Positioning and Sterile Preparation

The operating room must be kept warm to avoid hypothermia. All prep solutions are warmed to body temperature. The infant is intubated in the lateral position or supine with a doughnut-sponge on the back to accommodate the sac. The pelvis is propped up with a horizontal roll to render the lumbar spine hyperlordotic and give maximum relaxation of the skin and soft tissues. The membranous portion of the sac, containing the exposed neural placode, is irrigated profusely with antibiotic-reinforced saline (betadine compounds have been shown to be neurotoxic), and the surrounding skin scrubbed with the usual antiseptic agents.

Opening the Sac

Magnification is used from the very beginning. The sac is entered through the diaphanous leptomeningeal membrane halfway between the margin of healthy skin and the edge of the placode (eFig. 60-4). Neural tissue of the placode is recognizable by its pink, felty surface, transverse wrinkles, and a straight, longitudinal median raphe (Fig. 60-1A) and since the epithelialized membrane itself is relatively avascular, any substantial bleeding from the cut edges signifies breaching of the neural tissue (Fig. 60-1B). Bleeding is controlled with a pair of ultrafine irrigating bipolar cautery forceps. After the initial gush of CSF and collapse of the cyst, the edge of the neural placode is gently flipped up to identify the ventral nerve roots. Several crossing blood vessels may have to be coagulated to free the placode margins. The pearly epithelium must be meticulously trimmed circumferentially from the placode to avoid later occurrence of inclusion dermoid cyst. At the caudal extreme of a terminal placode, the epithelial membrane may remain thin, or one may encounter a band-like thickening probably representing remnant of the medullary cord that must be divided to free the tip. At the rostral extreme of the placode, careful incision of the epithelium–neural tissue junction on both sides exposes the delicate bevel-shaped transition between the neurulated cylindrical spinal cord and the un-neurulated, flat placode (eFig. 60-5). At the apex of the bevel, the central canal of the normal cord can be seen unfurling into the median raphe of the placode, from which CSF sometimes slowly oozes.

FIGURE 60-1 A, Open myelomeningocele with a terminal neural placode. Note longitudinal median raphe on the neural placode. Small arrowheads outline margin of placode. Glistening leptomeningeal membrane stretches between placode and skin. Large arrowheads outline skin edge. Triangular end of the sac is toward the anus. B, Cutting into the leptomeningeal membrane just outside the margin of the placode. C, Neural placode after dural edges from all sides have been defined and dissected free, ready for dural closure. D, Closed dural tube. Paraspinous muscles and periosteum of bifid neural arch are exposed on each side of the closed dural tube. Anus to the right of picture.

Handling the Neural Placode

The neural placode is always handled gently with jeweler’s micro-forceps. If the placode is pliable and thin enough, it is rolled on itself and sewn up with 8-0 nylon sutures through the delicate pial edges (eFig. 60-6). There is no evidence that this improves neurologic function, but it reduces the “sticky” surface from a flat plate to a seam and theoretically lessens the chances of later tethering. If the placode is too thick or too stiff, it should not be forced to roll up for fear of strangulation. Also, tugging too hard on the proximal spinal cord will cause ascension of the neurologic level. In addition, the caudal end of the placode must be checked for the presence of a neurulated cord in the rare case of a segmental placode (see below).

Paradoxically, small and compact neural placodes are likely functional and therefore should be treated with extreme care, whereas large, thin, and “spread-out” placodes commonly found in middle to high thoracic lesions are usually nonfunctional. It is sometimes advisable to cordectomize these large placodes, for whilst they do not convey volitional movements, the decentralized neuronal pools within them can produce pathologically hyperactive local reflexes underlying high-pressured, spastic bladders and prominent ureteral reflux. Reflux is less prevalent when the bladder is rendered flaccid following elimination of this type of placode.

Dural Closure

The margins of the dural flaps are created by sharply incising the ventral dura from the lumbosacral fascia and periosteum (Fig. 60-1C and eFig 60-7; see also eFig. 60-4A and B), and a new dural tube is reconstructed in the midline (Fig. 60-1D). One aims to obtain as capacious a dural sac as possible commensurate with the size of the placode, the theory being if the placode passively flops freely within a large CSF space, it is less likely to adhere to the dorsal dura. This is almost always achievable with the patient’s own dura, even if some of the periosteum overlying the bifid neural arches have to be mobilized with the dura proper to enlarge the sac. In the rare event of insufficient dura, bovine pericardium can be used as a graft because its texture is compatible with newborn dura and because it seldom has suture-hole leakage of CSF. At the end of closure, the suture line is tested with a Valsalva maneuver held at a pressure of 30 to 35 cm of H₂O for 10 seconds.

Skin and Myofascial Closure

The true size of the skin defect is only apparent after completely excising the epithelialized membrane back to full thickness skin. Defects up to 5 or 6 cm diameter can usually be closed primarily after the subcutaneous layer is mobilized a short distance centrifugally, just enough to reduce the tension on the skin edges. The subdermal layer is closed with interrupted absorbable sutures, and the skin with fine nylon sutures. A number of surgical techniques have been developed to minimize suture line tension in large defects. Lateral relaxing incisions with bipedicle flap closure in the midline have been effective,¹¹ but the relaxing incisions themselves then require skin grafting at the same time or at a later date. Complex multiple rotation skin flaps have also been tried, but this necessitates extensive skin undermining and still does not altogether eliminate all tension spots. For impossibly large defects, I favor using composite skin–muscle (myocutaneous) flaps. There are three advantages to this technique. Cadaver vaso-latex studies show that a rich vascular anastomosis exists in the skin overlying the gluteus maximus and the latissimus dorsi muscles on each side and across the midline; the muscles themselves are supplied by the gluteal and thoracodorsal arteries, respectively. As long as these arterial pedicles are preserved, blood supply to the lumbosacral skin is well maintained when the four muscles are apposed in the midline, even if the short paraspinous arterial perforators deep to the muscle are taken.¹² Secondly, because there is no undermining of the subcutaneous tissues, the skin tension is mostly absorbed by the muscle closure. Finally, this method results in a triple-layered (muscle, subcutaneous tissue, and skin) closure with extra insurance against CSF leak. The flaps survival rate even for enormous defects is greater than 92% with this technique, and the blood loss and extra anesthetic time are quite acceptable.¹²

It is important to avoid large wound seroma or hematoma, which increases skin tension and prevents flap adherence to the underlying tissue from which revascularization for the flap must be derived. If the flaps are large and “wet,” a small drain without negative suction may be left in for 24 hours. Suction drains may perpetuate a CSF fistula and are best avoided.

Early Complications (First Postoperative Week)

The operative mortality for children undergoing repair of an ONTD should be close to 0.^13–15 The most common cause of postoperative death is related to hindbrain dysfunction (73%),^10,16 but this seldom occurs acutely in the first week of life. Most of the immediate complications pertain to the wound itself.

