Early mechanical outcomes of two expandable interbody cages in minimally invasive TLIF: a comparative single-surgeon study

Introduction

Interbody fusion using the transforaminal lumbar interbody fusion (TLIF) or posterior lumbar interbody fusion (PLIF) approach has become a mainstay for degenerative lumbar pathology, offering circumferential fusion through a single posterior corridor. The TLIF technique, introduced by Harms and Rolinger (1), uses unilateral or bilateral facetectomy and disc-space access through Kambin’s triangle (2), minimizing retraction of neural elements. Refinements led to minimally invasive surgery TLIF (MIS-TLIF), which employs tubular retractors or pedicle screw–based retraction (3) and percutaneous instrumentation to reduce muscle disruption, blood loss, and postoperative pain (4). Multiple studies have shown that MIS-TLIF achieves clinical outcomes and fusion rates comparable to open TLIF, with faster recovery and shorter hospitalization (3-12). Navigation (13) and bone graft substitutes such as bone morphogenetic protein (14) have contributed to shorter length of stay.

One evolving aspect of TLIF is interbody cage design. Traditional static cages [polyetheretherketone (PEEK) or titanium] are inserted in a fixed form and can require aggressive endplate preparation and impaction, which, while ensuring a snug fit, may contribute to endplate injury or subsidence if not performed carefully (15-18). Subsidence is undesirable because it produces loss of disc height and lordosis and potentially foraminal stenosis or hardware loosening, and it is described more frequently after posterior than after anterior or lateral fusions (19-26). Established measures to reduce subsidence include placing the cage on the strong peripheral apophyseal ring and using a cage with a sufficiently large footprint to distribute load (15,19,21,27,28). Cage material also contributes: high-modulus titanium can increase endplate stress relative to PEEK (29-32), and larger cages are associated with lower subsidence through broader load distribution (5,30).

Expandable cages were developed to address some limitations of static implants (33). These devices are inserted at reduced height and expanded in situ, in principle allowing a larger final footprint or height without excessive impaction and enabling improved disc-height and foraminal restoration with less insertional endplate trauma. Expandable TLIF cages have been shown to produce greater early postoperative disc and foraminal height than static cages (28,34-36).

Recent series have, however, raised concerns about the mechanical behavior of certain expandable designs, reporting non-trivial rates of subsidence, secondary collapse of the expansion mechanism, and pseudarthrosis requiring revision (37). Cage collapse—failure of the expandable implant itself, distinct from subsidence into bone—has been sparsely characterized. A notable report by Hejrati et al. (37) examined the Catalyft PL expandable cage (Medtronic Spine & Biologics, Memphis, TN, USA)—the same device used in the present study—in a single-center series of 53 patients. They described a high secondary collapse rate: 20 of 92 implanted cages (21.7%) collapsed postoperatively, affecting 20.8% of patients, at a mean of approximately 4 months, with some cases requiring revision. Younger patients (<70 years) were at higher risk. The authors also reported a 13.2% pseudarthrosis rate and a 7.6% reoperation rate. These findings underscore that the failure modes of expandable cages require scrutiny.

While several expandable devices are now available, comparative data assessing whether different designs behave equivalently under similar surgical conditions are scarce. Most reports focus on a single implant and do not control for variability in surgical technique, disc preparation, or fixation, making it difficult to disentangle device-related from surgeon-related factors.

The present study reports a single-surgeon experience with two distinct expandable TLIF cages—Catalyft PL and X-PAC (Expanding Innovations, Mountain View, CA, USA)—implanted using an identical disc-preparation set and pedicle screw construct. Our primary aim was to compare the rate of secondary cage collapse between these devices at the patient level. Secondary aims were to compare the incidence and severity of subsidence and to describe pseudarthrosis and reoperation. The devices differ in expansion mechanism: the Catalyft PL cage uses an anteriorly threaded inner screw to drive expansion, a mechanism that is not positively constrained, whereas the X-PAC device uses a unidirectional ratcheting mechanism (38). We hypothesized that, under otherwise comparable surgical conditions, the Catalyft PL cage would be associated with a higher rate of collapse than the X-PAC cage. We present this as a hypothesis-generating comparison rather than a test of mechanism, given the observational, sequential-cohort design. We present this article in accordance with the STROBE reporting checklist (39) (available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0135/rc).

Methods

Ethics

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was reviewed and approved by the Institutional Review Board of University of Nevada, Reno (approval No. 2466873-1). The requirement for individual informed consent was waived owing to the retrospective analysis of de-identified data.

