Article Reviewed Narayanan R, Ezeonu T, Heard JC, Lee YA, Yeung CM, Henry T, Kellish A, Kohli M, Canseco JA, Kurd MF, Hilibrand AS, Vaccaro AR, Schroeder G, Kepler C, & Kaye ID. Which North American Spine Society disc herniation morphology descriptors are most associated with improvements in clinical outcomes after microdiscectomy? N Am Spine Soc J. 2024 Jun 1;19:100336. https://doi.org/10.1016/j.xnsj.2024.100336 Commentary The reviewed study by Narayanan et al, is a retrospective cohort series that aimed to evaluate inter-rater reliability and correlation to patient reported outcomes for surgically treated disc herniations characterized by the 2001 North American Spine Society (NASS) naming system.

To that end, lumbar microdiscectomy patients at a single academic institution between 2014 and 2021 were retrospectively studied. Inclusion criteria were patients who underwent a single-level lumbar decompression with discectomy and were above the age of 18. Exclusion criteria included the lack of preoperative MRI, or the lack of patient reported outcome measures (PROM) at either 3- or 12-month follow-up.

Structured Query Language searches and manual chart reviews were used to extract patient demographics, surgical characteristics, and admissions data. Preoperative MRIs were assessed and classified by an orthopedic surgery resident and spine fellow using NASS nomenclature. PROM data was assessed at baseline, 3-month postoperatively, and 12-month postoperatively. Changes in PROMs were analyzed in relation to literature-reported minimal clinically important difference (MCID).

With this data, backwards multivariate logistic regression identified NASS descriptors associated with MCID in a variety of PROM metrics, including Oswestry Disability Index (ODI), mental component score (MCS), physical component score (PCS), and Visual Analog Scale (VAS) in the back and leg. Cohen’s kappa coefficient was also computed to judge interrater reliability.

A total of 213 patients met study criteria and were included. Mean age was 44.1+/-12.8 years old. As expected, most discectomies were performed at L5-S1 (61.0%) or L4-5 (34.3%). Disc herniation characteristics included the following: area of disc herniation (average 115.4 mm2), broad based (19.7%), extrusion (43.7%), sequestered (10.8%), subarticular (61.5%).

The highest inter-rater reliability was found in the central canal area (κ=0.88) and sequestration status (κ=0.88), while the lowest inter-rater reliability was found in the direction of migration (κ=0.53). Predictive descriptors of MCID in PROM metrics at three-months post-op included sequestration predicting ODI (OR=0.11) and MCS (OR=0.25), axial disc herniation area predicting ODI (OR=1.03) and MCS (OR=1.02), and extrusion predicting PCS (OR=0.37). At 12 months, predictive descriptors included sequestration predicting ODI (OR=0.07), axial disc herniation area predicting VAS in leg (OR=1.05), central canal area predicting VAS in leg (OR=0.98), percent central canal involvement predicting MCS (OR=1.07), and extrusion predicting MCS (OR=0.34) and VAS in leg (OR=0.4). Notably, there were only three descriptors that predicted MCID at both 3- and 12-month follow up: sequestration status, disc herniation area, and extrusion status. Extrusion status, however, was not considered as a viable predictive factor as there was poor reliability in the metric.

The results of the reviewed study are of clinical interest. Unto itself, utilization of the NASS naming system underscores the importance of this common nomenclature. As the authors note, standardization of terminology is important for the interpretation of studies.

Additionally, the concept of correlating clinical outcomes (as measured by PROMs) with different disc morphologies is very useful. Clearly, disc herniations correlated with most positive clinical outcomes may be best candidates for surgical intervention. Understanding these differences may be of clinical utility to both surgeons and patients.

As the authors acknowledge, there are limitations to the current study. For one, patients without 12-month follow up PROM data were excluded from the study, potentially introducing selection bias into the patient sample. Low reliability of some of the measures may have, unto itself, limited the predictive nature of those measures. Finally, the outcomes of the surgically treated cohort were not compared to the outcomes of those treated nonoperatively.

Overall, this reviewed study helped define the reliability and predictive efficacy of disc herniation descriptors as defined by the 2001 NASS nomenclature system. While reliability varied between descriptors, many descriptors, including central canal area and sequestration status, demonstrated strong inter-rater reliability. Furthermore, sequestration status and disc herniation area were the most predictive descriptors of clinical outcomes.