often associated with significant blood loss, and the usual practice is to secure suitable crossmatched blood before surgery.

Plain radiography initially, computed tomography (CT) subsequently, and magnetic resonance imaging (MRI) more recently have been used in clinical practice for both preoperative planning and postoperative follow-up. These modalities are also used in research and for exploring the pathophysiologic mechanisms implicated in causing the skull deformity.

Three-dimensional (3D) CT (3D-CT) scans, though typically not necessary to diagnose craniosynostosis, are capable of providing useful anatomic information. These clearly demonstrate the abnormally fused suture(s) and allow accurate preoperative planning. Of surgical importance is the configuration of the bony skeleton of the orbits, the anterior fossa floor, and the temporal regions. These are the sites at which osteotomies are commonly performed during corrective surgery.

Using reconstructed 3D-CT scans (see the images below), surgeons can readily appreciate the morphology of the sites of interest and can formulate a suitable operative plan—for instance, to determine how far the supraorbital bar must be advanced to correct the orbits. CT venography can be helpful in planning occipital craniotomies that will expose portions of the torcula and transverse sinuses.

No medical treatment exists for craniosynostosis, and helmet therapy alone does not correct the head shape. Because of the progressive nature of the cranial deformity, most children with craniosynostosis are recommended for surgery. However, children with mild deformities or those who present late without signs of increased intracranial pressure (ICP) are occasionally treated without surgery.

The literature is mixed with respect to the ability of surgery to affect cognitive outcomes.[13]  Nevertheless, there has been some evidence to suggest that earlier more comprehensive cranial vault reconstructions are associated with improved cognitive outcomes as compared with less comprehensive procedures or comprehensive procedures performed later.[14, 15]  Even with early comprehensive surgery, studies suggest that subtle learning disabilities persist.[16]

Surgical options

Although there are several options for the treatment of single-suture craniosynostosis (SSC), each is designed to return the skull shape toward a more normal contour and allow for unrestricted cranial expansion. The two basic types of techniques are as follows:

• Cranial vault reconstruction (CVR)
• Suturectomy (strip craniectomy), which is most often followed by the use of a molding helmet

CVR has the advantage of immediately correcting the cranial shape but the disadvantage of being a more extensive operation associated with larger incisions, higher rates of transfusion, and longer hospital stays.

A national longitudinal comparison of strip craniectomy (n = 251) with CVR (n = 1811) for craniosynostosis management identified socioeconomic disparities between strip craniectomy patients and CVR patients.[17] CVR was found to be more commonly performed in underrepresented minorities and patients with Medicaid, whereas strip craniectomy was common in the White population and patients with private insurance. Although CVR was associated with higher hospital charges and complication rates than strip craniectomy, the differences were less than expected.

Alternatives to these procedures have also been described, including spring-mediated[18] and external distraction devices.[19] Spring-assisted surgery has gained popularity. It originated in 1997 with Lauritzen in Sweden.[20]  Lauritzen’s group in Gothenburg has preferentially used spring cranioplasty in children up to age 6 months and the modified pi-plasty in older children.[21]  Initial use of this approach focused on SSC, particularly sagittal synostosis.

Valetopoulou et al performed a systematic review (22 studies; N = 1094) comparing endoscopic strip craniectomy followed by postoperative molding helmet therapy (ESC-H; n = 605) with spring-assisted cranioplasty (SAC; n = 489) for correction of nonsyndromic sagittal SSC.[22]  They found outcomes to be broadly similar and concluded that the available evidence was insufficient to establish either ESC-H or SAC as superior in this setting. Much of the literature consisted of single-center retrospective studies of low methodologic quality; international multicenter randomized controlled trials comparing ESC-H with SAC would be required in order to yield definitive and generalizable data.

Distraction osteogenesis for unicoronal craniosynostosis has been described.[23]

Timing of surgical intervention

Although there is not a standard of care regarding the exact timing of surgery for craniosynostosis, surgery is generally recommended during the first year of life. The age at which surgery is performed is somewhat technique-dependent. Endoscopic-assisted suturectomies followed by molding helmet therapy are generally performed prior to 4 months,[24] before the compensatory deformities become more of an issue and while the skull is more malleable. Suturectomy relies on early brain growth to passively change the shape of the skull. More comprehensive CVRs are generally performed in the second half of the first year of life and actively change the shape of the skull.