Study design and patient population

This is a retrospective, non-randomized cohort analysis of consecutive patients who underwent one- or two-level MIS TLIF for degenerative lumbar pathology by a single surgeon at Renown Regional Medical Center and Northern Nevada Medical Center. Catalyft PL and X-PAC cases were performed sequentially, the Catalyft PL cases between March 2022 and February 2025 and the X-PAC cases between May 2025 and March 2026, with an 89-day interval between the final Catalyft PL case and the first X-PAC case; the device change reflected a complete transition in implant preference rather than concurrent selection. Patients were divided into two cohorts by interbody implant: (I) the Catalyft PL cohort (Medtronic Catalyft PL titanium expandable cage; 112 consecutive patients, 144 cages), and (II) the X-PAC cohort (Expanding Innovations X-PAC cage; 56 consecutive patients, 80 cages. Baseline characteristics are summarized in Table 1, and the study population is summarized in Table 2.

Because devices were used in sequential, non-overlapping periods, this design permits assessment of an association between cage type and early radiographic collapse but cannot establish that the expansion mechanism itself caused any observed difference; unmeasured temporal factors (e.g., drift in endplate-preparation technique, evolving postoperative activity guidance, or radiographic interpretation) cannot be fully excluded.

Surgical technique and instrumentation

All procedures used a standardized unilateral MIS TLIF Wiltse approach through tubular retractors with facetectomy and discectomy; bilateral facetectomies were performed with anatomic reduction as indicated (3-5,40). Percutaneous pedicle screw fixation used the Voyager system (Medtronic Spine & Biologics, Memphis, TN, USA) in all cases (Figure 1). Intraoperative navigation with O-arm imaging and StealthStation guidance (Medtronic Spine & Biologics, Memphis, TN, USA) was used in every case for screw placement, disc-space localization, endplate preparation, and cage trajectory; fluoroscopy confirmed final cage position. Disc preparation included meticulous discectomy and careful endplate preparation preserving subchondral bone. Given the retrospective design, surgical technique, grafting strategy, and instrumentation were comparable and did not differ substantially between cohorts, rather than being experimentally controlled or held constant.

Figure 1 Intraoperative image showing retractor placement for MIS TLIF. Bilateral Wiltse exposures were performed through 4 cm incisions with bilateral facetectomies. The Voyager pedicle screw–based retractor system (Medtronic Spine & Biologics, Memphis, TN, USA) with percutaneously placed pedicle screws was used. MIS, minimally invasive surgery; TLIF, transforaminal lumbar interbody fusion.

Grafting strategy

Local autograft harvested during facetectomy and decompression was used in every patient, with supplemental cancellous allograft as needed. Recombinant human bone morphogenetic protein-2 (rhBMP-2; Infuse, Medtronic Spine & Biologics, Memphis, TN, USA) was used in all cases at one extra-small kit per level, placed within the interbody space. Iliac crest autograft was not routinely used. The grafting strategy was comparable across cohorts.

Cage type and dimensions

The device used throughout this series was the Catalyft PL, not the longer, hooked-end Catalyft PL40. Cage footprints differed between cohorts: Catalyft PL implants were predominantly 10 mm × 27.3 mm (approximately 273 mm² footprint) and X-PAC implants predominantly 12 mm × 32 mm (approximately 384 mm² footprint)—a 41% larger contact area in the X-PAC cohort. This difference is central to the interpretation of the subsidence findings and is addressed in the analysis and discussion.

Cage placement

Sequential navigated dilators were used to release the disc space, with undersized navigated shavers for endplate preparation. The disc space was pre-packed with approximately two-thirds of the available graft and rhBMP-2, and the cage was placed under fluoroscopy with the target position on the anterior apophyseal ring confirmed with live imaging. Expansion was performed under fluoroscopic guidance using combined imaging and tactile feedback to determine final height. Remaining graft and rhBMP-2 were packed through the cage. Capped rods were then placed, undersized where possible, to restore the posterior tension band and improve lordosis.

Data collection and radiographic outcomes

Demographic and operative variables included age, sex, body mass index (BMI), operated level(s), single- versus two-level status, smoking status, osteoporosis, and prior surgery at the operated segment (obtained from clinical histories rather than the operative record). Standing lateral radiographs were obtained immediately postoperatively or intraoperatively (baseline) and at approximately 6 weeks, 3 months, and 6 months as part of routine clinical follow-up. The primary endpoint was radiographic secondary cage collapse at the patient level, adjudicated from the final available follow-up radiograph (target approximately 6 months). All radiographic outcomes were scored de novo under the IRB-approved protocol by three authorized reviewers who independently assessed each study using predefined criteria. When reviewer determinations differed, discrepancies were discussed jointly and resolved by consensus; the consensus score was recorded as the final analytic value.

Subsidence severity was graded using a four-grade scheme adapted from Yamagata et al. (41) as a radiographic guide: grade 0 (none), grade 1 (<1/3 loss of vertebral body height vs. baseline imaging), grade 2 (1/3–2/3 loss), and grade 3 (>2/3 loss). For the present study, any grade 1 or higher was classified as subsidence present. Secondary cage collapse was defined as observable structural loss of implant height from mechanical deformation or failure of the expansion mechanism, distinct from subsidence into the endplate, using an operational threshold of >2 mm height loss relative to immediate postoperative or intraoperative baseline imaging. Cases were categorized as collapse with associated subsidence or isolated structural collapse. Radiographic pseudarthrosis at approximately 6 months was defined as absence of visible bridging bone, progressive hardware loosening or halo formation, or clear nonunion on standing films; computed tomography (CT) was obtained selectively when clinically indicated.