Wound Dehiscence

A study of the nutritional status of newborn infants who have had myelomeningocele surgery using body weight, nitrogen balance, serum protein, and total lymphocyte count as parameters, showed that these neonates undergo an initial period of severe catabolic changes that do not readjust themselves for as long as 1 month after surgery. This nonspecific catabolic response is caused by rises in circulating levels of ACTH, cortisol, thyroxin, growth hormone, and antidiuretic hormone, stimulated by the extreme stress of surgery, general anesthesia, and blood transfusion.⁸ During this period, the resistance to infection is lowered, and all anabolic processes, including wound healing, are temporarily slowed. This metabolically unstable time also coincides with feeding difficulties caused by hydrocephalus, postoperative ileus, neurogenic dysphagia (due to brainstem compression), and prematurity. It is no surprise that wound dehiscence is the single most common complication during the first postoperative week.⁸

Local factors, mostly avoidable, also contribute to this problem. A large sac means higher wound tension and precarious blood supply. An untreated kyphus adds stretching to the suture line and aggravates the local ischemia. Any additional external pressure caused by a tight dressing or improper patient positioning also interferes with healing.

It is common to see erythema along an 8- to 10-mm strip on either side of the suture line, particularly in areas of high tension. Sometimes the skin flaps may even look deep red to dusky as a result of venous stasis. These color changes often pass after a few days. When there is necrosis, the intensely dark red skin edges will turn black, but the necrosis may be limited to the epidermis, and the dermis and subdermis may survive, which should make adequate coverage. If full-thickness necrosis occurs, the blackness extends farther laterally. The junction between dead and viable skin demarcates, and the surrounding skin becomes erythematous and edematous. The sloughing skin edge also begins to pull away from the sutures, and serous exudate from subjacent fat necrosis seeps from the exposed subcutaneous tissues. Demarcation and sloughing are usually complete by the 7th or 10th day.

Sloughing of only the epidermis in small areas requires only simple dressing changes because the wound eventually epithelializes over the underlying dermal and subdermal layers. Skin grafting is unnecessary. If the skin necrosis is full thickness but there is healthy muscle underneath, the wound edges should be carefully debrided back to bleeding skin. It may then be dressed for second intention healing from below, but this will take some time and delay CSF shunting. A faster way would be partial-thickness skin grafting, which should take well over a well vascularized bed. If full thickness necrosis exposes the dural tube, some measure of immediate coverage must be instituted to prevent desiccation and meningitis. This usually means a more radical and extensive flap rotation or even pediculated full thickness skin flap grafting.¹⁷

Finally, parenteral or enteral hyperalimentation should be set up to ensure adequate nutrition.

Wound Infection

Considering how badly the exposed neural placode is contaminated during and shortly after birth, it is surprising how rarely wound (extradural) infections occur after closure of an ONTD. The wound infection rate is about 1.5% to 2.5%, which is only moderately higher than clean neurosurgical procedures.⁸ However, if one counts the intradural infections, the infection rate rises to 7% to 10% even for early closure.^15,18

The initial period of obligatory catabolism in these infants lowers both their cellular and humoral defenses. In as much as the infant must depend on transplacentally acquired maternal antibodies during the first 3 months of life, and IgA does not cross the placenta well, infections from enteric bacteria are particularly common. Chief among local predisposing factors to infection is a large sac with redundant, folded membranes. It is virtually impossible to sterilize all the creases and folds of the wobbly sac and the placode is therefore recontaminated during surgery. During closure, the contaminated placode is put inside a closed intradural space, which explains why intradural infection is more common than extradural infection.

Systemic signs of sepsis due to gram-negative meningitis are usually present 1 to 3 days after closure. In neonates, these early signs tend to be nonspecific, such as poor feeding, lethargy, or an ashen complexion. It is more common to see hypothermia than pyrexia, and the systemic white blood cell count often drops below 4000/mm.³ If the dural sac is well invested with a myocutaneous coverage, an intradural abscess may eventually form without any external signs. It is important to obtain CSF for culture from a ventricular tap if there is clinical suspicion of sepsis, for the long-term prognosis of gram-negative ventriculitis in the newborn depends almost solely on the promptness of diagnosis and treatment.

If the infection is confined to the extradural space, the wound will become red and fluctuant on day 5 to 7. The surrounding skin will also appear edematous but, unlike CSF infection, there is often no systemic signs of sepsis and the infant may continue to feed and move normally. A red, fluctuant wound should be diagnostically aspirated for purulent material. An abscessed wound must be opened immediately, widely debrided, irrigated with antibiotic solution, and reclosed over suction drains. Associated vascular occlusion and myonecrosis may require refashioning of a new myocutaneous closure. Depending on whether the CSF indices indicate intradural infection, the dura may have to be opened to rule out an intradural abscess. The patient is then put on broad spectrum, CSF-penetrating antibiotics.

Cerebrospinal Fluid Leak

The dura nearest its lateral margin may be severely attenuated in large myelomeningocele defects. This plus the often tense and precarious myocutaneous closure makes large lesions particularly prone to leak CSF. Also, the timing of the slowly climbing CSF pressure happens to coincide with weakening of the tenuous suture line, around the 5th to 8th postoperative day. A small amount of transdural CSF leak probably occurs through the suture holes in most cases, considering the thinness of the newborn dura. The appearance of slight fluctuance under the skin flaps during the first few postoperative days is likely due to a combination of CSF and blood. As long as the skin closure holds, and there is no outward leak of CSF, there is no risk of infection. This small amount of seepage is self limiting. A large transdural leak causes a tense subcutaneous accumulation that will eventually threaten the viability of the suture line. When CSF actually breaks through the skin barrier, the risk for gram-negative infection rapidly rises, and treatment must be promptly initiated.

A shunt is effective in preventing CSF leak but is not recommended after the leak has sprung, especially if the leak has already breached the skin closure. Even a low pressure shunt maintains a constant lumbar CSF pressure of 5 to 6 cm H₂O, still considerably higher than that in the subcutaneous pocket, which is near atmospheric. The preferential passage of CSF is still out through the back wound and not through the shunt. If a CSF leak persists in the presence of a shunt, the latter becomes infected sooner or later. The external ventricular drain (EVD) is a much better means of decompression after a substantial leak has already existed. The drainage chamber can be lowered to subzero pressure to siphon CSF away from the back wound. With elimination of outward leak, the probability of infection is mitigated. The skin edges can now be oversewn, and the infant may have to be sedated to minimize the milking action of muscles overlying the thecal sac. Many leaks can be successfully managed by such measures without a reoperation. If the leak persists post-EVD, the wound needs to be explored to close the dural defect.

Dorsal Lipoma

The lipoma–cord interface is entirely on the dorsal surface of the lumbar spinal cord, sparing the distal conus (Fig. 60-2A). The junctional demarcation between lipoma, cord, and pia, the fusion line, can always be traced neatly along a roughly oval track, separating fat from the dorsal root entry zone (DREZ) and dorsal nerve roots laterally (Fig. 60-2B and eFig 60-10). The lipoma therefore never contains nerve roots. The lipomatous stalk runs through an equally discrete dorsal dural defect to blend with extradural fat. The uninvolved conus often ends in a thickened filum terminale.

