Detailed reviews by Harrison et al [35–37] and Rhee et al  suggest that preserving a normal sagittal spinal contour may be important for long-term health. De Jonge et al  described how correction of lateral scoliotic curvatures caused a spontaneous restoration of the sagittal spinal curves, suggesting that loss of sagittal spinal curves may somehow be related to scoliotic curvatures.
Scoliosis places otherwise symmetrical muscle groups under longstanding, isometric, asymmetrical loads [38–41], which may compromise circulation within the muscle, ultimately leading to myofascial trigger points and chronic inflammation . Weinstein et al  reported that scoliosis patients may retain high levels of function in later life, but do report higher instances of chronic back pain.
In addition to higher instances of chronic back pain, significant psychological issues may arise from concern over cosmesis and conventional treatment. Freidel et al  measured the self-perceived quality of life in women with scoliosis using the SF-36 questionnaire. They concluded that the psychosocial aspects of scoliosis and scoliosis treatment should be addressed in the treatment of this group of patients. Similarly, Sapountzi-Krepia et al  described the psychological distress that adolescents encounter while going through bracing treatment for scoliosis. A case-control study by Danielsson et al  identified a potential negative impact on the ability to function sexually due to conventional treatment restraint or self-consciousness of physical appearance.
Aside from back pain and psychological disturbance, several studies also suggest that scoliosis affects more than the musculoskeletal system. Curvatures of the thoracic spine are associated with restrictive lung disease due to ribcage deformity and decreased chest wall compliance . Chest wall compliance is inversely proportional to the magnitude of the Cobb angle down to 10°, and vital capacity is reduced by decreased chest wall compliance directly [46, 47]. Exercise endurance is also inversely diminished with increasing Cobb angle, even in patients with normal resting vital capacity . Thoracic scoliosis may also cause shortness of breath and recurrent respiratory infections [46, 49]. Indeed, scoliosis affects more than the musculoskeletal system.
Concerning coronal Cobb angle measurement for scoliosis, manual radiographic measurement has consistently shown good to excellent inter- and intra-observer reliability [50–53]. Previous studies demonstrate a manual Cobb angle measurement error on full-spine radiographs of 2.5 – 4.5° [51–53]. However, to achieve this low error, it is imperative that the same end vertebrae, same protractor, and same endplates are consistently chosen. Importantly, these measurement errors were extracted from data collected on full-spine radiographs. Patient positioning can significantly negatively impact measurements on full-spine radiographs . The Cobb angle measurements in our study were taken from sectional radiographs, which reduce the positional distortion caused by inconsistent patient positioning. It is unknown to what extent the use of sectional radiography has on Cobb angle measurement error, if any.
The treatments outlined here required home care exercises, as described earlier. However, these exercises, which take up a combined 60 minutes per day, can be done in private, away from scrutiny by peers, neighbors, or relatives. This is in contrast to bracing treatment, where the brace must be worn at least 18 hours per day to achieve a good clinical result .
We placed the headweight, shoulderweight, and hipweights in areas designed to reduce our patient's specific spinal distortion patterns on radiograph. The patient was evaluated radiographically while wearing the headweight and shoulderweights to determine optimal position and weight. Our repeated clinical observation has demonstrated that patients may visually appear to improve with a shoulderweight in a certain position. However, they can look dramatically different on radiograph (migration away from the vertical axis) than they appear in visual posture analysis. This is consistent with recent failed attempts to objectify visual posture analysis as a valid clinical tool . It is prudent to develop alternative methods of evaluation to avoid unnecessary radiation exposure to patients.
Because of the anterior wedging from T7–T10 in case #2, it is not surprising that over time a thoracic hyperkyphosis and swayback developed in this patient. As a result, marked anterior weight bearing of the head was required to maintain a horizontal eye level, thus satisfying the postural reflexes [56–61] Additionally, the marked forward head posture elicits the pelvo-ocular reflex, causing a forward shift of the pelvic girdle under the forward head position . Therefore, the postural distortions seen in this case may represent compensatory changes over time as a result of thoracic buckling, a posture known to commonly cause increased mechanical stress at the spinal transition areas [4, 24]. Correcting these compensatory postural changes proved to be a challenge, given that the impetus for them (the anteriorly wedged thoracic vertebra) could not be immediately, if ever, changed. However, within the confines of the Hueter-Volkmann law, we postulate that sustained correction of the asymmetrical mechanical spinal loading may theoretically help these vertebrae to remodel to some degree. Although the forward head posture is a compensatory reaction to the hyperkyphosis, the cervical spine soft tissue has likely remodeled to the forward head posture, given the likely duration of its existence . Therefore, we felt that direct correction of the forward head posture must also be achieved to improve overall sagittal alignment, given the neurological control and importance of head position on upright spinal position . This hypothesis remains to be definitively evaluated.