Although earlier surgery may be associated with improved cognitive outcomes, the risk of surgical morbidity and mortality must also be considered. In addition, there is evidence to suggest that earlier surgery is associated with a higher incidence of recurrence and a more frequent requirement for repeat procedures.[25]  The true impact of timing of surgery on neurodevelopmental outcomes has not been defined with certainty.[26]  

Preparation for surgery

CVR procedures to correct craniosynostosis are known to result in extensive blood loss. Bleeding is the main cause of mortality after surgery to correct craniosynostosis. Meticulous hemostasis and early transfusion can mitigate the results of blood loss.[27, 28]

Tranexamic acid (TXA) has been described in the literature as an adjuvant for reducing blood loss and transfusion requirements. A double-blind, placebo-controlled trial was performed with TXA during correction of craniosynostosis.[29] Patients were loaded with 50 mg/kg of TXA after induction of anesthesia, before incision, which was followed by infusion of 5 mg/kg/hr during surgery. These patients were compared with those receiving placebo and found to have lower perioperative mean blood loss (65 mL/kg vs 119 mL/kg) and lower perioperative mean blood transfusion (33 mL/kg vs 56 mL/kg).

Another group, in a randomized double-blind study, pretreated patients with erythropoietin (600 U/kg) once a week for 3 weeks leading up to surgery.[30] The volume of packed erythrocytes transfused was reduced by 85% intraoperatively and by 57% throughout the study period. Other studies in children undergoing cardiac surgery or spinal surgery for scoliosis have found similar benefits.

The mechanism by which TXA works is not well understood; moreover, the studies have been criticized for the heterogenous populations and attenuated power of the studies.[31] Research has continued in this area, in particular with regard to TXA dosing. Overall, many centers have adopted the use of TXA to lower transfusion rates given the current data available. 

Controlled hypotension to reduce blood loss during fronto-orbital advancement was studied by Seruya et al.[32] Mean arterial pressure (MAP) and calculated blood loss were evaluated. An inverse relation between MAP and calculated blood loss was discovered; however, on further evaluation, it was found that blood loss was the cause of changes in MAP.

In sagittal synostosis, the skull is long and narrow. Correction requires reconstruction of the skull so that it is rendered wider as compared with its anterior-posterior length. One factor that must be taken into account during preoperative planning and repair is compensatory growth, which can be anterior, posterior, or both. 

Surgical goals are to shorten the skull and to widen it with correction of severe compensatory growth. Additionally, a bifrontal craniotomy is required to correct the frontal bossing. In similar fashion, if the compensatory growth involves the occipital bone, an occipital craniotomy may be beneficial. If compensation involves both frontal and occipital bones, then surgery often must be performed with the patient in a modified prone position and must include both a bifrontal and an occipital craniotomy. However, some surgeons do not advocate frontal craniotomy, on the grounds that this compensation has been shown often to improve over time with surgical correction during infancy.[33]

Permanent marker is used to delineate the planned incision line with the patient in the prone position for this case example. The incision line is planned in a curved manner to allow a more cosmetically pleasing result as the child ages. (See the image below.) The focus of this correction will be on the parietal bones and occipital narrowing.

The bicoronal incision is made, and Raney clips are placed on both sides of the wound to maintain meticulous hemostasis. One of the more significant risks to an infant undergoing CVR is blood loss. The surgical team and the anesthesiology team must work together to reduce risks and treat blood loss quickly so as to avoid complications. The image below depicts full exposure of the skull with scalp flaps retracted in prone position. The margins of the planned cuts to remove the sutures with the craniotome are also drawn with a marker. The central line marks the fused sagittal suture, with lines on each side of it marking a central segment for removal.

The anterior fontanelle must be teased away from the surrounding skull before the craniotome can be used to create the cuts that will remove the sutures of interest. Care must be taken not to disturb the superior sagittal sinus. The cuts to remove the suture are made with the craniotome. The central segment including the sagittal suture is then easily removed and placed aside to use as part of the final construct if needed. Bilateral closed wedge-shaped osteotomies are performed near the patent lambdoid sutures to allow shortening of the anterior-posterior length.

In the final configuration (see the image below), the parietal bones have been widened at their lateral bases, and the occipital bone has been flattened and brought closer to the parietal bones to shorten anterior-posterior length. In young infants, the vertex segment of bone may be discarded or used as bone graft between the parietal bones. In older children, this segment is used for bone graft; filling cranial defects from surgery is less reliable in this population.

After closure (see the image below), the improvement in width and occipital flattening is apparent.

The unilateral coronal craniosynostosis produces a forehead that is typically bossed on one side and recessed on the other. In this case, a bifrontal craniotomy is required, with reconstruction of the frontal bone. In particular, the bossed area must be recessed and reduced. An orbital rim advancement is also required.

In patients with right unilateral coronal craniosynostosis, the bossed left forehead and the recessed right forehead are readily apparent (see the image below).

Upon reflection of the scalp, one can appreciate the recessed right forehead and a bossed left forehead with much more clarity (see the image below). The goal of correction is to bring the right forehead forward and the left forehead back slightly, in relation to each other, so that as the child grows, the forehead will be even when viewed from above. Some centers advocate “overcorrecting” the recessed/synostotic side to prevent recrudescence. 