FIGURE 60-2 A, Dorsal lipoma on MRI. Sagittal image shows intact conus caudal to lipoma stalk. Axial images: upper shows site of lipoma attachment to cord; lower shows free conus just caudal to the level of lipoma attachment. B, Intraoperative picture shows neat oval fusion line around lipoma–cord interface on a horizontal plane. Note intact conus and caudal sacral roots. C, Transitional lipoma. Left: Sagittal MRI shows lipoma begins dorsally but involves entire conus. Ventral side of neural placode is free of fat. Right: Plane of the fusion line begins dorsally and then cuts obliquely toward the tip of the conus. The array of DREZ and dorsal roots is also forced to slant dorsoventrally. D, Transitional lipoma. Intraoperative picture showing massive lipoma but very distinct dorsoventral fusion line separating fat from the DREZ and dorsal roots, which always lie lateral and ventral to the fusion line. The ventral side of placode is always free of fat in a regular transitional lipoma. DREZ, dorsal root entry zone.

Transitional Lipoma

The rostral portion of this type is identical to that of a dorsal lipoma, with a discrete fusion line and easily identifiable DREZ and dorsal roots. Unlike the dorsal type, however, which always spares the conus, the transitional lipoma then plunges caudally to involve the conus as the plane of the fusion line cuts ventrally and obliquely towards the tip of the conus likened to making a slanting, beveled cut on a stick (Fig. 60-2C). The lipoma–cord interface thus created may be undulating and tilted so that the neural placode is rotated to one side or even spun into a parasagittal edge-on orientation, but the neural tissue is always ventral to it and the DREZ and the nerve roots are predictably localizable lateral and ventral to the fusion line and therefore do not course through the fat (Fig. 60-2D). There may or may not be a discrete filum. The dorsal dural defect extends to the caudal end of the thecal sac and may be much larger on the biased side.

Terminal Lipoma

Unlike the dorsal and transitional types, terminal lipomas insert into the caudal extremity of the conus without blending with the spinal cord or its root entry zones. All the sacral roots unmistakably leave the conus rostral to the lipoma, and in most cases the conus itself looks normal. The dural sac and the dorsal myofascial coverings are intact. The lipoma either replaces the filum entirely, or is separated from the conus tip by a short, thickened filum (Fig. 60-3A and B).

FIGURE 60-3 A, Terminal lipoma. Drawings showing insertion of fat directly on the end of the conus, which is itself intact. Resection plane is right through a clear transition plane. Typical MR appearance is shown at right. B, Intraoperative pictures for terminal lipoma. Top shows the typical terminal lipoma. Middle shows separation of “hitch-hiker” nerve roots from lipoma. Lower shows conus after transaction through separation plane. C, Choatic lipoma. Intraoperative picture showing fat ventral to placode and on one of the sacral roots (arrow). Note absence of discrete fusion line. D, Lipomyelomeningocele with the lipoma stalk, cerebrospinal fluid sac, and part of the conus extending out of the spinal canal through a dorsal defect.

Chaotic Lipoma

This previously undescribed type is so named because it does not “follow the rules” of either the dorsal, transitional, or terminal lipoma. It begins dorsally in an orderly fashion as in a dorsal or transitional lipoma, but its caudal portion is ventral to the neural placode and does engulf neural tissue and nerve roots (eFig. 60-11). The fusion line may be distinct rostrally but quickly becomes blurred distally, and the location of the DREZ and nerve roots is less predictable. The moniker “chaotic” depicts the sometimes confusing blend of the ventral fat and neural placode, and the often impossible task of separating fat from neural tissue at surgery (Fig. 60-3C). Chaotic lipomas are uncommon but are characteristically seen with sacral agenesis.

The literature^30,31 describes one other lipoma type, the lipomyelomeningocele, in which part of the distal conus extends into the extraspinal compartment, dragging with it a small collar of dural sac (Fig. 60-3D). The basic structure is either that of a transitional or a dorsal lipoma. Accordingly, we choose to classify this type as either a transitional or dorsal lipoma with a descriptive qualifier of “extraspinal extension.”

Embryogenesis of Dorsal and Transitional Lipomas

In the embryo, a progressive disparity exists between the spinal cord and vertebral column as a result of the faster growth rate of the latter.^32–35 The caudal end of the cord ascends gradually from opposite the coccyx in the 30-mm human embryo to the L₁–L₂ level at birth.^34–37 Proper ascent of the cord requires a well-formed neural tube and a smooth pia-arachnoid covering. If during early development a dorsal defect exists in the dura (duraschisis) and neural tube (myeloschisis), mesodermal elements from the surrounding mesenchyme will enter the dural sac and form attachment with the sliding neural tube in the form of a fibro-fatty stalk, resulting in its entrapment. This theory features a fundamental defect in neural tube closure during primary neurulation (secondary neurulation does not involve dorsal neural fold closure), and thus applies only to the dorsal and transitional lipomas (see below). It is compatible with the observation that these two types of lipomas are always associated with neural arch defects.

The embryologic error leading to the mesodermal invasion of the neural tube probably lies in premature disjunction between the cutaneous and neural ectoderms^38–40; that is, the separation of one from the other occurs before the converging neural folds fuse with each other. This allows the paraxial mesenchyme to roll over the still gaping neural folds and enter the central canal. Once contact between mesenchyme and ependymal neuroectoderm is made, further closure of the neural tube is permanently prevented and a segmental dorsal myeloschisis is created (eFig. 60-12A and B). Alternatively, the fault may lie in a delay in neural folds fusion secondary to an insufficiency of the paraxial mesoderm in impelling their dorsal convergence,^41–47 so that ectodermal disjunction again precedes neural folds fusion. Lastly, faulty fusion of the neural folds due to metabolic disturbance of the cell membrane-bound glycosaminoglycans, which are vital to cell–cell recognition and adhesion,^48–51 could likewise reverse the temporal relationship between disjunction and neural folds fusion.

Experimental studies show that the pluripotent mesenchyme forms derivatives according to the inductive properties of the adjacent neuroectoderm (eFig. 60-12C).^6,9 The ependymal side of the neural tube induces mesenchyme to form fat, muscles, collagen, and occasionally bone and cartilage, whereas the outer surface of the neural tube induces the formation of meninges.⁵² However, no dura can now form over the dorsal opened portion of the neural tube, and the dural defect neatly surrounds the evolving lipomatous stalk, which tethers the neural tube to the subcutaneous adiposity. In like manner, deficiencies in the overlying myofascial layers (from myotomal mesoderm) and neural arches (from scleromesoderm) also neatly surround the lipomatous stalk (eFig. 60-12D).

Within the neural tube, the intramedullary fat and muscles fuse with the developing alar and basal plates. Since the dorsal root ganglions develop from neural crest cells at the outer aspect of the neural fold lateral to the site of failed fusion, the dorsal nerve roots grow outward ventrolateral to, but never traverse, the lipomatous stalk. The DREZ must correspondingly lie very near, but always lateral to, the exact junctional boundary between lipoma and spinal cord. This boundary, called fusion line, is of tremendous surgical significance^22,53,54 (eFig. 60-12D). Meanwhile, the cutaneous ectoderm, long detached from the neuroectoderm, heals over in the dorsal midline to form wholesome skin over the subcutaneous lipoma.