Reviewer assessment, reliability, and ascertainment bias

Radiographic measurements were independently performed by three reviewers using predefined criteria. When reviewer determinations differed, the case was discussed jointly and a single consensus result was recorded. Individual reviewer-level ratings were not retained in the analytic dataset after consensus adjudication; formal inter-observer reliability statistics (Cohen’s κ for binary collapse and weighted κ or an intraclass correlation coefficient for subsidence grade) therefore cannot be computed retrospectively. This is acknowledged as a limitation. Because the two devices have visually distinct radiographic profiles, blinding to cage type was not feasible; this introduces potential ascertainment bias in collapse detection, which we sought to mitigate through independent multi-reviewer assessment, consensus adjudication, and a pre-specified operational threshold. We acknowledge that 2-dimensional (2D) radiographs carry projection variance of approximately ±1–2 mm and that the >2 mm collapse threshold approaches this range; the magnitude of the observed between-group difference (16.7% vs. 1.2% of cages) substantially exceeds what projection variance alone would be expected to produce, but marginal cases cannot be excluded, and CT confirmation was not routinely available.

Statistical analysis

The primary analysis was the between-group comparison of secondary cage collapse at the patient level, with each patient contributing a single observation. In response to the reviewers, the patient-level endpoint was adopted as the primary analysis, and a clustered cage-level analysis and a footprint-restricted subsidence sensitivity analysis were added to address within-patient correlation and cage-size confounding, respectively; the six-comparison multiple-testing framework was specified at the time of the original analysis. Cage-level analyses, in which multiple cages within a patient are not independent, were reported as supportive secondary analyses and interpreted as descriptive; to address non-independence, a clustered generalized estimating equation (GEE) logistic model (binomial family, logit link, exchangeable working correlation, patient as the clustering unit, with robust sandwich variance estimation) was fitted for collapse and for subsidence and was consistent with the between-group association after accounting for within-patient correlation [collapse odds ratio 15.2, 95% confidence interval (CI): 2.0–116.0, P=0.009; subsidence odds ratio 3.5, 95% CI: 1.5–7.9, P=0.003]. The wide CI for the collapse odds ratio reflects the single collapse event in the X-PAC arm, and this clustered estimate is interpreted as supportive of the patient-level primary result rather than as an independent confirmatory test. In two-level cases, both cages contributed to cage-level counts and to the clustered model, while the patient-level endpoint recorded a patient as affected if any implanted cage met the outcome definition.

Between-group comparisons of categorical outcomes used Fisher’s exact test throughout, for consistency across outcomes with small or zero event counts. Effect sizes were reported as absolute risk difference, relative risk (RR), and number needed to harm, with 95% CIs; proportions used the Wilson score interval and relative risks the log-scale method with a continuity correction where a zero cell was present. Six comparisons were pre-specified—(I) collapse at patient level, (II) collapse at cage level, (III) subsidence at patient level, (IV) subsidence at cage level, (V) pseudarthrosis, and (VI) the time-to-collapse comparison—giving a Bonferroni-corrected threshold of α=0.008. Unadjusted P values are reported alongside this threshold, and results are interpreted accordingly.

The pre-specified time-to-collapse comparison (VI) could not be validly performed. Patients underwent serial follow-up radiography at approximately 6 weeks, 3 months, and 6 months, but the analytic dataset did not preserve a reliable exact date of collapse onset or first occurrence for every event. Collapse was therefore adjudicated at the final available radiographic follow-up, and the event-time and number-at-risk information required for a valid Kaplan-Meier analysis could not be reconstructed consistently. Accordingly, no Kaplan-Meier or log-rank analysis was conducted, and secondary collapse is reported as cumulative incidence at final follow-up rather than by survival analysis. The Bonferroni divisor was retained at six pre-specified comparisons, which is conservative given that five were ultimately executed. A sensitivity analysis for subsidence, added to address cage-size confounding, restricted the comparison to cages with a footprint ≥360 mm² to reduce confounding by cage size. A broad multivariable subsidence model was not planned because the number of outcome events, particularly in the X-PAC group, was insufficient relative to the number of clinically relevant covariates requested and the clustered cage-level structure to support stable adjustment without substantial overfitting and unstable estimates. No fully adjusted multivariable subsidence model was therefore fitted; the subsidence comparison is reported as exploratory. Analyses were performed with IBM SPSS Statistics, version 29.

