Achondroplasia and Other Dwarfisms

Published on 26/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 1971 times

CHAPTER 219 Achondroplasia and Other Dwarfisms

Benjamin S. Carson, Sr., Daniele Rigamonti, Edward S. Ahn

Hypochondroplasia, achondroplasia, and thanatophoric dysplasia are clinically related skeletal dysplasias caused by gain-of-function mutations of the fibroblast growth factor receptor 3 (FGFR3) gene. The phenotype, disproportionately short stature with rhizomelic shortening of the extremities, results from defective formation of endochondral bone.

Hypochondroplasia yields the mildest phenotype, which varies within and between families and frequently lacks the neurological complications often seen in achondroplasia, such as hydrocephalus, cervicomedullary compression, and spinal stenosis. Life expectancy in this group may approach normal. The pathologic features of micromelic short stature—reduced or unchanged interpedicular distances in the lumbar spine, disproportionately long fibulas, squared and shortened pelvic ilia, and reduced subischial leg length—are significantly less in hypochondroplasia than in achondroplasia. Achondroplastic patients have increased age-specific mortality rates at all ages, with the highest increase occurring in childhood. Cervicomedullary compression probably accounts for the excess deaths in children; however, cardiovascular causes of death appear to be increased in adults. Individuals with thanatophoric dysplasia seldom survive to adulthood.

Given our present understanding of the short-limb dysplasias, neurosurgical approaches to achondroplastic, thanatophoric, and hypochondroplastic patients are similar. Hydrocephalus and cervicomedullary compression are typically pediatric concerns, whereas spinal stenosis has traditionally been treated in adults; however, increased sensitivity to signs of spinal stenosis and improved surgical technology have allowed earlier treatment of clinically significant spinal compression. The initial symptoms frequently do not have strictly neurosurgical resolutions; thus, a comprehensive treatment plan involving a multidisciplinary team of physicians that includes a neurosurgeon, neurologist, pulmonary specialist, sleep specialist, geneticist, anesthesiologist, neuroradiologist, orthopedic surgeon, and otolaryngologist is useful.¹^–³ Because these patients are at risk for brainstem compression, comprehensive testing is directed toward the detection of central and obstructive apnea and cervicomedullary compression, all of which contribute to the risk for sudden death.

Several areas of research are leading to new initiatives in the management of skeletal dysplasias. First is the application of advanced diagnostic tools and imaging techniques.⁴ Second is an emphasis on outcome studies, a movement based on the well-supported assumption that different operative strategies can yield marked long-term differences in the patient’s health.⁴^,⁵ Third, a better understanding of molecular biology has led to fresh ideas for targeted molecular treatment of achondroplasia, namely, those that inhibit the excess FGFR3 signals that are present in these individuals. A fourth area, the use of recombinant growth hormone to treat pediatric patients, is outside the scope of this chapter. Here, we describe our current approach to the evaluation and management of short-limb dysplasias and summarize the molecular biology, diagnostic findings, and outcome studies, mostly from our own institution.

Genetics and Epidemiology

Achondroplasia is an autosomal dominant disorder that results from a guanine (G)-to-adenine (A) mutation at base 1138 (G1138A) in the FGFR3 gene on chromosome 4 at 4p16.3.⁶^–⁸ The gene codes for a tyrosine kinase receptor expressed in developing bones. The nucleotide (missense) mutation results in an amino acid glycine (Gly)-to-arginine (Arg) substitution at amino acid 380 (Gly380Arg) in the transmembrane domain of the protein. Penetrance of the Gly380Arg mutation is 100%; in other words, all individuals with the mutation will have achondroplasia. A few patients with achondroplasia have been demonstrated to instead have a Gly375Cys mutation of the FGFR3 gene.⁹ Achondroplasia appears to be genetically homogeneous. No significant racial difference has been detected in North America, Spain, Korea, Japan, China, or Sweden.¹⁰^–¹⁵

The FGFR3 hypochondroplasia mutation pattern is more diverse than that for achondroplasia.^14,¹⁶^–¹⁸ In many patients, a C1620G or C1620A mutation results in an asparagine-to-lysine substitution at amino acid 540 (N450K) in the FGFR3 proximal tyrosine kinase domain. In a substantial number of patients who express a mild phenotype, the mutation has yet to be identified.¹⁶ The phenotypic diversity confounds estimates of incidence data. Accurate prenatal ultrasonographic diagnosis is rare.¹⁹ Before identification of the specific nucleotide mutations, some achondroplasia case reports probably included patients with hypochondroplasia. This overlap should be considered in efforts to compare current and past case series.

