Differences in TLIF and posterolateral fusion outcomes limited in terms of cost-effectiveness


A study from Norton Leatherman Spine Center, Louisville, USA, has shown that, while TLIF (Transforaminal lumbar interbody fusion) may generate better one-year ODI (Oswestry disability index) outcomes for spondylolisthesis, it may not produce any other significantly better outcomes than posterolateral fusion (PSF). The results also call into question the common categorisation of PSF as a “smaller” surgery.

The study was presented by Steven Glassman at the 30th annual meeting of the North American Spine Society. The researchers analysed a large national outcomes registry, N2QOD (the National Neurosurgery Quality and Outcomes Database), using a propensity matching technique to match TLIF and PSF patients according to similar demographics. The aim of the study was to discover if the theoretical benefits of TLIF over PSF—potentially increased fusion rate, deformity correction and more complete foraminal decompression, for example—could be shown in practise.

N2QOD is a qualified, prospective, multi-centre registry, enrolling patients based on diagnostic criteria. The researchers used N2QOD data to analyse the relative benefits of TLIF and PSF for particular diagnoses, including spondylolisthesis, stenosis, and adjacent segment disease. To account for the fact that treatment by TLIF or posterolateral fusion is often distributed across different demographic groups—PSF in older patients, and TLIF in smokers—the team used a propensity matching technique to compare patients with similar demographics.

The study was limited by the predominance of cases of spondylolisthesis, stenosis and adjacent segment disease in the N2QOD database. The researchers did not have enough participants to utilise the propensity-matching technique in the diagnoses of disc herniation, recurrent disc herniation and mechanical disc collapse, although in the latter case this may be due its late addition to the registry.

The 109 spondylolisthesis patients propensity-matched from the registry revealed almost identical estimated blood loss (474cc vs 406cc) and operative times (201.3 mins vs 201.6 mins) for PSF and TLIF, respectively. Very similar results were produced between the procedures in the stenosis (439.6cc vs 360.8cc, 193 mins vs 178.9 mins, n=63) and adjacent segment disease (439.3cc vs 390.5cc, 207.5 mins vs 190 mins, n=47) cohorts. Glassman said, “We tend to think of TLIF as a ‘bigger’ operation, but in this fairly large group, that is not what the data showed.”

As well as similar surgical variables between the two procedures, the one-year HRQOL outcome scores were almost identical. In all three diagnoses back pain, leg pain and EuroQOL-5D were not significantly different. The only significant difference was the ODI change for spondylolisthesis (20.8 for PSF vs 29.4) for TLIF. Also, the minimal clinically-important difference threshold was achieved 18% more frequently in TLIF than in PSF. For the other diagnoses, however, this difference was not observed.

Glassman reported that, whilst TLIF did generate somewhat better outcomes for spondylolisthesis patients, its overall benefits were not dramatic. Quality-adjusted life years (QALYs) are driven by procedural outcomes, he claims, and these results might make it difficult to prove the cost-effectiveness of TLIF over PSF. “This really demonstrates the need to better define the exact patient in whom TLIF provides a better outcome, because those are the only patients where we have a hope of demonstrating cost-effectiveness.