The genesis of a dorsal lipoma perfectly exemplifies mistimed disjunction during primary neurulation. Its fibro-fatty stalk always involves cord segments above the conus, which mainly forms from secondary neurulation. Furthermore, failure of primary neural tube closure appears to be segmental, normal closure takes place “business-as-usual” immediately following the abnormal event. This “square pulse” nature is illustrated by the fact that the sharp fusion line between fat, spinal cord, and pia-arachnoid can be neatly traced circumferentially around the lipomatous stalk^53–55 (see Fig. 60-2B and eFig. 60-10). Dorsal lipomas therefore result from a segmental closure abnormality involving only primary neurulation. They are found in less than 15% of spinal cord lipomas in our series.^22,26

In transitional lipoma, the myeloschisis involves much more than an isolated segment of the primary neural tube. Even though its rostral part resembles the dorsal lipoma, the involvement of the whole of the caudal spinal cord means that not only primary but also secondary neurulation have been profoundly disturbed by the mesodermal invasion. This is supported by the observations that in many transitional lipomas the filum is incorporated into the distal fat, and within the lipomas are often spaces resembling the vacuoles within the secondary neural tube. Also, while the rostral part of the transitional lipoma is always dorsal and aptly reflects premature disjunction of primary neurulation, the distal part sometimes involves both the dorsal and ventral aspects of the conus, a situation compatible with misguided mesenchymal inclusion during the much less orderly events of secondary neurulation. Intramedullary mesenchyme may migrate within the neural tube after invasion and travel caudally across the boundary from the primary to the secondary neural canal, since the two neural canals are in continuity.⁵⁶ In fact, the hypothesis that the rostral part of the transitional lipoma arises from aberrant primary neurulation (involving only the dorsal cord) and the caudal lipoma arises from abnormal condensation of the secondary neural cord (affecting more ventral aspects of the conus) furnishes at least one explanation for the dorsoventral obliquity of the lipoma–cord interface.

Embryogenesis of Chaotic Lipomas

Chaotic lipomas do not quite fit into either the dorsal or transitional schema. They often do not have a distinct dorsal part with the symmetry of a dorsal lipoma, and the lipoma–cord interface is irregular and ill-defined, with fat running through the neural placode to the ventral side in large and unruly measures. Even in the context of the less orderly transitional lipoma, the interplay between lipoma and cord in this type of lesion seems to be in constant chaos.

This degree of anatomic unpredictability in chaotic lipoma and its strong association with caudal agenesis (82% in our series²²) suggest that the embryogenetic error occurs during the early stage of secondary neurulation as part of the general failure of the caudal cell mass (eFig. 60-13).^57,58 Secondary neurulation comprises three distinct stages: (1) condensation of neural material from the caudal cell mass to form the solid medullary cord, (2) intrachordal cavitation of the medullary cord^56,58,59 and its integration with the primary neural tube, and (3) partial degeneration of the cavitary medullary cord through massive apoptosis to result in the thin filum terminale.^34,56 It is possible that formation of the chaotic lipoma involves the entanglement of lipogenic mesenchymal stem cells with cells from the caudal cell mass during aberrant condensation of the medullary cord, forming an inseparable mixture of neural tissue and fat, with nerve roots projecting out haphazardly.²²

Embryogenesis of Terminal Lipoma

Terminal lipomas result from abnormal secondary rather than primary neurulation, as evidenced by the unexcepted rule that the lumbar and upper sacral cord segments, products of primary neurulation, are never affected in a terminal lipoma. Furthermore, dorsal myeloschisis and duraschisis, both hallmarks of failed (primary) neural fold fusion, are never seen. Lastly, the terminal lipoma either replaces or forms part of a thickened filum, which temporally places the pathogenetic process at secondary neurulation. The fact that the distal conus in terminal lipomas always remains fat-free argues against an abnormal condensation phase during early secondary neurulation. On the other hand, the terminal lipoma often contains (disorganized) spinal cord elements and ependymal tubules,^60,61 which suggests rather an incomplete or ineffective degeneration phase at late secondary neurulation.

Step 1. Exposure

The skin and soft tissue incision should go straight through the subcutaneous lipoma if one is present. Removing this will leave behind a large subdermal space into which CSF could collect under tension and consequently hinder wound healing. Frequently a discrete fatty stalk connects the subcutaneous with the intraspinal lipoma through a defect in the lumbodorsal fascia. This stalk is continuous with the spinal cord and should not be tugged on during fascial dissection.

The upper extent of the bony exposure should include one level above the rostral end of the lipoma. This reveals for proper orientation the “last” normal set of nerve roots and DREZ before starting lipoma resection. Wide laminectomy is essential to afford full access to the lateral edges of the dural sac (see below). Visualization of the normal dura rostral to any lipoma gives a depth perspective as to how far out neural tissue and CSF sac might have extruded beyond the plane of the dura. The heavy bulk of extradural fat could then be safely lopped off to give room for intradural dissection and lighten the tug on the conus.

Step 2. Detachment of Lipoma from Dura

The dura rostral to the lipoma is opened in the midline. For dorsal lipoma, the dural incision is carried circumferentially around the discrete lipoma stalk and then down the middle again to expose the conus (eFig. 60-10). For transitional lipoma, the dura is cut close to the lipoma edge on each side as far caudal as possible, though the two side incisions seldom could meet distally. The dural edges are then tautly and widely retracted with sutures. This is a crucial maneuver because full lateral exposure of the intradural span, made possible by the generous bone removal, reveals the “crotch” where the far lateral fringe of the lipoma attaches to the inner surface of the dura (see Fig. 60-6A).

Next, the fusion line is identified where pia, spinal cord, and lipoma join in a continuous furrowed border that travels from rostral to caudal outlining the entire attachment of the lipoma stalk to the cord. In a dorsal lipoma, the fusion line forms a neat complete oval or circle from side to side, usually upon a leveled horizontal plane, often bilaterally symmetrical, and always sparing the conus below (Fig. 60-2B and eFig 60-10). In a transitional lipoma, the rostral fusion line starts distinctly enough but then edges ventrally towards the tip of the conus and tends to wander laterally and asymmetrically, often becomes sheltered by the overhanging fat, and never meets its mate from the other side at the caudal end (see Fig. 60-2C and D).

True to the events of embryogenesis, the DREZ and dorsal roots are always lateral to the fusion line, and at the rostral end of the fatty stalk of both lipoma types, this orderly arrangement can be depended upon on both sides, thus presenting a convenient place to start the dissection. In most transitional lipomas, however, the more caudal nerve roots are quickly covered from view by the overflowing fat which tends to fuse with the dura at a far lateral point (Fig. 60-5A). Hence the term “crotch dissection,” which depicts the key step of grasping the overhanging fat and pulling it medially under tension against the tagged dural edge, then sharply separating the fat-dura attachment with dissecting scissors (Fig. 60-5B and eFig. 60-14). It is absolutely requisite to lean the round curve of the scissors firmly against the inner lining of the dura while cutting this attachment to avoid blindly injuring the nerve roots, which project from the cord slightly medial to the “crotch” and lie just deep to the fat. The hidden roots should spring into view wherever the detached fat is pulled back, and can be gently coaxed away from the dura by blunt dissection toward the exit foramina (Fig. 60-6B). At the same time, the free CSF space ventral to the dorsal roots, and the fat-free, pia-covered ventral surface of the neural placode, hitherto hidden by the overhang, now “pop” into view (Fig. 60-6C).