The superior orbital rim will be advanced on the right as well, with the use of some autologous bone from the suturectomies. The frontal bones are removed with the craniotome. The superior orbital rim is removed with a combination of a high-speed saw and osteotomes. The periorbita is protected with a malleable retractor. Once the superior orbital rim is removed, it is remodeled.

In this case, the superior orbital rim is remodeled by extending the right side with autologous bone and absorbable plates. Once the superior orbital rim has been remodeled to correct the recessed right side and bossed left side, it is ready to be reattached with absorbable plates to the surrounding bones, including the frontal bones, which have already been reshaped. The corrected frontal bones and superior orbital rim are reattached with absorbable plates. The right frontal region is no longer recessed, and the left frontal region is no longer bossed.

With the closure complete (see the image below), the symmetry of the frontal bones is more obvious. The advancement of the right side and the partial recession of the left frontal region can be appreciated.

The bilateral coronal synostosis produces a skull that is excessively tall and short. The surgery to correct this should produce a skull that is longer in the anterior-posterior dimension and shorter in the superior-inferior dimension. As with the unilateral coronal synostosis, an orbital rim advancement is required. Decreasing the vertical height of the skull can be challenging but can be accomplished using a variety of techniques, each with their particular advantages and limitations.

In the example illustrated (see the images below), surgical treatment was performed in two stages. First, a biparieto-occipital osteotomy was performed, with external detractors designed to lengthen the skull but also to decrease the vertical height. After the bone consolidated, a standard fronto-orbital approach was performed to correct the recessed forehead and orbital rims.

Metopic synostosis is characterized by trigonocephaly. The forehead appears ridged, and the patient has hypotelorism and proptosis (see the images below).

This condition is repaired by advancing the orbital rims, which are noted to be recessed, in addition to removing the fused metopic suture. The forehead requires careful reconstruction. Some institutions perform an endoscopic strip suturectomy through a small incision and then helmet the child to reshape the head as the child grows.[34]  In this example, a fronto-orbital reconstruction is performed, much as for the unilateral coronal patient but with a focus on correction of the bilateral flattening. After forehead scalp reflection, the bony deformity is exposed (see the image below).

A bifrontal craniotomy is performed removing the forehead, and this is followed by  removal of the supraorbital bandeau (bar). This supraorbital bar is deformed, with bilateral narrowing and a triangular shape (see the image below). 

The flattening is contoured to a projected, more arched configuration, and it is stabilized to the parietal bones on each side with resorbable plates. The frontal bones are contoured to match the supraorbital bar and attached. The scalp is closed. (See the image below.)

Lambdoid synostosis is characterized by unilateral flattening of the ipsilateral parietal bone and, particularly, the ipsilateral occiput. Compensatory changes include overgrowth of the ipsilateral mastoid bone with inferior deflection of the ipsilateral ear, as well as overgrowth of the contralateral parietal bone. This creates a skull that is parallelogram-shaped when viewed from behind.

The goals of surgery are to create a rounder ipsilateral parieto-occipital region, to remove and elevate the ipsilateral overgrown mastoid, and to reduce the contralateral overgrown parietal bone. (See the image below.) Care should be taken to protect the major venous sinuses during this approach; preoperative computed tomography (CT) venography can be helpful. Multiple options are available to accomplish the reshaping and are less standardized than the options for other craniosynostoses.

After cranial vault remodeling procedures, patients are monitored in the pediatric intensive care unit (ICU). Serial hemoglobin and coagulation parameters are evaluated until stabilization is achieved. Patients receive transfusions as needed according to set cutoffs. Postoperative imaging may be performed to assess for early postoperative complications and to demonstrate the new arrangement of the bony architecture of the cranial vault. 

Complications are rare after craniofacial surgery. Hypovolemic shock can occur if significant intraoperative blood loss has not been replaced in a timely manner. Blood loss during surgery has been shown to increase with longer operating times, particularly when such times exceed 5 hours. Additionally, recognized craniofacial syndromes and pansynostosis have also been associated with increased blood loss during surgery.[35]

Intraoperative dural tears that remain unrecognized can cause postoperative cerebrospinal fluid (CSF) leaks and resultant infection or subgaleal fluid collections. Epidural or subdural hematoma can occur.

Almost all patients develop facial swelling postoperatively, more prominently around the eyes, which rarely causes problems; however, parents and caregivers should be counseled appropriately. Wound infections are generally rare. The frequency of these complications is less than 10%.

Restenosis, though rare, can occur. Long-term follow-up is warranted. Patients are also assessed at regular intervals to monitor for the ossification of cranial defects left during the reconstruction. Serial head circumference measurements are obtained to confirm proper growth of the skull. Although ophthalmologic evaluations do not effectively rule out intracranial hypertension, they can be a useful adjunct in monitoring for its occurrence.