An arginine-to-cysteine (R248C) substitution in the extracellular domain of the receptor has been found in thanatophoric patients. Other mutations are likely because at least two phenotypes have been identified, thanatophoric dwarfism type I and type II. Less is known about this dysplasia because the mutation is almost always lethal neonatally.²⁰

Compound mutations for achondroplasia and hypochondroplasia have been reported in children whose achondroplastic father or mother had the G380R mutation and the other parent had the hypochondroplasia N450K mutation.²¹^,²² The phenotypic expression of the compound mutation appears to be more severe than that of achondroplasia.²²

New mutations account for about 80% of children born with achondroplasia. In other words, most infants with FGFR3 mutations are born to parents without FGFR3 mutations. As in many autosomal dominant conditions, a positive correlation exists between advanced paternal age and the occurrence of new mutations. It was initially thought that factors influencing DNA replication or repair during spermatogenesis, but not during oogenesis, may predispose to occurrence of the G1138A mutation in the FGFR3 gene.²³ However, more recent evidence suggests that sperm with the FGFR3 mutation have a selective advantage over sperm with a normal FGFR3 gene, thereby increasing the number of affected sperm with age.²⁴^–²⁶ Offspring of couples who are both affected by achondroplasia have a 25% chance of inheriting both parental achondroplasia alleles, thereby resulting in homozygous achondroplasia, which is almost universally fatal within the first year of life.²⁷ The skeletal features of achondroplasia are highly exaggerated in the homozygous condition: significantly shorter limbs, a smaller chest size, and a smaller foramen magnum. A brief summary of medical complications by age is presented in Table 219-1.

TABLE 219-1 Medical and Neurological Complications of Achondroplasia

Approximately 150 skeletal dysplasias have been identified, a number of which are associated with neurological symptoms.²⁸ Achondroplasia is the most common in clinical series. Although frequency estimates cluster between 1 in 10,000 and 1 in 35,000 live births, the true frequency must be recalculated because these data were generated before the FGFR3 mutations were identified.^6,^7,^29,³⁰

The FGFR3 gene encodes one of four tyrosine kinase receptors for fibroblast growth factor (FGFR1 to FGFR4) in mammals.³¹ Mutations of FGFR3 in achondroplasia have been shown to cause a gain of function, which correlates with the severity of the clinical phenotype.³²^,³³ FGFR3 normally functions as an inhibitor of linear bone growth by acting negatively on both the proliferation and differentiation of growth plate chondrocytes.³⁴^,³⁵ In achondroplasia, the normal function of FGFR3 is exaggerated. Numerous cellular mechanisms have been proposed that describe the increase in receptor tyrosine kinase activity that is common in all of the FGFR3 mutations in achondroplasia.³²

Medical Complications

Most individuals with achondroplasia have normal intelligence. A cross-sectional survey (using self-reported 36-item short form health survey [SF-36] data) of the functional health status of adults with achondroplasia revealed that the mental component summary scores did not differ significantly from scores in the general population. In contrast, the physical component summary scores were significantly lower starting in the fourth decade of life.³⁶ In children, motor milestones are delayed, partly because of generalized hypotonia and partly because of the mechanical disadvantage imposed by short limbs.^1,^27,³⁷^–³⁹ Psychosocial problems arising from short stature include lack of acceptance by peers and the tendency for adults, including parents and teachers, to treat children with achondroplasia appropriately for their height rather than their age.⁴⁰^,⁴¹ Quality-of-life issues on which the patient’s perceived status is suboptimal require special attention during adolescence.⁴² Involvement in support groups with other families who have children of short stature can improve self-esteem and assist parents in guiding their achondroplastic children through the difficulties of growing up in a culture that often equates stature with status.

Reproductive difficulties have not been conclusively documented, but reduced fertility, frequent fibroid cysts, and early menopause have been reported. The decreased reproductive rates in achondroplastic individuals may have been due in part to the social stigma present in those with reduced height in finding potential mates. However, with the establishment of organizations for those with reduced height, such as Little People of America, these individuals are now more likely to marry and have children.³² Women with achondroplasia must deliver their infants by cesarean section because of cephalopelvic disproportion,⁴³ and the administration of spinal anesthesia is strongly discouraged.⁴⁴