FIGURE 60-5 Drawing depicting relationship among lipoma, neural placode, nerve roots, and dural sac in an axial slice. A, The lipoma–cord assembly is suspended at the dural edge at far lateral adhesion points like a hammock against side hinges. The dotted transverse line that joins the two side hinges divides the assembly into a dorsal disorderly fibro-fatty half that completely blocks the surgeon’s view to a much more orderly ventral half containing the important anatomic landmarks of fusion line, dorsal root entry zone, dorsal roots, fat-free ventral placode, and pristine, ventral cerebrospinal fluid space. B, After detaching the far lateral adhesion points (the hinges) by careful “crotch dissection,” and folding-in the fatty mass, the ventral anatomic landmarks can now be visualized.

FIGURE 60-6 Transitional lipoma. A, Exposure of the rostral lipoma, showing the normal cord to the right and the swath of yellow fat indicating beginning of the lipoma (small arrow). The large arrows locate the far lateral adhesion points (the “crotches”) between the flanges of the lipoma and the inner dura. The rostral starting point of the fusion line is also indicated (open arrow). B, Exposure of the pristine ventral CSF space (arrow), nerve roots, and fat-free ventral side of the neural placode. C, Exposure of all relevant anatomy vital to the next stage of the actual resection of the lipoma: the fusion line, dorsal root entry zone, dorsal roots, ventral side of placode and free CSF space, in dorsal-ventral order. D, Complete bilateral detachment of the adherent hinges and terminal untethering; the neural placode now “sinks” into the basin of the dural trough. CSF, cerebrospinal fluid.

It is clear from this description that in each successive axial slice, all lipomas big or small, dorsal or transitional, are roughly divided by a transverse line joining the points of the far lateral fat-dura attachment on each side, where the lipoma–cord assembly is in effect suspended like a hammock against two lateral hinges over an uncluttered ventral CSF pool. Dorsal to this transverse line is the visible but disorderly, massive, and unrevealing fat and ventral to this line is the orderly fusion line, DREZ, dorsal nerve roots, neural placode, and ventral CSF space but all initially rendered invisible to the surgeon by the overhanging fat (Fig. 60-5A). The purpose of the “crotch dissection” is therefore to release this suspension so that the hammock of neural placode and nerve roots can be folded inward enough to be identified and preserved during the next phase of lipoma resection (Fig. 60-5B).

This laborious but indispensable step of “crotch dissection” is carried all the way caudally (Fig. 60-6C) until all the “useful” nerve roots are identified and the entire neural placode, with a profusion of lipoma still attached, is completely unsuspended from the dura and has literally “fallen” to the basin of the dural trough (Fig. 60-6D).

Step 3. Lipoma Resection

Resection of the lipoma begins at the rostral end where the anatomic relationships between fat, DREZ, and nerve roots are clearly decipherable (Fig. 60-7 and eFig. 60-15). Sharp dissection with microscissors is used to locate a thin but distinct silvery white plane between fat and cord at the demi-lune of the rostral fusion line (Fig. 60-8A). It takes some determination, for the initiate, to cut into this traditionally forbidden place, seemingly straight into the spinal cord right at the fusion line, but with experience, this white plane can be found in every case. We strongly discourage using the CO₂ laser because it chars the surface (no more white plane!) and negates the tactile feedback through the microscissors on which the surgeon depends to differentiate between cutting through the grittiness of fibrous fat and the formless softness of spinal cord. Bleeding on the white plane can be handled with the ultra-fine irrigating bipolar cautery (0.2-mm tips) and a very low current setting. The cold irrigation mitigates against sticking but more importantly it dissipates heat rapidly from the cord. Minimal cauterization is used on the DREZ to avoid postoperative dysesthesia.

FIGURE 60-7 Basic technique for transitional lipoma resection: resection on an asymmetrical and oblique fusion line along an occasionally undulating lipoma–cord interface plane, showing pre- and post-operative pictures.

FIGURE 60-8 A, Transitional lipoma. The “white plane” of thin glistening fibrous netting separating fat from spinal cord, bounded always by the fusion line. Note detached rostral portion of the lipoma lifted away from the white plane to show the resection edge. B, Chaotic lipoma. Note ventral pia-covered fat medial to ventral nerve roots (being stimulated by concentric microprobe), and dorsal fat perched on the dorsal side of the placode. C, Transitional lipoma. Completed neurulation. D, Completed expansile graft duraplasty with bovine pericardium.

As long as all the activities are rendered medial to the fusion line, thus also medial to the DREZ and nerve roots, dissection along the white plane can be conducted safely all the way to the end with no damage to the cord or nerve roots (eFigs. 60-16 and 60-17). In a dorsal lipoma, this is a simple feat because the white plane is basically horizontal and flat, the two banks are symmetrical, and the caudal end well defined rostral to the conus so that a completely circumscribed attack on the fat is possible from multiple angles (eFig. 60-18). In a large transitional lipoma, navigating the white plane is more difficult because it always slopes ventrally, often undulates, and one side may be tilted so steeply that the corresponding DREZ and nerve roots are shifted ventrally and the placode so rotated that its ventral surface now faces the side. Such a white plane is almost turned vertically “on edge,” its orientation confusing unless one remembers the transverse line concept dividing one “clean” ventral hemisphere from the “messy” dorsal one.

The white plane sometimes seems never-ending in large transitional lipomas and the caudal thecal sac is thronged with fat admixed with suspicious strands. This is when systematic stimulation and identification of the ventral nerve roots become invaluable in localizing the termination of the functional spinal cord. As soon as two to three pairs of sphincter-activating sacral roots are identified, any tissue distal to the last pair can be considered nonessential and be cleanly cut across to consummate the final liberation of the placode (eFig. 60-19). A good chunk of the now isolated distal fatty stump should be excised to prevent reconnection with the terminal placode.

In chaotic lipomas, electrophysiologic determination of the functional extent of the neural placode may be the only way to achieve final untethering; the caudal fat-cord-fibrous jumble can only be sorted out functionally and not anatomically by direct stimulation of the placode and the projecting nerve roots. The handling of the white plane on the dorsal side of a chaotic lipoma is the same as with the other lipoma types, but the billows of fat on the ventral side of the placode should be left alone and its smooth pial surface left unviolated (Fig. 60-8B and eFig. 60-20). It is always the dorsal and never the ventral part (unless iatrogenically invaded) of the lipoma that actually tethers the spinal cord.²²

Step 4. Neurulation of the Neural Placode

Total resection of the dorsal fat and thorough unhinging of the placode convert a bulky, transfixed lipoma–cord complex into a free-floating, thin, supple, purely neural plate (eFig. 60-17), eminently suitable for pia-to-pia, midline, dorsal closure with interrupted 8-0 nylon sutures without strain or strangulation to the neural tissue (eFig. 60-21). It is helpful to leave a narrow cuff of pia along the cut edges of the white plane to accommodate the sutures, which are tied with inverted knots. Neurulation thus transforms a broad, wafery, sticky sheet into a trim, sturdy, pia-covered tube bearing a single seam, evocative of the natural neurulation process (Fig. 60-8C).