Data quality

Radiographic measurements were performed independently by three experienced observers with consensus resolution of discrepant determinations. Electronic database interrogation used validation rules and range checks, with cross-reference of dates against operative reports and picture archiving and communication system (PACS) records. Each included patient had baseline imaging and at least one postoperative standing lateral radiograph sufficient to adjudicate the primary endpoint; 95% of patients had radiographic follow-up through approximately 6 months. For patients without an approximately 6-month study, the final available follow-up radiograph was used for endpoint adjudication. Baseline nonradiographic data were >95% complete.

Results

Study population

We analyzed 168 consecutive MIS TLIF patients: 112 with 144 Catalyft PL cages and 56 with 80 X-PAC cages (Table 2). All included patients had evaluable baseline and postoperative imaging for the primary endpoint, and 95% had radiographic follow-up through approximately 6 months; the final available follow-up radiograph was used for endpoint adjudication. Mean radiographic follow-up was 6.19 months in the Catalyft PL cohort and 7.39 months in the X-PAC cohort. Because the X-PAC cohort, which had the lower collapse rate, also had the longer mean follow-up, it had more rather than less opportunity for late collapse to be detected; the follow-up difference therefore does not account for the observed between-group difference. Baseline demographics were broadly comparable between cohorts with respect to age, smoking status, osteoporosis, and prior segmental surgery (Table 1). Cage footprint differed systematically (see Methods). These imbalances are considered in the analysis and Discussion, particularly for subsidence.

Secondary cage collapse (primary endpoint)

Secondary cage collapse occurred in 23 of 112 Catalyft PL patients (20.5%, 95% CI: 14.1–28.9%) compared with 1 of 56 X-PAC patients (1.8%, 95% CI: 0.3–9.4%), corresponding to a relative risk of approximately 11.5 (95% CI: 1.6–83.0) and a number needed to harm of approximately 5. This patient-level difference was statistically significant by Fisher’s exact test (P=0.001) and remained significant after Bonferroni correction (α=0.008) (Table 3).

At the cage level, collapse affected 24 of 144 Catalyft PL cages (16.7%, 95% CI: 11.5–23.6%) versus 1 of 80 X-PAC cages (1.2%, 95% CI: 0.2–6.7%) [risk ratio (RR) 13.3, 95% CI: 1.8–96.7; P<0.001]. Because cages within a patient are not independent, this cage-level result is supportive rather than primary; the clustered GEE model accounting for within-patient correlation was consistent with this association (odds ratio 15.2, 95% CI: 2.0–116.0; P=0.009). The 95% CIs for the relative risk and the odds ratio are wide owing to the single collapse event in the X-PAC arm, and the point estimates should be interpreted with corresponding caution.

Temporal analysis of collapse

Within the Catalyft PL cohort, collapse rates increased across sequential thirds of the series (10.8%, 27.0%, and 23.7%), a pattern opposite to what an operator learning effect would predict. X-PAC collapse rates remained low throughout (0%, 5.3%, 0%), with near-absence of failure from the first cases onward. Because collapse-onset timing was not recorded, these sequential-cohort rates are reported as cross-sectional cumulative incidence rather than as a time-to-event analysis. A representative Catalyft PL collapse and a representative X-PAC case are shown in Figures 2,3, respectively. The rising within-cohort collapse rate, combined with the short 89-day separation between cohorts and unchanged technique, argues against operator learning as the explanation but does not by itself establish a mechanistic cause.

Figure 2 Representative Catalyft PL case in which secondary collapse occurred, shown intraoperatively and at 6 months. Progressive loss of cage expansion is demonstrated over follow-up.

Figure 3 Representative X-PAC case, shown intraoperatively, at 6 weeks, and at 6 months. Note the endplate coverage of the larger cage. Only one X-PAC cage collapse occurred and subsidence was infrequent.

Subsidence

Subsidence occurred in 36 of 112 Catalyft PL patients (32.1%, 95% CI: 24.2–41.3%) and 41 of 144 cages (28.5%, 95% CI: 21.7–36.3%), versus 8 of 56 X-PAC patients (14.3%, 95% CI: 7.4–25.7%) and 8 of 80 cages (10.0%, 95% CI: 5.2–18.5%) (Table 3). No X-PAC case demonstrated grade 2 or 3 subsidence, whereas 5 of the 36 Catalyft PL cases with subsidence (13.9%) were grade 2 or higher. The patient-level comparison did not reach the Bonferroni-corrected threshold (Fisher P=0.02; α=0.008); the cage-level comparison was nominally stronger (P=0.001) but is subject to the non-independence noted above.

Critically, the subsidence comparison is confounded by cage geometry and accompanied by an additional procedural imbalance. First, X-PAC cages had a 41% larger endplate footprint, and footprint is one of the most robust independent predictors of subsidence in the lumbar interbody literature. A sensitivity analysis restricted to cages with footprint ≥360 mm² showed attenuation of the between-device difference, consistent with cage geometry accounting for a meaningful share of the observed subsidence differential. Second, the cohorts differed in multilevel proportion (X-PAC 42.9% vs. Catalyft PL 28.6%), an additional imbalance that was not adjusted for given limited event counts. For these reasons we do not attribute the subsidence difference to expansion mechanism, and we report subsidence as an exploratory, hypothesis-generating finding rather than a primary result.