Obstructive sleep apnea, or sleep-disordered breathing secondary to a small upper airway, is common. Tonsillectomy and adenoidectomy decrease the degree of upper airway obstruction in most children. The majority do not have significant obstructive or central apnea, but a considerable minority are severely affected.⁴⁵^–⁴⁸ The cause of different patterns of sleep disorders may be related to localized alterations in chondrocranial development.⁴⁶ Many infants sleep with their necks in a hyperextended position, which functionally increases the size of the upper airway. Although the hyperextended neck position relieves intermittent obstruction, it can also exacerbate the neurological sequelae of cervicomedullary compression related to a small foramen magnum. Abnormal respiratory sinus arrhythmias may be present.⁴⁹ A small thoracic cage can result in restrictive pulmonary disease in infancy. Respiration may be further compromised by aspiration secondary to gastroesophageal reflux, swallowing dysfunction, or both, and result in recurrent pneumonia. Anesthesia can be given safely to children, with special consideration for limited neck extension and the use of appropriately sized endotracheal tubes.⁴⁷

A relatively high rate (about 3%) of jugular bulb dehiscence—complete absence of the roof over the jugular bulb—was identified in a series of 126 achondroplastic children. This increased incidence may account for unexplained hearing loss, tinnitus, and self-audible bruits in these children and poses a risk for difficult-to-control bleeding at myringotomy.⁵⁰

Evaluation and Diagnosis

Cervicomedullary Compression

Clinical Findings and Pathology

Cervicomedullary compression stems primarily from a reduction in the diameter of the foramen magnum in both the sagittal and coronal dimensions, a reduction that is sometimes more than 5 SD less than normal.⁵¹^–⁵⁶ Cervicomedullary compression warrants early and aggressive treatment because it results in high cervical myelopathy and increases the risk for sudden death by central respiratory failure.⁵⁷^–⁶⁰ A prospective evaluation of achondroplastic infants found radiographic evidence of craniocervical stenosis in 58% of the patients studied, and a diagnosis of cervicomedullary compression was made in 35%.⁶¹ Although these figures are derived from a selected population and are therefore higher than for the general population, they are a strong argument for careful evaluation and treatment of achondroplastic children. A retrospective study found increased mortality (in comparison to population standards) in achondroplastic children younger than 4 years, with sudden death from brainstem compression identified as the cause of half of the deaths. The same study also found a 7.5% risk for sudden death in the first year of life.²⁷

Chronic medullary and upper cervical cord compression may exist as a neurologically asymptomatic lesion and exhibit neither signs of root compression in the arms nor symptoms of cranial nerve impairment. Nonetheless, microcystic histopathologic changes, cervical syringomyelia, necrosis, and gliosis have been reported in autopsies of achondroplastic children who died unexpectedly. Presumably, lesions of this type interrupt the neural respiratory pathways from the nucleus tractus solitarii to the phrenic nerve nucleus, thereby arresting the muscles of inspiration and resulting in sudden death in some cases. We therefore consider infants with a history of sleep apnea or other severe respiratory or neurological abnormalities to be at increased risk for respiratory complications resulting from occult cervicomedullary compression. Some authors have recommended performing sleep and imaging studies on all children with achondroplasia.⁶² We believe that a careful history and neurological examination should precede costlier and more uncomfortable diagnostics. A composite profile of patients with cervicomedullary compression includes upper or lower extremity paresis, apnea or cyanosis, hyperreflexia or hypertonia, and delay in motor milestones beyond achondroplastic standards. These patients can present a striking contrast to the usual floppy, hypotonic achondroplastic infant.⁶³

Evaluation

Once a high-risk patient with respiratory or neurological symptoms or signs has been identified, we advise comprehensive testing. Parents should be carefully interviewed about the health and medical history of their child, with an emphasis on respiratory symptoms. A general physical examination should be performed, including chest measurements. Respiratory evaluation should include evaluation of blood pH, a chest radiograph, and overnight polysomnography. Electrocardiograms and echocardiograms should be performed to see whether there is cardiac evidence of chronic oxygen deprivation during sleep. A neurological examination for signs of brainstem compression, such as hyperreflexia, hypertonia, paresis, asymmetry of movement or strength, or abnormal plantar responses, is essential. Brainstem auditory evoked potential and upper extremity somatosensory evoked potential evaluation should be considered as an adjunct, especially in patients with normal results on neurological examination. Imaging studies of the craniocervical junction are necessary, particularly magnetic resonance imaging (MRI) in the sagittal plane. We also strongly recommend cardiac-gated MRI flow studies of cerebrospinal fluid (CSF) at the foramen magnum. There has been some evidence that dynamic MRI with the head in flexed and extended positions may increase the sensitivity for positional pathology in CSF dynamics.⁶⁴

Indications for Surgery

Concerns have been expressed about the indications for surgical decompression of the foramen magnum in this population.⁶⁵^–⁶⁷ Radiographic studies during the first years of life show some degree of compression at the foramen magnum level. MRI evidence of spinal cord compression, such as indentation or narrowing of the upper cervical cord, is a common finding that is usually graded as “marked” or “severe” in the MRI report.⁶⁵ In one report, myelomalacia was observed in 13 of 30 achondroplastic children.⁵⁵ Yet other than the routinely observed generalized hypotonia seen in the achondroplastic population, the majority of these children are asymptomatic and outgrow their developmental delays.^66,⁶⁸^–⁷⁰ Cervicomedullary decompression (CMD) as a standard routine prophylactic measure is therefore not warranted.