Step 5. Expansile Graft Duraplasty

The argument for a graft dural closure comes from the belief that if the neurulated placode could slosh about in ample CSF within a capacious sac, the likelihood of reattachment to the dura would be diminished. Thus we prefer a slightly full-bodied yet texturally compatible (to infant dura) material such as bovine pericardium that can maintain its shape, to a filmy-soft graft such as autologous fascia lata that may swell and ebb with respiration and body movements and thus collapse on the cord. The bovine graft is carefully measured and shaped to prevent inward folds. Running prolene sutures are used to achieve watertight closure, confirmed with Valsalva maneuvers (Fig. 60-8D).

Terminal Lipoma

There is no controversy surrounding the technique of resecting terminal lipomas, and the surgical result is essentially that of filum surgery. The terminal lipoma inserts directly into the caudal extremity of the conus. The junction between lipoma and conus is usually readily identifiable by the yellow color of the fat, the sparse vasculature compared to the conus, and the lack of nerve roots exit. Roots may become adherent to the ventral surface of the lipoma. It is important to identify and preserve “hitch hiker” roots arising from the conus but become adherent to the pia over the lipoma (see Fig. 60-3B). The conus–lipoma junction is then cut across transversely.

Type I SCM

Planning the skin incision requires knowledge of the exact vertebral level of the median septum. A linear midline skin incision is made to span at least two laminar levels above and two below the septum. In a type I SCM, the bony septum is always extradural, being completely surrounded and excluded from CSF by the medial walls of the double dural tubes which form a complete sleeve for the bone in the sagittal midline. The septum itself is frequently fused with, and thus hidden under, the neural arches so that it is not immediately visible at exposure. Noting the adjacent bony anatomy on the preoperative scan such as a bifid or eccentric spinous process, exostoses, and abnormal fusion is helpful in localizing the level. Another useful hint is that the septum is often located where the spinal canal is widest, or where the neural arches and spinous processes are hypertrophic and fused with adjacent laminae.

The extent of the laminectomy should include at least one level rostral and one caudal to the septum-bearing laminae. The hypertrophic laminae are rongeured away piecemeal around the attachment of the septum until only a small island of lamina is left attached to the dorsal end of the septum (Fig. 60-11A, upper). This affords a circumferential view of the bone spur still within its dural sleeve so that its dural attachment can be safely dissected off the bone deep within the cleft. However, once the dorsal support of the septum (by the laminae) is eliminated, the septum is no longer held rigidly at both ends and could be pushed from side to side depending on how solidly it is anchored ventrally. Excessive lateral movement of the septum thus may injure the subjacent hemicords and must be avoided. A Woodson dental elevator with a thin, angulated, sharp edge is well suited to peel off the dura with minimal lateral “wedging” motions (Fig. 60-11A, middle).

FIGURE 60-11 A, Exposure of a type I bony septum. Upper: After the laminae rostral and caudal to the midline septum are resected, the septum-bearing laminae are removed carefully around the dorsal “stump” of the septum where it is attached to the ventral surface of the hypertrophic neural arches. Middle: Surgeon then performs subperiosteal separation of the median dural sleeve from the bony septum with a small dental elevator. Lower: Bony septum is then avulsed from its ventral attachment, which is often slender. B, Upper: Dural opening is made from rostral to caudal along the medial edge of the median dural sleeve (dashed line). Lower: After lateral retraction of the dura, the median dural sleeve is isolated except ventrally. Note the paramedian dorsal roots attaching the medial aspect of each hemicord to the dural sleeve. C, Upper: Dural sleeve is resected flush with the ventral dural surface, proceeding from the rostral “free part” to the caudal end. Note the thickened filum terminale. Lower: Ventral dural defect is left after complete resection of the dural sleeve flush with the vertebral body. Note the sectioning of the thickened filum after it has been cauterized. The conus is now completely untethered.

For most type I septa, their dorsal attachment is broad but their ventral junction with the vertebral body is usually flimsy; they could thus be easily avulsed from the dural cleft (Fig. 60-11A, lower). If the ventral attachment happens to be stout and bony, the septum may be carefully chipped away using a small rongeur. Complete extradural removal of the bone spur at this stage greatly facilitates later resection of the dural sleeve. Also, most type I septa are either purely bony or partly cartilaginous. Purely cartilaginous septa are rare. Embedded in the septum is a fairly constant central artery which can bleed briskly on avulsion. A deft plunge with some bone wax on a cotton patty deep into the cleft should handle the bleeding.

The dura is opened on both sides of the now-empty dural cleft to isolate the median dural sleeve (Fig. 60-11B, upper). The medial aspect of each hemicord is often tightly adherent to the dural sleeve by fibrous bands that must be cut. Paramedian dorsal nerve roots, when present in a type I lesion, typically stretch from the dorsomedial aspect of the hemicords to end blindly within the median dural sleeve (Fig. 60-11B, lower). These are nonfunctional and must be cut prior to resection of the dural sleeve. The dural sleeve is always wedged against the caudal reunion site of the hemicords, and the split cord rostral to the septum is therefore a safe, relatively spacious area to begin resection of the dural sleeve. Proceeding caudally from the rostral edge of the sleeve where the hemicords are least adherent, the surgeon cauterizes the ventral attachment of the sleeve to seal the central vessels and then cuts it flush with the ventral dural wall (Fig. 60-11C, upper). Not having the enclosed stump of the bone spur greatly simplifies the low resection of the sleeve. The most hazardous part of this undertaking is at the caudal end where the reunited “crotch” of the hemicords hugs tightly against the caudal end of the sleeve. The pressure on the crotch is readily felt here, and upward migration of the cord is often seen right after the sleeve resection.

Complete resection of the dural sleeve exposes the ventral extradural space in the sagittal midline (Fig. 60-11C, lower). Any remaining bony stump must now be trimmed down until it is no longer in contact with the ventral surface of the hemicords (Fig. 60-12). Closure of the anterior dural defect is unnecessary because of the abundant adhesions of the ventral dura to the posterior longitudinal ligament which would naturally prevent CSF leakage. Anterior dural closure is actually undesirable because the anterior suture line can potentially tether the hemicords.⁸⁴ Posterior dural closure ultimately converts the double dural tubes into a single sac.

FIGURE 60-12 Resection of type I split cord malformation septum. A, Bone septum is isolated after laminectomy to expose the double dural sleeve. B, Bone septum has been removed and dura opened, exposing median dural sleeve and hemicords. C, Median dural sleeve is being pulled caudally to expose the rostral free space between the hemicords. D, Free hemicords and reunited cord after complete resection of median dural sleeve.

Frequently, a fibroneurovascular stalk containing paramedian dorsal nerve roots, fibrous bands, and large blood vessels (myelomeningocele manqué) tautly tethers the hemicords to the dorsal dura. These bands always penetrate the dura at a level more caudal than their origin from the hemicords, and they often form an exuberant tuft of fibroadipose tissue which, unlike ordinary loose extradural fat, clings tenaciously to the outer surface of the dura, and marks the underlying myelomengocoele manqué. These bands must be cut flush with the hemicords to complete the untethering process.