Pseudarthrosis and reoperation

Radiographic pseudarthrosis occurred in 2 of 112 Catalyft PL patients (1.8%, 95% CI: 0.5–6.3%) and 0 of 56 X-PAC patients (0.0%, 95% CI: 0–6.4%); Fisher’s exact test showed no significant difference (P=0.55). This analysis is severely underpowered and cannot exclude a clinically meaningful difference. Both Catalyft PL pseudarthrosis cases were symptomatic, CT-confirmed, and associated with documented cage collapse, which is hypothesis-generating for a mechanistic link between mechanical failure and nonunion but cannot support that inference statistically. Mean follow-up (6.19 months in the Catalyft PL cohort and 7.39 months in the X-PAC cohort) is shorter than the 12-month minimum generally recommended for fusion adjudication, and pseudarthrosis results should be interpreted only as early observations.

Discussion

Expandable interbody cages are increasingly used in TLIF because expandability permits insertion through a narrow corridor while still pursuing sagittal restoration, foraminal decompression, and a broad endplate footprint. A recognized concern is that the initial gain in disc-space height may not be maintained, with post-implantation settling and, for some designs, mechanical failure of the implant itself—a mode that cannot occur with static cages (33,37,42).

Key findings

In this consecutive single-surgeon series, the Catalyft PL screw-based expandable cage was associated with a significantly higher rate of early secondary collapse than the X-PAC ratcheting cage at the patient level (20.5% vs. 1.8%; P=0.001), a difference robust to Bonferroni correction and supported by cage-level and clustered analyses. The magnitude and direction parallel the independent series of Hejrati et al. (37), which reported 21.7% collapse of Catalyft PL implants. Subsidence was also more frequent with Catalyft PL, but this comparison is confounded by the 41% larger X-PAC footprint and accompanied by a higher proportion of multilevel procedures in the X-PAC cohort (42.9% vs. 28.6%); the patient-level subsidence comparison did not survive correction for multiple testing. We therefore report subsidence as exploratory and do not attribute it to mechanism.

Strengths and limitations

Strengths include consecutive case inclusion, a single surgeon using comparable technique and identical instrumentation across cohorts, independent radiographic review by three observers with consensus adjudication of discrepancies, a within-cohort temporal analysis that argues against an operator learning effect, a patient-level primary endpoint with a clustered supportive analysis, and effect-size reporting with CIs.

Limitations are substantial and shape the conclusions. First, the design is retrospective and non-randomized, with devices used in sequential, non-overlapping periods; unmeasured temporal factors cannot be excluded, so the data show association rather than causation. Second, radiographic reviewers could not be blinded to cage type owing to distinct device profiles, introducing potential ascertainment bias despite mitigation; although initial assessments were performed independently, individual reviewer-level scores were not retained after consensus adjudication, so interobserver reliability could not be quantified retrospectively. Third, follow-up was short and relied on plain radiographs without routine CT, which is insufficient for definitive fusion or pseudarthrosis adjudication; conclusions regarding pseudarthrosis are correspondingly reserved. Fourth, the subsidence comparison is confounded by cage footprint and accompanied by a higher multilevel proportion in the X-PAC cohort and is reported as exploratory. Fifth, the number of outcome events relative to the requested covariates and clustered data structure was insufficient to support a stable fully adjusted multivariable model without substantial overfitting, so we relied on the size-restricted sensitivity analysis rather than broad covariate adjustment. Finally, this is a single-surgeon, single-center experience, and results may not generalize; the study is underpowered for pseudarthrosis and other low-frequency outcomes. Taken together, these features place the study in the category of an early mechanical-performance comparison rather than a definitive fusion-outcomes trial.

Comparison with similar research

Cage collapse of the Catalyft PL device was previously described by Hejrati et al. (37), a single-center retrospective study from a Swiss institution examining the Catalyft PL expandable titanium spacer in TLIF and PLIF between July 2022 and November 2023. Across 53 patients and 92 implanted spacers, mostly at L4/5 and L5/S1, the authors reported a secondary collapse rate of 21.7% of spacers (20.8% of patients), occurring on average approximately 4 months postoperatively; pseudarthrosis developed in 13.2% of patients and 7.6% required revision. Patients under 70 years were at nearly fourfold higher risk of collapse, while other patient and surgical variables were not significantly associated. Notably, collapse did not significantly affect segmental lordosis or clinical outcomes at last follow-up, with 54.7% achieving good or excellent MacNab scores—suggesting some patients had begun forming fusion mass before collapse. The authors discontinued the device in December 2023. Our patient-level collapse rate (20.5%) is closely concordant with their patient-level figure (20.8%), lending external consistency to the present finding.