Several groups have published guidelines for surgical intervention.^68,^69,⁷¹ Our indications are based on lower limb symptoms, polysomnography, and MRI flow studies. The underlying principle must be to identify patients who are at risk for neurological damage or sudden death. We recommend that patients with cervicomedullary compression be identified and treated prophylactically, before abrupt and irreversible changes occur. For the purpose of diagnosis, we define clinically significant cervicomedullary compression to be (1) neurological evidence of upper cervical myelopathy or chronic brainstem compression (apnea, lower cranial nerve dysfunction, swallowing difficulties); (2) evidence of stenosis on imaging studies, including the absence of flow above and below the foramen magnum; and (3) frequently an otherwise unexplained respiratory or developmental abnormality.

Hunter and coworkers conducted a multicenter review of 193 patients with chondrodysplasias. The study reported data on rates of medical complications at specific age intervals (see Table 219-1). At age 4 the rate of cervicomedullary compression was 6.8%. The authors emphasized the important role of surgery, primarily because progressive neurological symptoms continue into adulthood. Ultimately, about 17% of the patients in the series underwent CMD.⁷²

Hydrocephalus

Clinical Findings and Pathology

Hydrocephalus in an achondroplastic patient is probably secondary to deformation of the cranial base. Constriction of the basal foramina, particularly the jugular foramina, is thought to reduce venous drainage and potentially raise intracranial venous pressure. Investigators have demonstrated a correlation between the degree of venous narrowing at the jugular foramina and the degree of hydrocephalus in achondroplaasia.⁷³ In theory, absorption of CSF into the sagittal venous sinus is thus reduced and results in hydrocephalus.⁵⁶ However, identifying patients at high risk for hydrocephalus is currently not possible. Hydrocephalus may resolve in some patients with continued growth of the skull base during puberty.

It is easy to suspect hydrocephalus in a patient with achondroplasia, given that macrocrania is a morphologic hallmark of the disease. Concerns about hydrocephalus may also arise because of the enlarged ventricles and the delayed acquisition of gross motor skills. Although hydrocephalus is associated with enlarged ventricles in the achondroplastic population, it generally resolves through growth and maturation of the cranial bones.⁷⁴ An achondroplastic child typically displays transient hypotonia, but the papilledema that is expected with symptomatic hydrocephalus is rare. Radiographically, mild to moderate ventricular enlargement, prominent cortical sulci, and an increased frontal subarachnoid space are apparent. Hydrocephalus severe enough to require shunting is often discovered after craniocervical decompression, when CSF leaks often complicate wound healing.

Evaluation

A degree of abnormal CSF dynamics is common in achondroplasia. We believe that routine surveillance of achondroplastic infants for hydrocephalus is best limited to noninvasive measures that are easily incorporated into the pediatric examination and should not require referral to special centers and complex diagnostic procedures. Longitudinally evaluating head circumference and tracking the acquisition of developmental milestones, with comparison to published standards specific for achondroplasia,⁷⁵^,⁷⁶ are usually sufficient. Imaging studies can be reserved for patients whose head circumference crosses percentiles on the achondroplastic chart or for those who exhibit unexplained neurological signs or developmental retardation.

Indications for Surgery

Stenosis of the jugular foramina contributes to the altered CSF dynamics in achondroplasia.⁵⁶ Jugular foramen decompression is an option.⁷⁷ However, we believe that this option should be used only in children with severe jugular stenosis and debilitating hydrocephalus in whom conventional ventriculoperitoneal shunting is contraindicated. Given the high percentage of complications, shunting is best reserved for those in whom the symptoms are severe and threatening.⁵⁴ CSF pressure profiles can be determined with an external transducer attached to an open fontanelle, with an epidural pressure monitor, or with an intraventricular catheter in older patients. If no critical pressure elevations are detected during a 48-hour period, shunt placement is not required, ventriculomegaly notwithstanding. However, the presence of severe clinical stigmata for hydrocephalus obviates such demonstrations.

