Whether to perform posterior decompression for all OPLL patients was yet to be controversial since the concept of K-line in OPLL was raised. Fujiyoshi et al.[4] analyzed the posterior shift of the spinal cord by intraoperative ultrasound immediately after decompression procedure and reported that negative K-line OPLL patients’ postoperative cord shift distance was insufficient for there were no subarachnoid space at the ventral side of the spinal cord, which indicated a poor prognosis. Furthermore, Epstein et al.[2] proposed that when CSM accompanied by cervical kyphosis, the surgical outcomes of posterior approach were limited for cervical lordosis is crucial for sufficient decompression. Hence regional kyphotic CSM and OPLL with negative K-line were both general accepted as contraindications of single-staged posterior decompression[8].
However, posterior cervical laminectomy is reserved for multilevel cervical stenosis and OPLL given to its direct enlargement of spinal canal to allowing direct or indirect decompression of the spinal cord with less additive morbidity, avoiding catastrophic complications of anterior approaches including retropharyngeal hematoma, main vessel injury and esophageal injury. With adequate fusion, laminectomy permitted dorsal migration of the cord from compressive origin and thus improved the vascular perfusion and neurological recovery[12, 13]. And posterior approach had additional potential merits in direct visualization, surgical duration and intraoperative blood loss compared to combined approaches. In present study, the operative time of PLF were 171.36 ± 39.69min for KSM group and 184.25 ± 85.26min for NKO group, as the intraoperative blood loss were 179.09 ± 134.20 ml for KSM group and 272.50 ± 265.80 ml for NKO group. For the risk of postoperative instability for CSM patients, we didn’t consider to perform laminoplasty which was reported had lower JOA recovery rate than laminectomy by previous studies[6, 14, 15].
Myelopathy in cervical segment is partially resulted from static risk factors like stenosis or tethering by kyphotic deformity[9], which cause direct mechanical compressional neurocyte death and ischemia[16]. The canal-occupying ratio was even higher in OPLL with negative K-line, thus the mechanical cord compression could be much severer despite the rather normal posterior structures. The negative K-line reflected the substance that such OPLL was a type of intraspinal kyphotic lesion in sagittal plane, in regard to the original definition of K-line, which the “K” represents “kyphosis”[4]. Zhang et al.[17] reported that by correcting the cervical kyphosis into lordosis in patients with negative K-line OPLL via posterior screw-rod instrumentation, the K-line turned from negative to positive, and the canal-occupying ratio decreased from 63–33% on average with the JOA score recovered from perioperatively 8.4 to 13.3 at follow-up. A recent study by Du et al.[18] emphasized the role of cervical lordosis in posterior decompression outcomes for multilevel CSM associated with kyphosis, indicating a similar effect of stenosis and cord tethering to negative K-line OPLL.
In spite of the correlation between spinal cord shape and the development of cervical myelopathy[19], the difficulties existed in how to quantitatively describe the pathophysiological effect. Mihara et al.[20] reported postoperative ultrasound evaluated the subarachnoid space and its correlation with better neurological recovery. To our knowledge the sufficient spinal cord shift after posterior decompression was reported to be associate with good clinical outcomes[21]. Our present study evaluated the perioperative C3-C7 column-cord distance by modifying the method by Shiozaki et al[12], which was 7.025 ± 3.057 mm in NKO group and 9.150 ± 2.312 mm in KSM group, reflecting the similar perioperative cervical spinal cord condition among two groups(p = 0.244). After posterior laminectomy and fusion, the C3-C7 column-cord distance at last follow-up improved to 16.625 ± 4.968 mm in NKO group (p = 0.047) and 21.317 ± 5.792 mm in KSM group (p = 0.004). The cord shift might be contributed to the compensatory space provided by PLF, and the alleviation of kyphosis by neutral position fusion.
Since the radiographic evidence is not completely corelated with the severity of myelopathy[22], we also evaluated the JOA score and its recovery rate among two groups. After PLF, The JOA score change of both groups was larger than 2.5 points regarding to the minimum clinically important difference of JOA by So Kato et al[1], showing favorable neurological recovery. For NKO group, the mean JOA recovery rate is 50.76%, which illustrated sufficiency of PLF for both regional kyphotic CSM and OPLL with negative K-line. Previous studies also investigated the outcomes of posterior decompression in patients with large intraspinal occupying lesion as OPLL or regional kyphotic CSM. Saito et al.[5] reported posterior double-door laminoplasty for OPLL with negative K-line showed favorable results, in which JOA score improved from 8.0 to 11.9 postoperatively with average recovery of JOA score was 43.6%. Furuya et al.[23] investigated 5-year outcomes of posterior decompression and fusion (PDF) for OPLL with negative K-line, and the JOA score recovered from 7.9 ± 2.4 to 11.7 ± 2.6 with a recovery rate of 39.0%, while this study emphasized the postoperative C2-C7 angle kyphosis of 2.7° on average. Katsumi et al.[24] also retrospectively analyzed PDF for negative K-line OPLL with a JOA recovery rate of 53.3%, and they reported a postoperative lordotic cervical alignment at least 2° correlated with an over 50% JOA recovery rate. Du et al.[18] investigated PLF for multilevel cervical degenerative myelopathy associated with kyphosis, showing significant differences before and after surgery in the JOA score and spinal cord shift.
The main concerning for our PLF procedure were over shift of the spinal cord which was reported to be a risk factor for C5 root palsy[25]. In our present study, the incidence of C5 root palsy were 10.5% in KNO group and 8.3% in KSM group at final follow-up. Postoperative axial neck pain was most common in the cervical kyphosis patient, in present study 2 patient (16.7%) in KSM group developed this complication, showing significant difference with NKO group. We recognized that OPLL patients had a better fusion environment for OPLL enhanced the stability compared to patients without OPLL after PLF. However, a more harmonic cervical alignment may alleviate the axial symptom of patients with regional kyphosis even their C2-C7 SVA were in a relative accredited degree.
Our present study first compared the outcomes of PLF managing regional kyphotic CSM and negative K-line OPLL and proposed the similar sufficient decompression effect for cervical spinal cord. Our indication selection of PLF renew the existing surgical strategy for these two exceptional conditions limited to anterior or combined approaches. Nevertheless, in the setting of global kyphotic cervical alignment with large C2-C7 SVA, dorsal drift of the spinal cord was limited[2] and we consistently thought it is a contraindication for PLF. The limitation of present study targeted to K-line was absence of analysis of the canal occupying ratio and locations of the OPLL lesion. The cervical alignment was not compared quantitatively between two groups. Our sample size is relatively small and there is a need for randomized controlled studies.