The difference between preoperative and 6-month follow-up VAS and ODI were assessed and data was statistically compared by paired Student t-test, and a value of <0.05 was considered significant.

All patients received 4 doses of antibiotics (1 previous night of surgery, 1 intraoperatively, and 2 after surgery) at 12 hours apart. All patients underwent surgery with a posterior approach. Needle docking was done on side of the symptoms and pain. Mid pedicular line entry point was used for patients having unilateral pain without evidence of contralateral lateral recess or foraminal stenosis (Figure 1A). These patients underwent ipsilateral lateral flavectomy only (flavum sparing).

Medial pedicular line was used for patients with central canal stenosis and patients with bilateral symptoms with evidence of bilateral lateral recess and/or foraminal stenosis with ligamentum flavum thickening leading to bilateral lateral recess or foraminal stenosis (Figure 1B). Patients presenting with bilateral symptoms with no evidence of leg pain, surgical entry side was decided on the side of pathology as evident on the MRI. Over the top contralateral decompression was done for these patients if pathology dictates.

After initial marking of entry site, dilator amd endoscopic sleeve was docked. Soft tissue clearance was performed to see raw bone using plasma energy and 90° wand. Cranial lamina and spinolaminar junction were identified (Figure 2A). The inferior articular process (IAP) was defined, and the inferior tip of IAP was identified and dissected using a combination of rongeurs and plasma energy (Figure 2B). Incremental osteotomy of the IAP performed with a sharp osteotome from caudal to cranial harvesting bone for interbody grafting at a later stage. Resection of the IAP was done till the inferior endplate of the cranial vertebra as confirmed on the lateral projection of the image intensifier, leaving behind part of the pars interarticularis intact to protect the exiting root during cage insertion (Figure 2C and D).

Cranial laminotomy was performed till visualization of medial splitting of the ligamentum flavum (Figure 3A). The medial edge of the superior articular process and the tip identified and sequentially resected using a combination of osteotome and Kerrison rongeurs. Caudal laminotomy performed to decompress the traversing root. Contralateral decompression performed in patients with bilateral symptoms. Ipsilateral resection of ligamentum flavum is performed to expose lateral border of thecal sac and extended contralaterally if required (Figure 3C). A safe corridor was created in the axilla between the exiting and traversing root with medial border at the medial pedicular line.

Annulotomy was performed and vertebral endplates were prepared using mechanical shavers and curettes (Figure 4A and B). Near to complete discectomy and end plate preparation was performed by tilting the working sleeve of the endoscope to gain access to corners of the endplate. The anterior disc space was packed with bone chips (harvested during the surgery) as graft. Appropriate size cage was introduced anteriorly using a semicircular retractor to protect the traversing root. The exiting root was protected by the part of the pars interarticularis which was left behind. Once the cage was placed, part of the left behind pars interarticularis was resected to decompress the exiting root. The decompression of exiting root was necessary only in patients with foraminal pathology. The incision was closed over a drain. All patients underwent posterior percutaneous pedicle screw fixation after interbody fusion.

The drain was removed on the next day and patients were mobilized & discharged on postoperative day 1 of surgery with oral antibiotics for 5 days. Patients diagnosed with osteoporosis, as confirmed by dual-energy x-ray absorptiometry, began treatment with teriparatide injections along with calcium and vitamin D supplementation.

Among the 53 patients, 19 were male and 34 females with an average age of 58.2 (range, 24–79) years. The most common affected level was L4–5 (36 of 53), followed by L5–S1 (13 of 53), L3–4 (3 of 53) and multiple in 1 case. All patients had back pain, and 49 patients had leg pain. Motor weakness was seen in 37 patients, with single myotomal weakness in 34 and foot drop in 3 patients. Only 1 patient had bladder involvement. The average duration of symptoms was 34.8 (range, 15 days–180 months) months. The mean follow-up period was 20.9 (range, 6–36) months.

The preoperative mean VAS score was 7.46 (range, 6–9) and mean ODI was 57.9 (range, 25–93). On postoperative follow-up at 6 months the mean VAS score was 1.65 (range, 0–9) and mean ODI score was 18.19 (range, 0–55) (Figure 5). The improvement in VAS and ODI was statistically significant (p<0.001). Neurological improvement (motor weakness) was seen in all patients with some residual weakness in 3 patients (all were patients with footdrop) (Table 1).

Mean duration of surgery lasted 182 (range, 124–240) minutes. This time gradually decreased owing to learning curve. The intraoperative blood loss was minimal.

Many of the patients were operated on within 24 months of the study. Long-term follow-up was not available in all the patients to assess the fusion rates. Stability for short-term follow-up was assessed with cage/screw position. The final radiological outcome was accepted stable and fused if no periscrew loosening or broken implants was present. Out of 53 patients, 2 had retropulsion of cage not beyond the posterior vertebral line. They were followed up, but no further retropulsion was noted on follow-up at 6 months or beyond.

One patient had persistent back on follow at 3 months and had to undergo medial branch block followed by rhizotomy at the adjacent segment. On follow-up at 6 months, he had minimal to no residual back pain. Two patients had incomplete retropulsion of the cage on initial follow-up x-rays and no further increase in retropulsion of cage was noted at 2-year follow-up as confirmed on the follow-up x-rays. One patient presented with persistent low back pain following an episode of urinary tract infection 3 weeks after the surgery. On evaluation, he had deep seated infection for which he underwent full-endoscopic debridement, removal of loose cage with retention of the posterior instrumentation. Six weeks injectable antibiotics followed by 6 weeks of oral antibiotics. The infection resolved as confirmed on clinical symptoms, erythrocyte sedimentation rate, C-reactive protein (CRP).

None of the patients had neurological complications in the postoperative period.