The significance of cases #2 and #3 lies in the location of the scoliotic curvatures. In the vast majority of cases, double major curvatures usually maintain a right thoracic/left lumbar pattern. In this case, the pattern was reversed, showing a left thoracic/right lumbar scoliosis. Several authors have previously discussed the unique presence of a left thoracic – right lumbar curvature pattern. McCarver et al  showed that only 1% of 550 patients with idiopathic scoliosis had double major curvatures consisting of a left thoracic – right lumbar configuration. Winter and Lonstein  maintained that any left thoracic curvature should be further evaluated for neurological abnormalities, such as neurofibromatosis, spina bifida, or syringomyelia. Finally, Schwend et al  also concluded that additional testing was necessary in left thoracic curvatures, given an observed higher incidence of neurological clinical signs. Case #3 seems to correlate these findings given the left thoracic scoliosis secondary to Scheuermann's Disease. It is important to note, however, that treating the Scheuermann's Disease itself was not our aim. Rather, our goal was to reduce the thoracic scoliosis secondary to it. We are not attempting to show that this treatment may affect the Scheuermann's Disease. In this case, however, additional testing was conducted at the initial time of discovery of the scoliosis. Further, my initial neurological examination also failed to produce any remarkable neurological findings.
Recently, several authors have discussed the relationship between the sagittal spinal contour and scoliosis [14, 15, 67, 68]. Harrison et al [35–37] have discussed the pathophysiologic changes associated with the loss of the sagittal curves. Based on this evidence, we decided that it was important to the long-term outcome to address these spinal parameters.
Cases #1 and #2 present what appears to be inconsistent findings. Case #1 initially had a 23° cervical lordosis, below asymptomatic 31–40° range identified by McAviney et al , and the normal 34° identified by Harrison et al . However, case #2 displayed a 32° initial cervical lordosis despite having a thoracic hyperkyphosis. In case #1, the patient had 31 mm of forward head posture. Since forward head posture reduces the magnitude of the cervical lordosis [69, 70], a 23° cervical lordosis may not be normal for this patient. Additionally, recent evidence suggests that sagittal balance may more closely correlate to symptoms than sagittal alignment  Cervical lordosis by itself may not provide an accurate assessment of normal for each patient. Therefore, we suggest that both the cervical lordosis and forward head posture be weighed before a patient's cervical spine may be considered "normal." In contrast, case #2 had a both a normal cervical lordosis and forward head posture (32° and 22 mm, respectively). Therefore, we classified this patient's cervical spine as normal, despite the thoracic hyperkyphosis. We feel that the 55° lumbar hyperlordosis is a direct compensation for the swayback posture created by the thoracolumbar vertebral remodeling. This is consistent with the post treatment reductions in the swayback posture and lumbar lordosis.
In the Pettibon system, most of the manipulative treatment is not administered on a vertebral segmental basis. Rather, it is delivered to a specific region of segments so that the entire region may be mobilized. The goal of manipulative therapy in the Pettibon system is to mobilize several vertebral joints so that the rehab procedures can target the joints while they temporarily have an increased range of motion .
The purpose of the Pettibon Weighting System™ is to artificially alter the centers of mass of the head, trunk, and pelvis, causing reactive corrections by the postural reflexes [72–74]. The goal of postural reflexes is to maintain efficient body stance and locomotion using the least energy expenditure possible [56, 63, 75]. In the present cases, each patient was instructed to continue with their home exercise routine on a once weekly basis in attempts to maintain the change in spinal configuration.
The procedures that comprise the Pettibon system have been previously examined in specific clinical cases [5, 76]. Although these techniques have been investigated for preliminary treatment of idiopathic scoliosis , they have not, until this point, been used in cases of scoliosis due to structural deformity or left thoracic primary curvatures. Given the perceived results of the cases outlined here, it is worthy of future investigations in such cases. However, case reports and case series designs do not provide substantive evidence of therapeutic effectiveness. This remains the realm of properly conducted prospective clinical trials.
Conservative treatment for scoliosis needs to be examined much more closely in the biomedical literature, as side effects [44–46] and compliance issues  make conventional treatments such as bracing less attractive to patients and parents of minor patients.