For patients who have undergone craniocervical decompression, we expect sustained intracranial pressure (ICP) to be less than 20 mm Hg during the immediate postoperative period. Transient increases above this level can be associated with activity or irritation in normal individuals. In situations in which ICP is abnormally elevated, we proceed with shunting. In situations in which the interpretation is equivocal, we extend the period of monitoring for 1 or 2 days. Occasionally, even when no elevation in ICP is documented, a persistent CSF leak or subgaleal collections of CSF develop soon after the ventriculostomy is removed but are not necessarily indicative of hydrocephalus. We take a more aggressive approach to shunting, however, if subcutaneous collections develop over the site of craniocervical decompression.

Spinal Stenosis

Clinical Findings and Pathology

Spinal stenosis is the most common complication of achondroplasia and usually becomes symptomatic in the third decade or later. The anatomy of the achondroplastic spine is distinctive in several respects, all of which contribute to spinal cord compromise and nerve root compression. The superior and inferior articular facets of the vertebral bodies are susceptible to hypertrophy, which results in a “mushroom” shape that is clearly evident on MRI. Abbreviated and thickened pedicles of the vertebral arches result from premature fusion. Intervertebral disks tend to bulge prominently, thus further aggravating neural encroachment by the enlarged vertebral body articular surfaces. The interpediculate distance decreases in the lumbar region of the spine, which results in a canal that tapers caudally, the opposite of normal (the canal normally widens caudally). The overall picture is one of dramatic stenosis in every dimension of the spine (Fig. 219-1).

FIGURE 219-1 Illustration of the thoracolumbar spine in a pediatric patient with achondroplasia that demonstrates the abnormal bone anatomy leading to early neural compression.

(Used by permission from Sciubba DM, Noggle JC, Marupudi NI, et al. Spinal stenosis surgery in pediatric patients with achondroplasia. J Neurosurg. 2007;106:372; illustration by Ian Suk.)

Although the problems related to hydrocephalus and cervicomedullary compression are frequently identified in infancy and childhood, neurological problems below the foramen magnum are traditionally thought to be manifested in late adolescence and adulthood. However, in our most recent surgical series of 44 pediatric achondroplastic patients who underwent spinal decompressive surgery, over half were younger than 12 years.⁷⁸ Because the achondroplastic spinal canal tends to have severe congenital constriction, more intensive early screening might reveal substantial numbers of young achondroplastic patients with occult symptoms of spinal stenosis. Estimates of the incidence of symptomatic spinal stenosis range from 37% to 89%, thus suggesting that a significant proportion will eventually have this problem.⁷² Thorough urologic and neurological testing plays a useful part in the prospective evaluation for occult stenosis. Interestingly, in our pediatric series of achondroplastic spinal stenosis, 27 of 44 (61%) patients who underwent laminectomies for decompression had previously exhibited signs of cervicomedullary compression. In these patients the average age at the time of CMD was 3.5 years, whereas the average age at the time of spinal decompressive surgery was 11.5 years.⁷⁸

In management of the achondroplastic spine, it is possible to distinguish between the neurosurgical and orthopedic aspects of this disease. The hypotonia that an achondroplastic infant typically exhibits suggests that muscular tone may be insufficient for adequate protection of pediatric skeletal structures in weight-bearing postures. Achondroplastic children are in fact developmentally delayed in supporting their heads independently, as well as in upright sitting and walking. In our opinion, parents should not encourage early sitting because of the potential for aggravation of thoracolumbar kyphosis in this posture. Sitting and standing postures affect the curvature of the spine, and in achondroplastic children, muscular weakness, short vertebral pedicles, and lax spinal ligaments complicate these mechanics. Attention has also been drawn to the dynamic effect of a small chest and a globulus abdomen in the progressive development of kyphosis.⁷⁹ Moreover, delayed standing predisposes to the development of a gibbus, with wedging of one or more vertebral elements. These wedged deformities are both debilitating and preventable. Because surgical repair has risks, effort is well spent on prevention. Orthopedic bracing is used prophylactically when the formation of a wedged gibbus seems likely. Parents should also be urged to not use any infant carriers, strollers, or baby seats that exaggerate the thoracolumbar kyphosis.

In an adult, compromise can result from abnormalities such as hyperlordosis, minor disk bulging, hypertrophic osteoarthritis, or ligamentous hypertrophy. The presence of thoracolumbar kyphosis is also positively correlated with symptomatic spinal stenosis.⁷⁹ Although low back pain is a common complaint in achondroplastic patients, symptomatic neurogenic claudication can develop in those with severe stenosis. Prolonged walking produces first paresthesias and later weakness of the lower extremities, which is usually bilateral. These symptoms are promptly relieved by resting, squatting, or leaning forward, which straightens the lordosis and increases the transverse diameter of the lumbosacral canal.⁸⁰ With progressive stenosis, the distance walked before claudication ensues decreases, thus making this symptom a useful clinical parameter.

