The cephalad attachments of VEPTR devices are circumferential rib “cradles,” while the caudal attachments may be ribs, spine, or pelvis [10]. Non-VEPTR rib-based techniques generally use upward-directed distraction spine hooks as rib anchors beneath the caudal surface of the ribs. Acute rib anchor failure is usually rib cutout or fracture and associated with excessive stress or poor quality of rib bone such as bone dysplasia or nutritional osteopenia [32] (Fig. 50.6). Preoperative assessment of bone density and treatment of osteopenia or osteoporosis will minimize this complication. Rarely acute dislocation of the entire rib at the costovertebral articulation occurs if excessive distraction force is applied to the rib. Acute loss of rib fixation is best avoided by distributing distraction force to multiple anchor points (“load sharing”), either by encircling more than one rib (VEPTR I) or staggered multiple anchors (VEPTR II) [33] (Fig. 50.7) or similar configurations with multiple spine hooks used on the ribs. Where appropriate, an expansion thoracostomy [28] can diminish the necessary distraction force needed to achieve control of combined chest and spine deformity. Kyphotic deformities corrected by cantilever forces applied via the upper rib attachment may fail acutely if too much deformity is corrected. Chronic migration or “drift” of rib anchors is a common phenomenon, particularly in rib-to-spine more than rib-to-rib distraction constructs [10, 28–31, 34, 35]. Campbell [31] reported drift of rib attachments in 7 of 27 patients. Usually rib anchor drift is not functionally significant, as the drifting rib attachment gradually pulls with it a solid bone attachment and remains functionally connected to ribs, but some lose functional connection and require revision, which can usually be done at the time of a planned lengthening. Distal migration of VEPTR iliac S-hooks is common over time, particularly in ambulatory patients or unilateral devices [36, 37]. Distal drift of an S-hook can be troubling, as the S-hook becomes buried in the ilium, gradually drifting toward the acetabulum. Earlier revision of drifting S-hooks should be considered to avoid the need for extensive bone removal to access buried S-hooks.

Injury to the brachial plexus is unique to VEPTR and rib-based distraction techniques which commonly use the uppermost chest wall for attachment points [30, 38, 39]. Concomitant congenital rib and shoulder anomalies may contribute to the occurrence of injury, but two definite etiologies for brachial plexus injury in primary VEPTR or rib-based surgery are recognized. The brachial plexus can be injured directly by an implant placed too cephalad and laterally in the uppermost thorax. Campbell has described the boundaries for safe upper rib cradle placement, suggesting devices should remain medial to the scalene muscles and never cephalad to the second rib [10]. Probably the most common etiology for brachial plexus injury is compression of the plexus between the acutely cephalad-displaced upper chest wall and the clavicle or upper humerus at the time of the initial distraction and expansion procedure. Nassr et al. [38] has validated this explanation experimentally. Brachial plexus palsy may be delayed, as the compression gradually takes effect and postoperative swelling occurs. Awareness of the possibility of brachial plexus palsy and attention to upper extremity motor and sensory monitoring intraoperatively can provide early warning [39]. If extensive displacement of the thorax is planned or the chest wall soft tissue covering is stiff, preliminary clavicular osteotomy, as done for correction of Sprengel’s deformity, and preliminary implantation of a tissue expander may help avoid brachial plexus compression. VEPTR when used as a pure spine distraction device, such as in the minimally incisional technique described by Smith [34, 35], is not associated with brachial plexus injury, as there is much less acute chest expansion.

Rib-based attachments such as VEPTR or chronic contact between the chest wall and either rib-based or spine-based rods can produce local chest wall scarring and fusions between otherwise normal ribs. This phenomenon is readily seen clinically at the time of revision or on CT and has been documented [40, 41]. The clinical significance of chest wall scarring and fusions between ribs in the EOS patient is not clear. If the preoperative condition included a congenitally stiff, small thorax such as that seen with congenital rib fusions, spondylocostal or spondylothoracic dysplasia, or some myopathies, then the stiffness created by rib-based devices, or spine-based devices contacting the ribs, is likely not significant, as the result of treatment remains the creation of a larger albeit still stiff thorax. However, if the chest wall was mobile preoperatively and there were not extensive congenital rib fusions, then the scarring, rib fusions, and stiffness associated with treatment may be relatively detrimental to thoracic function when compared to techniques that do not directly affect the chest wall. The leather-like scarring on the chest wall beneath spine- or rib-based devices is readily apparent to the surgeon at the time of device exchange. Re-fusion of previously separated congenitally fused ribs can occur and may be a cause of deformity progression and inability to continue with lengthening. Repeat osteotomy of rib fusions can lead to improvement in deformity and continued lengthening [41]. Rib re-fusions are typically medial in location or beneath VEPTR devices. Excision of the bony bridge and release of the adjacent chest wall scar usually permits resumption of device lengthening (Fig. 50.8). If VEPTR lengthening becomes increasingly difficult, a search for rib fusion by CT is appropriate.

Scapulothoracic stiffness, subscapular bursa formation, and spontaneous fusion of the scapula to the VEPTR device and ribs can occur. The location of upper rib-based anchors beneath the scapula stimulates bursa formation and may contribute to shoulder stiffness. If the original procedure included a thoracostomy, the incision of the scapular stabilizing muscles may contribute to scapulothoracic stiffness or dysfunction. Repetitive incisions for lengthening in the area also contribute to scapulothoracic scarring. Attention to surgical technique, early encouragement of range-of-motion exercises, and placement of lengthening incisions away from the scapula may help preserve scapulothoracic function in rib-based devices. If scapulothoracic motion is limited, a CT may reveal bridging bone between scapula and ribs at the location of the subscapular rib-based attachment (Fig. 50.9). At the time of device exchange or as a separate procedure, it is possible to mobilize the scapula from the underlying thorax, freeing scar and adhesions, and excising bony bridges. Postoperative physical therapy is needed to retain mobility.

Wound problems may limit the duration and success of rib-based distraction devices. Integrity of the wound is particularly important for the initial procedure as well as multiple subsequent lengthening procedures. Wound dehiscence or superficial wound infection may lead to deep infection involving the implant, a difficult problem at best with implant removal sometimes needed (Fig. 50.10). Experience with implant infection in VEPTR and rib-based devices is well documented by Campbell, Smith, and others [42, 43]. Preoperative nutrition, soft tissue health, and soft tissue handling are important deterrents to perioperative infection. Predisposing factors to wound problems include poor nutrition, prior infections, or prior incisions in the area such as in myelodysplasia and many neuromuscular deformities, where there may be insensate skin or an uncooperative patient. Nutrition should be assessed and must be optimal preoperatively, even to the point of establishing enteral feeding to encourage weight gain. Preoperative planning of skin flaps and use of tissue expanders can help avoid leaving surgical incisions directly over prominent devices. For expansion thoracotomies, the author prefers to create flaps in which the muscle layer is longer than the overlying skin, making dehiscence less likely (Fig. 50.11). Excessive tension on the wound must be avoided. Prominent devices need to be protected from pressure in the post-op period, and a donut-like padding is incorporated into the post-op dressing.