Guven et al. (38) reported the largest clinical series of the X-PAC cage, from the Hospital for Special Surgery (Weill Cornell Medicine): 131 patients and 160 cages implanted between 2021 and 2023. Anterior and posterior disc height were significantly restored and maintained at 6 months, though the initial segmental lordosis gain was not statistically sustained at that timepoint. No complete cage collapses were observed, cage height loss occurred in only 1 case (0.8%), and grade II or higher subsidence was seen in only 1 patient. Revision surgery was required in 6.1% of patients, none for cage-related reasons. The authors acknowledged the retrospective design, absence of a comparison group, reliance on plain radiographs, and short follow-up, and concluded that the screwless design appears to be a safe option that may address the collapse vulnerability of screw-based expandable cages. Our X-PAC collapse rate (1.8% of patients, 1.2% of cages) is consistent with their near-zero collapse experience.

More broadly, the role of expandable cages in TLIF remains under active evaluation. They allow placement of a larger device through a small annulotomy, but questions remain about long-term maintenance of disc height and lordosis and about cost relative to static cages (43-48). The present findings speak specifically to differences among expandable designs and should not be extrapolated into a general claim about expandable versus static cages.

Explanations of findings

Within the limits of an observational comparison, the pattern of failure is consistent with loss of the Catalyft PL expansion mechanism in a subset of patients, whereas subsidence appears more closely related to endplate coverage and cage geometry. The two devices differ mechanistically—the Catalyft PL cage expands via an anteriorly threaded inner screw that is not positively constrained, whereas the X-PAC cage uses a unidirectional ratcheting mechanism with discrete mechanical stops (Figures 4,5)—and it is biologically plausible that a non-constrained screw-based mechanism is more vulnerable to gradual loss of expansion under cyclic loading, while larger footprints are associated with less subsidence. We emphasize, however, that our data establish an association, not causation: the cohorts were treated sequentially, the devices differ in more than one respect, and we did not perform bench testing, retrieval analysis, or CT-based mechanical assessment. The proposed mechanism is therefore offered as a hypothesis consistent with the observed pattern and with the manufacturer’s own 2025 field-safety action, not as a demonstrated cause.

Figure 4 Schematic of a typical screw-based expandable TLIF cage similar to the Catalyft PL device. Expansion is achieved through an internal screw driven forward, deflecting the superior and inferior anterior flanges open. The system is not positively locked; if the screw unthreads, loss of expansion can occur. This schematic is illustrative of the general mechanism class and is not derived from bench testing or retrieval analysis. TLIF, transforaminal lumbar interbody fusion.

Figure 5 Schematic of a ratcheting expandable cage similar to the X-PAC system. As the cage is expanded, cams advance along ratchet teeth; the driver is removed once expansion is achieved. This constrained mechanism is intended to resist loss of expansion. This schematic is illustrative of the general mechanism class and is not derived from bench testing or retrieval analysis.

Implications and actions needed

These findings support two practical considerations, stated at the appropriate level of certainty. First, surgeons using screw-based expandable cages such as the Catalyft PL device should be aware that early secondary collapse may occur and should follow such patients accordingly; our data, the independent series of Hejrati et al. (37), and the Food and Drug Administration (FDA) class 2 recall of the Catalyft PL and PL40 Expandable Interbody System—initiated by Medtronic on 16 September 2025 and posted by the FDA on 21 November 2025 for potential cage height loss or collapse with possible migration, subsidence, and/or neurological injury (49)—together raise sufficient concern to warrant careful device selection and independent prospective validation. We do not conclude that screw-based mechanisms should be avoided or abandoned; rather, that these convergent signals raise concern and warrant further independent validation. Second, the technical aspects of cage placement matter: anterior apophyseal placement and use of the largest footprint the anatomy allows are associated with lower subsidence, and subsidence was infrequent when a 12 mm × 32 mm cage was used.

Conclusions

In this consecutive single-surgeon MIS TLIF series, the Catalyft PL screw-based expandable cage was associated with a significantly higher rate of early secondary collapse than the X-PAC ratcheting expandable cage, a difference that persisted across the study period and was not explained by an operator learning effect. Because the cohorts were treated sequentially and differed in cage geometry and multilevel proportion, these data demonstrate an association rather than a causal effect of expansion mechanism, and the subsidence difference in particular is confounded and exploratory. The collapse findings are directionally consistent with an independent published series and with a 2025 FDA class 2 recall of the Catalyft PL system. Collectively, they warrant cautious device selection, longer follow-up with CT-based fusion assessment, and independent prospective validation before firm mechanistic or comparative conclusions are drawn.