Evaluation

In more advanced cases of stenosis, neurological abnormalities, such as weakness of the lower extremities (particularly the dorsiflexors of the toes and feet) and hypoesthesia, often at a truncal level, persist at rest. Occasionally, a partial Brown-Séquard syndrome is seen. Spasticity and hyperreflexia of the legs typically indicate compression of the thoracic cord but may signify coexistent cervical compression. The results of neurological examination in patients with claudication often remain otherwise normal until late in the disease process. Straight leg raising and reverse straight leg raising test results are usually normal unless a superimposed disk problem is present. Therefore, the thoracolumbar spine must be evaluated in all symptomatic achondroplastic patients, even in the absence of neurological findings. The study of choice is MRI.

Indications for Surgery

We recommend decompressive laminectomies for achondroplastic patients whose ambulation is severely limited by claudication, who show significant weakness (other than mild extensor hallucis longus weakness) at rest, or who have urinary incontinence or urgency that is attributable to compression of the cauda equina or spinal cord. There is a high incidence of urologic complications after laminectomy and a high correlation between preoperative urologic abnormalities and postoperative complaints. We therefore recommend thorough urologic testing as part of the preoperative evaluation for patients undergoing laminectomy.

Operative Management

At our institution we use a common high-speed drill technique for both craniocervical and spinal decompression. Although laminectomy is a widely practiced spinal procedure, we believe that modifications that address several of its shortcomings, including inadequate decompression and secondary spinal instability, are necessary for its use in achondroplastic patients.⁸¹^,⁸²

Cervicomedullary Compression

Craniocervical surgical decompression for cervicomedullary compression in children with achondroplasia has been used at several centers and generally yields good results.⁵⁷^,⁸³ Decompression of the cervicomedullary junction has been shown to bring about dramatic and sustained improvement in neurological and respiratory function when it is combined with other therapy as needed for respiratory compromise.⁶¹^,⁶³ The surgeon must understand the anatomic difficulties presented by achondroplastic patients (Fig. 219-2). Clinical evaluation, moreover, is frequently difficult because achondroplastic patients can have respiratory difficulties for many reasons, some of which are unrelated to the neurological compromise. Long-term follow-up data that would allow definitive assessment of craniocervical decompression have also been lacking.⁶⁵ As with any surgical procedure, detailed prior consultation must be conducted with the parents to inform them of the potential risks and expected benefits for their achondroplastic child.

FIGURE 219-2 Illustration demonstrating normal findings (left) and findings in the cervicomedullary junction in patients with achondroplasia (right). Note the horizontal orientation of the posterior rim of the foramen magnum in achondroplasia, which is causing kinking of the cervicomedullary junction. Also note obstruction of the cerebrospinal fluid space at the level of the foramen magnum, which is causing obstructive hydrocephalus. The area of bone removed during cervicomedullary decompression is depicted in green in the illustration of achondroplasia.

(Used by permission from Bagley CA, Pindrik JA, Bookland MJ, et al. Cervicomedullary decompression for foramen magnum stenosis in achondroplasia. J Neurosurg. 2006;104:166; illustration by Ian Suk.)

A large operating room is used to accommodate all the equipment and personnel necessary to decompress the craniocervical junction. Before coming to the operating room, the patient is sedated and antibiotics are administered. Patients also receive steroids preoperatively to protect the spinal cord and brainstem from local trauma. Patients are positioned prone on the operating table, with the head and neck carefully supported in slight flexion with the use of a padded pediatric horseshoe headrest. Upper extremity somatosensory evoked potentials are assessed routinely during positioning, as well as during the decompression procedure itself. The anesthesiologist is situated for easy access to the patient and is also in close proximity to the evoked potential monitoring equipment.

To prevent postoperative CSF leaks, for which achondroplastic patients are at significant risk, a right frontal external ventricular drain is inserted in the most obviously symptomatic hydrocephalic children. As an alternative to more invasive measures of ICP, we have become more reliant on intraoperative ultrasonography to determine the adequacy of CSF flow once the bony decompression has been completed.