In 5% to 10% of type I SCMs, the bony septum is oblique. It arises from the midline posterior surface of the vertebral body but then cuts diagonally across the spinal canal to divide it into two asymmetrical compartments. Without exception, the hemicord contained within the larger compartment (major hemicord) is much larger than the hemicord in the smaller compartment (minor hemicord), sometimes by a factor of 2 or 3. Moreover, the larger hemicord frequently possesses one set of lateral ventral roots but two sets of dorsal roots, whereas the smaller hemicord only gives off a single set of ventral roots. In these unusual cases, the exposure of the minor hemicord is hampered because it is partly sheltered by the overhanging oblique bone spur as well as being ventrally rotated away from the surgeon’s view. Moreover, the smaller hemicord is extremely delicate. It can thus be injured inadvertently during the removal of the bone spur. This unusual pattern of asymmetric splitting must be recognized through preoperative scan as a signature of heightened risk.

Type II SCM

In all cases of type II SCM, some form of fibrous (mesenchymal) septum or band is found within the midline cleft. Three patterns of such nonrigid median septa are encountered in type II lesions: (1) The least common is a complete fibrous septum stretching between the ventral and dorsal surfaces of the dural sac. Except for it being nonosseous, the complete fibrous septum transfixes the hemicords to the surrounding dura in the same manner as does the type I bony septum. (2) Slightly more common is the purely ventral fibrous septum. Its intimate adherence to the ventromedial aspects of the hemicords in effect anchors the cord ventrally where the septum arises from the ventral dura. (3) By far the most prevalent kind is the purely dorsal septum that attaches the dorsomedial aspects of the hemicords to the dorsal dura. A tuft of fibrovascular tissue in the extradural space is sometimes found connected to the septum through a small defect in the dorsal dura (Fig. 60-13B).

FIGURE 60-13 Resection of type II split cord malformation septum. A, Computed tomography myelogram shows dorsal septum. B, Drawing of the median septum and MMQ. C, Fibrous septum and a large blood vessel, and an MMQ. D, Muscles within fibrous septum. MMQ, myelomeningocele manqué.

Hypertrophic and fused laminae, common in type I lesions, are seldom found in type II SCMs. In fact, the neural arches of type II lesions are often attenuated or even bifid. Laminectomy for these SCMs is technically easy and safe. A midline dural opening immediately exposes a dorsal or complete septum (Fig. 60-13) but a purely ventral septum has to be sought for either between the hemicords or by gently rotating the hemicords to one side (Fig. 60-14). Like the type I bony septa, all type II fibrous septa are found near the caudal end of the split. However, the length of the split in type II SCMs is, by comparison, much shorter than that of the type I SCM, and, because all fibrous septa are thin, the hemicords in type II SCM are apposed much closer together, with very little “free” part. Intracleft exploration is not recommended. Fortunately, the excessive width of the spinal canal should allow the hemicords to be rolled gently to one side for ventral inspection.

FIGURE 60-14 Type II split cord malformation with ventral fibrous septum. A, Computed tomography myelogram shows VS. B, Drawing of VS with a skirting artery being cut. C, VS with skirt blood vessel. D, VS being cut. VS, ventral septum.

The point of attachment between hemicords and septum is invariably rostral to the point of attachment between dura and septum (Figs. 60-13 and 60-14). This is true of all three kinds of fibrous septa, giving testament to the fact that the fibrous septum was dragged upward by rostral movement of the cord after formation of the primordium of the septum. This upward dragging converts the incomplete (ventral or dorsal) septum into an oblique backward-pointing sheet and the complete septum into a V-shaped sheet with the apex pointing rostrally. Blood vessels often skirt the edge of the septum (Fig. 60-14C) or are loosely incorporated within the centromedian fibroneural bands. Very rarely, a true arteriovenous malformation is found within the median cleft. Unless these large vessels are contributing to the tethering of the hemicords, they should be left alone. Paramedian dorsal nerve roots and myelomeningocele manqués are commonly found in type II SCMs (Fig. 60-13), coursing dorsally from the dorsomedial aspect of the hemicords. The puny nerve roots usually end blindly in the septum but the more robust myelomeningocele manqués often penetrate the dorsal dura caudal to their hemicord origin (Fig. 60-13). These bands contribute to the tethering and must be cut. Untethering is completed by simply cauterizing the central vessels and excising the median fibrous septum. The ventral septum and bands come through a small defect in the ventral dura, presumably where the original endomesenchymal tract arose, but this small defect never leaks CSF and does not need to be repaired.

Composite SCMs and Multiple SCMs

A composite SCM consists of two or more SCMs of differing types occurring in tandem, with no normal cord in between. The most common constituents of a composite SCM are a type I–type II–type I combination (Fig. 60-15A). Each individual component is typical of its kind: the type I lesion has an extradural bone spur and a median dural sleeve, and the type II lesion has a median fibrous septum within a single dural tube. The three median elements are continuous, suggesting that the entire lesion results from a single (but very large) endomesenchymal tract in which meninx primitiva precursor cells have been included at both ends to cause the type I lesion, but not in the middle where the median septum remains fibrous. The total length of the septa can sometimes span as many as seven vertebral levels. The split cord is coextensive with the septa so that the rostral and caudal reunion sites of the hemicords hug tightly the rostral margin of the rostral bone spur and the caudal margin of the caudal bone spur, respectively.

FIGURE 60-15 A, Composite SCM. Right: CTM at T12 shows a type I SCM with double dural sacs and an oblique bone septum dividing the cord into a small right hemicord and a large left hemicord. Middle: CTM at L1 shows a type II SCM with a single dural sac and an oblique fibrous septum (small arrowheads). Left: CTM at L4 shows a second type I SCM with an oblique bone septum. All three septi are oblique in the same plane. B, Intraoperative drawings of a composite SCM after the two type I bony septa have been removed. Upper: Midline type II fibrous septum fills the Interval between the median dural sleeves of the two type I lesions. Lower: Fibrous septum has been resected, which provides room in the middle; the dural sleeves now can be resected safely from a middle free space toward the respective “crotches” of the split cord. CTM, computed tomography myelogram; SCM, split cord malformation.

The dural opening for composite SCMs begins rostral to the rostral dural sleeve, skirts around it, returns to the midline over the type II lesion, and finally skirts around the second dural sleeve (Fig. 60-15B, upper). Resection of the middle type II fibrous septum should then be done to gain a “free” area within the midportion of the median cleft. This will provide working room to begin resection of the type I dural sleeves above and below: for the rostral dural sleeve, resection goes from caudal to rostral, whereas resection goes from rostral to caudal for the caudal dural sleeve. This way, the manipulation of the dural sleeve is always directed toward and never away from the tight reunion site of the hemicords at either end (Fig. 60-15B, lower). Release of the cord is accomplished with total excision of all three septa (eFig. 60-24).

If two or more SCMs occur in the same patient but are separated by an interval of normal spinal cord, they are true multiple SCMs. These are rare because they result from multiple endomesenchymal tracts, that is, from multiple embryologic errors in the same neural tube. The individual SCMs may be all type I or type II, or of mixed types.²³ Their surgical treatment is as described.