Acknowledgments

We wish to acknowledge Neil Babu from the University of Nevada, Reno BS-MD Program, who aided with data analysis, review, and manuscript preparation.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0135/rc

Data Sharing Statement: Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0135/dss

Peer Review File: Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0135/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0135/coif). L.H.S.S. serves as an unpaid editorial board member of Journal of Spine Surgery from December 2024 to December 2026. L.H.S.S. reports consulting fees from Medtronic Spine & Biologics, Kyocera, Katomed, and Johnson & Johnson; and is the founder of, and holds an equity/leadership interest in SpinePoint, a company that produces static interbody cages. L.H.S.S. reports no financial relationship with Expanding Innovations, the manufacturer of the X-PAC device. L.H.S.S.’s continued use of other products from Medtronic (including navigation, pedicle screw systems, and graft materials) is noted for transparency. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Review Board of University of Nevada, Reno (approval No. 2466873-1), and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Harms J, Rolinger H. A one-stager procedure in operative treatment of spondylolistheses: dorsal traction-reposition and anterior fusion (author's transl). Z Orthop Ihre Grenzgeb 1982;120:343-7. [Article in German]. [Crossref] [PubMed]
2. Tumialán LM, Madhavan K, Godzik J, et al. The History of and Controversy over Kambin's Triangle: A Historical Analysis of the Lumbar Transforaminal Corridor for Endoscopic and Surgical Approaches. World Neurosurg 2019;123:402-8. [Crossref] [PubMed]
3. Liu H, Li J, Sun Y, et al. A Comparative Study of a New Retractor-Assisted WILTSE TLIF, MIS-TLIF, and Traditional PLIF for Treatment of Single-Level Lumbar Degenerative Diseases. Orthop Surg 2022;14:1317-30. [Crossref] [PubMed]
4. Kalina R. Minimally invasive transforaminal lumbar interbody fusion (MIS TLIF) in treatment of degenerative diseases of lumbosacral spine compared to modified open TLIF: a prospective randomised controlled study. Neurol Neurochir Pol 2024;58:503-11. [Crossref] [PubMed]
5. Kalina R. Minimally invasive transforaminal lumbar interbody fusion (MIS TLIF) in treatment of degenerative diseases of lumbosacral spine compared to modified open TLIF: a prospective randomised controlled study. Neurol Neurochir Pol 2024;58:503-11. [Crossref] [PubMed]
6. Faulks C, Cabbabe K, Biddau DT, et al. A single straight expandable cage via a hybrid posterior-transforaminal approach with rhBMP-2 or allograft provides high fusion rates with low risk of subsidence. J Spine Surg 2025;11:1-14. [Crossref] [PubMed]
7. Rathbone J, Rackham M, Nielsen D, et al. A systematic review of anterior lumbar interbody fusion (ALIF) versus posterior lumbar interbody fusion (PLIF), transforaminal lumbar interbody fusion (TLIF), posterolateral lumbar fusion (PLF). Eur Spine J 2023;32:1911-26. [Crossref] [PubMed]
8. McMordie JH, Schmidt KP, Gard AP, et al. Clinical and Short-Term Radiographic Outcomes of Minimally Invasive Transforaminal Lumbar Interbody Fusion With Expandable Lordotic Devices. Neurosurgery 2020;86:E147-55. [Crossref] [PubMed]
9. Zhang B, Ma JS, Feng P, et al. Clinical efficacy of minimally invasive transforaminal lumbar interbody fusion (MIS-TLIF) in the treatment of II° lumbar isthmic spondylolisthesis: A retrospective cohort study. Medicine (Baltimore) 2023;102:e35420.
10. Virk S, Vaishnav AS, Sheha E, et al. Combining Expandable Interbody Cage Technology With a Minimally Invasive Technique to Harvest Iliac Crest Autograft Bone to Optimize Fusion Outcomes in Minimally Invasive Transforaminal Lumbar Interbody Fusion Surgery. Clin Spine Surg 2021;34:E522-30. [Crossref] [PubMed]
11. Gao G, Cao L, Du X, et al. Comparison of Minimally Invasive Surgery Transforaminal Lumbar Interbody Fusion and TLIF for Treatment of Lumbar Spine Stenosis. J Healthc Eng 2022;2022:9389239. [Crossref] [PubMed]
12. Kim CH, Easley K, Lee JS, et al. Comparison of Minimally Invasive Versus Open Transforaminal Interbody Lumbar Fusion. Global Spine J 2020;10:143S-50S. [Crossref] [PubMed]
13. Wang Y, Hu Y, Liu H, et al. Navigation Makes Transforaminal Lumbar Interbody Fusion Less Invasive. Orthopedics 2016;39:e857-62. [Crossref] [PubMed]
14. Faundez A, Tournier C, Garcia M, et al. Bone morphogenetic protein use in spine surgery-complications and outcomes: a systematic review. Int Orthop 2016;40:1309-19. [Crossref] [PubMed]