A midline suboccipital incision is made for the decompression, and the ligaments and musculature are dissected subperiosteally to expose the occiput and the spinous processes and laminae of C1 and C2. The arch of C1 is then removed with a high-speed drill and small curets (Video 219-1). One frequently sees a thick, fibrous band or pannus above the level of C1, which should be left in place during dissection of the bone to protect the underlying dura and cord from incidental injury. Compression of the cervical cord occasionally necessitates removal of the arch of C2 or extension of the decompression even farther in a caudal direction. The posterior rim of the foramen magnum is thinned gradually with a high-speed drill and removed with small, straight and angled curets. Invariably, the bone of the foramen magnum is thickened, oriented more horizontally than usual, and severely indenting the underlying dura. Once decompression of the bone is complete, the fibrous pannus is removed as well, which often reveals a transverse dural channel that offers dramatic evidence of the extent of the dural constriction; consequently, adequate attention must be paid to the soft tissue aspects of the decompression. If a duraplasty is being performed, the dura is opened in the midline along the area of constriction. Adequate cord pulsations and CSF flow can be confirmed, and a dural patch graft can be performed with paraspinous fascia, pericranium, or commercially available human cadaveric dura. A watertight seal is confirmed, and the wound is copiously irrigated and closed in several layers; it is not drained, however, to avoid potentiating the development of a CSF fistula. In our early experience, CSF leaks developed in a significant number of patients despite efforts in making the closure watertight and required a ventriculoperitoneal shunt for definitive treatment. Therefore, it has been our recent practice to not routinely perform a duraplasty as part of the decompression. Instead, after removal of the bone and fibrous tissue, we use intraoperative ultrasonography to assess for residual compression of the cervicomedullary junction and to visualize the CSF spaces. If ultrasonography reveals adequate decompression and ample CSF spaces, we do not open the dura for a duraplasty. In our recent experience, only 2 of 43 patients underwent duraplasty, and both cases were complicated by postoperative CSF leaks.⁸³

After completion of the operation, somatosensory evoked potentials are evaluated before the patient is undraped in the event that it is necessary to re-explore the wound. Once movement is confirmed in all four extremities, the patient is sent to the pediatric intensive care unit. Extubation is often performed immediately postoperatively; however, in some cases, facial and laryngeal edema makes this procedure inadvisable for 12 to 24 hours. After surgery, primary attention is directed toward monitoring ICP as part of postoperative nursing care if a ventriculostomy is present.

The surgeon should bear in mind several important pitfalls when undertaking CMD in achondroplastic patients. First, the patient’s head must not be overflexed during positioning because such a position often reduces the subarachnoid space at the cervicomedullary junction. Second, the surgeon should avoid placing any instruments beneath the posterior arch of C1 or beneath the rim of the foramen magnum, even though these patients are pretreated with steroids. The cervicomedullary junction is already under tremendous constrictive pressure, and even brief introduction of instruments into the already compromised space can be disastrous. The spinal cord and brainstem of a child with achondroplasia are small; therefore, the decompression should be correspondingly small. To gain an accurate conception of the size of the spinal cord and brainstem for performing an appropriate decompression, the surgeon must study results of the preoperative MRI. Moreover, the decompression must be extended not only along the dorsal surface of the cervicomedullary junction but also sufficiently along the lateral dimensions of the medulla to adequately decompress the stenosis at the level of the foramen magnum.

Frequently, engorged veins exist beneath the ligamentum flavum; they are located dorsally and laterally and sometimes insinuate through the ligamentum flavum. These veins can create enormous bleeding and must be controlled rapidly if they are compromised. The possibility of air embolization through these veins also exists, and the surgeon should ensure that the patient is not in any degree of reverse Trendelenburg position. Finally, once the bony decompression is complete, the underlying dura, which in many cases is itself severely constricted, must be checked carefully. The dura is often fused with the ligamentum, and this soft tissue band serves to constrict the underlying neural tissues, even without the presence of the overlying bone. If this situation occurs, the band must be divided and a dural patch placed. When the band is divided, a significantly engorged annular sinus is commonly present beneath the foramen magnum; the band should be slowly divided and measures taken to control the annular sinus as it is encountered.

Spinal Stenosis

Decompression of the achondroplastic spinal canal is difficult because of the extent and severity of the stenosis. The quantitative magnitude of this stenosis has been well documented.⁸⁴ Moreover, poor postoperative results of spinal decompression were relatively common in achondroplastic patients.^81,^82,⁸⁵ Before the era of computed tomography (CT) and MRI, the degree and extent of the spinal compression were often not appreciated with conventional myelography because of the lack of adequate diffusion of contrast medium. Insertion of bulky instruments under the laminae during the conventional techniques frequently traumatized neural tissue. Another source of poor results was postoperative instability resulting from overly wide laminectomies.