Associated Dermal Sinus Tract and Dermoid Cyst

A dermal sinus tract is formed when the original connection between the endomesenchymal tract and the cutaneous ectoderm is retained (eFig. 60-25A). Because of this embryologic relationship, the deep end of the sinus tract is always in continuity with the mesenchymal median septum, but because the septum is either extra- or intrathecal depending on the SCM type, the clinical significance of the retained dermal sinus tract also varies with the SCM type.

For a type I lesion, the dermal sinus tract can be traced all the way from the skin pit through defects in the myofascial layers and neural arches to the bone spur (eFig. 60-25B). Thus, the tract is always extradural and usually does not contribute to the tethering. It is removed together with the bone spur (eFig. 60-25A). Even if the tract may occasionally encyst to form a dermoid cyst, it lies outside the dural sac and seldom becomes large enough to cause compression of the hemicords.

In a type II SCM, however, the dermal sinus tract is of necessity intradural where it retains connection with the median fibrous septum. In its intradural course, the sinus tract is often densely adherent to the hemicords, thereby exerting a separate tethering effect (eFig. 60-26). In addition, over 50% of dermal sinus tracts in a type II SCM will develop a dermoid cyst within the dura, often large enough to cause cord compression. The entire intradural sinus tract and cyst must be excised to prevent recurrence. The cyst is first collapsed by intracapsular evacuation of its cheesy content and the cyst wall is then carefully peeled off the pial surface of the cord. Its deep end is removed with resection of the fibrous median septum (Fig. 60-16).

FIGURE 60-16 Intraoperative pictures of type II split cord malformation and associated dermoid cyst showing dermal sinus tract continuous with the median fibrous septum and tract.

Associated Myelomeningocele and Hemimyelocele

Approximately 25% to 35% of SCMs have an associated ONTD.^20,22,23 Depending on whether one or both hemicords are involved in the dysraphic sac, the lesion may contain a hemimyelocele or a full-blown myelomeningocele, respectively.

In most myelomeningoceles, the open neural placode is terminal and thus is caudal to the SCM. Less commonly, the open placode is segmental and may therefore be rostral to the SCM. Because the open dysraphic sac is usually treated at birth, by the time the SCM is diagnosed by later neuroimaging studies and explored via a second surgical procedure, the original neural placode would have already developed dense adhesions to the dorsal dura. Because the placode is always adjacent to the split cord and may well be contributing to the tethering, it is always freed from the dura when the SCM is being treated. Depending on whether the placode is rostral or caudal to the SCM, it is carefully detached from the dorsal dura by sharp dissection either before or after the median septum is excised.

Most hemimyeloceles are segmental (eFig. 60-27). The unaffected hemicord is usually hidden from view by the median septum during the original sac closure, when the hemiplacode is mistaken to be the whole lesion. During the definitive procedure for the SCM, the entire dural sac (double or single) must be exposed to give access to both hemicords (Fig. 60-17). This often requires cutting into dense scar tissue. Removing the adjacent normal laminae helps to define the full width of the dura rostral to the scar. After identifying and resecting the median septum, the hemiplacode is detached from the dorsal dura with sharp dissection.

FIGURE 60-17 Intraoperative pictures of type I split cord malformation with associated hemimyelocele. A, 8-mm thick stalk of neural tissue arising from the right hemicord just rostral to the median dura-clad cartilaginous septum. B, Both hemimyelocele stalk and median septum have been removed.

Associated Neurenteric Cyst and Other Intestinal Anomalies

Since the ventral endomesenchymal tract is continuous with and partially composed of yolk sac endoderm, it is no surprise that enterogenous remnants and other intestinal anomalies coexist with SCMs. Neurenteric (enterogenous) cysts are rare (presumably because extrinsic induction of endodermal progenitor cells is fastidious) but have been found along the entire endomesenchymal tract, from subcutaneous layers to the prevertebral space in front of the split cord⁸⁴ (eFig. 60-28). These cysts can cause compression when found in between the hemicords (Fig. 60-18) or as intramedullary masses within the hemicords (Fig. 60-19). They must be resected in toto to avoid recurrence.

FIGURE 60-18 Intracleft neurenteric cyst in a type I split cord malformation. A, Sagittal magnetic resonance image shows the median septum. B, Computed tomography myelogram shows the cyst within the median cleft. C, Shows bony median septum (MS) within median dural sleeve. D, Shows neurenteric cyst (Neu cyst) rostral to the dura-clad bony median septum (MS) within the median cleft. Hc, hemicord.

FIGURE 60-19 Intracleft neurenteric cyst and intramedullary neurenteric cyst in a patient with type II split cord malformation, who also has a lipoma in the left Hc. A, magnetic resonance imaging shows Hc’s and lipoma in the L Hc. B, Shows intracleft neurenteric cyst (Neu cyst) after the R Hc is lifted up. C, Shows intramedullary neurenteric cyst with L Hc, adjacent to a dermoid cyst. Hc, hemicord; L Hc, left hemicord; R Hc, right hemicord.

Intestinal malrotation results when the ventral endomesenchymal tract remains as a tethering band in the proximal intestine (Ladd’s band) that prevents normal clockwise rotation of the primitive fore-gut (eFig. 60-29). If the portion of the endomesenchymal tract continuous with the yolk sac differentiates into mature bowel epithelium, a patent intestinal diverticulum linking small bowel with the prevertebral tissue may occur as an impressive mass lesion in the thoracic cavity (eFig. 60-30). This will require a combined thoracotomy and spinal exposure for complete resection.

Neurologic Injury

Worsening of neurologic function occurs in less than 3% of patients following surgery for SCM. The surgical morbidity is higher with a type I SCM due to hemicord injury during removal of the bone spur. This is especially true when the bony septum is oblique and the delicate minor hemicord is tucked under the overhanging septum. The risk of hemicord injury is minimized by the following precautions:

Cerebrospinal Fluid Leakage

Wound complications are seldom encountered in patients with SCM except when there is a previously treated open neural tube defect adjacent to the split cord. These children often have poor myofascial coverage as well as tenuous, scar-ridden skin over the dural sac and may develop CSF leakage. This underscores the importance of a watertight dural closure. In addition, these children should be managed in the fully prone position for 5 to 7 days after surgery similar to the newborn with a newly repaired myelomeningocele. Heavy sedation may be necessary to prevent an infant from wiggling and pulling at the incision.

Postoperative Bladder Management

Transient detrusor weakness and urinary retention is not uncommon after the resection of a type I SCM. It can last from a few days to several months and tends to be more common in adults than in children. Permanent worsening in bladder function occurs in about 3% of patients.^20,23,24 A bladder catheter is left in the first two postoperative days when the patient is confined to complete bed rest. The catheter is removed on the third postoperative day as the patient is encouraged to ambulate and use bathroom facilities. If there is obvious difficulty in micturition or if a large postvoid residual urine volume is obtained repeatedly, the patient is given a cholinergic agent such as bethanechol chloride to strengthen detrusor contraction, or an α-sympathetic blocker such as prazosin hydrochloride to encourage relaxation of the internal urethral sphincter. If the postvoid residual urine volume remains high on medication, an intermittent catheterization program is started in the hospital and is continued on an outpatient basis. After 4 to 6 weeks, a cystometrogram is obtained to evaluate bladder function and to determine whether therapy can be discontinued.