The following procedure has been used at our institution with good results. MRI provides adequate anatomic delineation and intrathecally enhanced CT is used, but myelography is sometimes helpful as an adjunct. Based on the results of these studies, the surgeon can devise an operative plan for adequate decompression that includes at least three segments above the level of demonstrated blockage and three segments below (or to S2). The incision is midline, and dissection is carried subperiosteally to expose the spinous processes, laminae, and facet joints over the extent of the area to be decompressed. When adequate exposure is achieved, the laminae immediately medial to the facet joints are gradually thinned with a high-speed drill so that a groove approximately parallel to the longitudinal axis of the spinal column is formed. The drill head is held at an angle of about 45 degrees to the laminal surface rather than 90 degrees; this angle offers the surgeon the control necessary to avoid accidental perforation of the laminae. The groove is deepened until the dura can be seen through the thinned laminal mantle. Drilling is then concluded in this area and is continued on the opposite side in similar fashion. An opening is made at the caudal part of the groove on the first side, a thin surgical punch is inserted into the epidural space, and the laminectomy is carried along the groove. A small Leksell rongeur is inserted through this opening to complete the laminectomy by lifting the laminae off in piecemeal fashion. This technique minimizes dural tearing, preserves the facet joints, and protects the neural structures from injury. If the facet joints are violated on both sides, spinal fusion is necessary. In our experience, because the primary compressive dimension in the achondroplastic spine is cephalocaudal, not lateral, undermining of the facets is not recommended.

Removal of the spinous processes plus detachment of the paraspinal soft tissues creates a large and deep void, particularly at the lumbosacral junction. To minimize the risk of pseudomeningocele formation, an overlapping closure with paraspinal muscle in which the muscle masses are brought in to fill the dead space was developed. First, a paraspinal muscle encased in fascia is partially detached from the iliac spine and the lumbosacral laminae with the use of a split-thickness incision, if necessary, to mobilize the required tissue, which is then reflected around its pedicle. The edge of the flap is brought down to the opposite lateral end of the lamina and attached with heavy sutures to the inferior part of the paraspinal muscle mass on that side. The superior part of the muscle mass on that side is then retracted over the first flap, thereby completing the muscle closure and collapsing the void. As with craniocervical decompression, somatosensory evoked potentials are monitored intraoperatively. Postoperative care is generally routine, with the exception of a high incidence of transient urologic dysfunction. Staff should be advised to help the patient be prepared for this possibility.

A different viewpoint concerning the ideal strategy for spinal laminectomy recommends a wide decompression with foraminotomies and mandatory undermining of the facets.⁸⁶ The rationale for this strategy is that compression is lateral as well as longitudinal in the achondroplastic spinal canal. Our experience, however, does not bear this out despite the hypothesized impact of small lateral recesses in the achondroplastic vertebral foramen. Moreover, the stabilization problems encountered with wide laminectomies can be more debilitating than the initial disease. In fact, there is no reason why every spinal level could not be subjected to the laminectomy we describe without the need for concomitant spinal stabilization. Therefore, in light of the results obtained at our institution with spinal decompression, we believe that a narrow, extensive laminectomy is the spinal decompression method of choice for achondroplastic patients. Some surgeons have adopted this operative technique for their nonachondroplastic patients as well. The goal is adequate decompression of neural elements and not simply enlargement of bony canals.

Thoracolumbar kyphosis is commonly present in achondroplastic children, with a prevalence as high as 87% in children between 1 and 2 years of age.⁸⁷ Most of these patients, however, remain asymptomatic and may demonstrate spontaneous resolution by 3 years of age. Effective control of progressive thoracolumbar kyphosis of 30 degrees or greater with orthotic bracing has been demonstrated in the pediatric age group.⁸⁸ However, decompression of the spine over areas of existing kyphosis warrants concern for progression as a result of removal of the posterior ligamentous tension band. Ain and colleagues demonstrated postlaminectomy kyphosis in 10 consecutive skeletally immature achondroplastic children who underwent decompressive laminectomy for spinal stenosis.⁸⁹ They also demonstrated that achondroplastic children with thoracolumbar kyphosis can safely undergo instrumented fusion concomitant with laminectomies for neurological symptoms.⁹⁰ Therefore, it is our current practice to recommend instrumented fusion in all skeletally immature achondroplastic children with thoracolumbar kyphosis at the intended levels of decompression.

Outcome

Craniocervical Decompression

We reviewed our experience in 43 children with achondroplasia who underwent CMD with a mean follow-up period of 62.5 months.⁸³ There were no surgery-related deaths in these children. The most common complication was CSF leaks, which occurred in 7 patients from either the ventriculostomy site or the suboccipital incision. There were no instances of clinical deterioration immediately after the operation; however, 5 patients experienced worsening after a period of improvement. Imaging in these patients revealed recurrent stenosis at the foramen magnum, and all of them responded well to a revision decompression. The remainder of patients did not experience recurrence of clinical symptoms after the initial decompression during the extended follow-up of the study. We therefore concluded that CMD can be safely performed for cervicomedullary compression in achondroplastic patients with minimal morbidity, provided that the surgeon recognizes the intricacies of the anatomy in this population.