15. Levy HA, Pinter ZW, Reed R, et al. Transforaminal lumbar interbody fusion subsidence: computed tomography analysis of incidence, associated risk factors, and impact on outcomes. J Neurosurg Spine 2024;41:463-72. [Crossref] [PubMed]
16. de Kunder SL, van Kuijk SMJ, Rijkers K, et al. Transforaminal lumbar interbody fusion (TLIF) versus posterior lumbar interbody fusion (PLIF) in lumbar spondylolisthesis: a systematic review and meta-analysis. Spine J 2017;17:1712-21.
17. Chang MC, Kim GU, Choo YJ, et al. Transforaminal Lumbar Interbody Fusion (TLIF) versus Oblique Lumbar Interbody Fusion (OLIF) in Interbody Fusion Technique for Degenerative Spondylolisthesis: A Systematic Review and Meta-Analysis. Life (Basel) 2021;11:696.
18. Zhang JK, Greenberg JK, Javeed S, et al. Predictors of Postoperative Segmental and Overall Lumbar Lordosis in Minimally Invasive Transforaminal Lumbar Interbody Fusion: A Consecutive Case Series. Global Spine J 2025;15:425-37. [Crossref] [PubMed]
19. Macki M, Anand SK, Surapaneni A, et al. Subsidence Rates After Lateral Lumbar Interbody Fusion: A Systematic Review. World Neurosurg 2019;122:599-606. [Crossref] [PubMed]
20. Alan N, Vodovotz L, Muthiah N, et al. Subsidence after lateral lumbar interbody fusion using a 3D-printed porous titanium interbody cage: single-institution case series. J Neurosurg Spine 2022;37:663-9. [Crossref] [PubMed]
21. Shen S, You X, Ren Y, et al. Risk Factors of Cage Subsidence Following Oblique Lumbar Interbody Fusion: A Meta-analysis and Systematic Review. World Neurosurg 2024;183:180-6. [Crossref] [PubMed]
22. Zavras AG, Federico V, Nolte MT, et al. Risk Factors for Subsidence Following Anterior Lumbar Interbody Fusion. Global Spine J 2024;14:257-64. [Crossref] [PubMed]
23. Amorim-Barbosa T, Pereira C, Catelas D, et al. Risk factors for cage subsidence and clinical outcomes after transforaminal and posterior lumbar interbody fusion. Eur J Orthop Surg Traumatol 2022;32:1291-9. [Crossref] [PubMed]
24. De Stefano FA, Dharia AA, Guillotte AR, et al. Radiographic and Clinical Comparison of Lateral Lumbar Interbody Fusion Versus Transforaminal Lumbar Interbody Fusion With Expandable Cage. Int J Spine Surg 2026;20:46-52. [Crossref] [PubMed]
25. Ohiorhenuan IE, Walker CT, Zhou JJ, et al. Predictors of subsidence after lateral lumbar interbody fusion. J Neurosurg Spine 2022;37:183-7. [Crossref] [PubMed]
26. Wu H, Shan Z, Zhao F, et al. Poor Bone Quality, Multilevel Surgery, and Narrow and Tall Cages Are Associated with Intraoperative Endplate Injuries and Late-onset Cage Subsidence in Lateral Lumbar Interbody Fusion: A Systematic Review. Clin Orthop Relat Res 2022;480:163-88. [Crossref] [PubMed]
27. Bin Abd Razak HR, Dhoke P, Tay KS, et al. Single-Level Minimally Invasive Transforaminal Lumbar Interbody Fusion Provides Sustained Improvements in Clinical and Radiological Outcomes up to 5 Years Postoperatively in Patients with Neurogenic Symptoms Secondary to Spondylolisthesis. Asian Spine J 2017;11:204-12.
28. Canseco JA, Karamian BA, DiMaria SL, et al. Static Versus Expandable Polyether Ether Ketone (PEEK) Interbody Cages: A Comparison of One-Year Clinical and Radiographic Outcomes for One-Level Transforaminal Lumbar Interbody Fusion. World Neurosurg 2021;152:e492-501. [Crossref] [PubMed]
29. Liu SX, Zeng TH, Chen CM, et al. 3D-printed porous titanium versus polyetheretherketone cages in lateral lumbar interbody fusion: a systematic review and meta-analysis of subsidence. Front Med (Lausanne) 2024;11:1389533. [Crossref] [PubMed]
30. Faizan A, Kiapour A, Kiapour AM, et al. Biomechanical analysis of various footprints of transforaminal lumbar interbody fusion devices. J Spinal Disord Tech 2014;27:E118-27. [Crossref] [PubMed]
31. Yee-Yanagishita C, Fogel G, Douglas B, et al. Biomechanical comparison of subsidence performance among three modern porous lateral cage designs. Clin Biomech (Bristol) 2022;99:105764. [Crossref] [PubMed]
32. Fogel G, Martin N, Williams GM, et al. Choice of Spinal Interbody Fusion Cage Material and Design Influences Subsidence and Osseointegration Performance. World Neurosurg 2022;162:e626-34. [Crossref] [PubMed]
33. Lai TW, Chen PM, Li CH, et al. Clinical outcome and complications comparison between expandable and static cages in open TLIF surgery: A 2-year retrospective study. Medicine (Baltimore) 2025;104:e44042. [Crossref] [PubMed]
34. Kucharzyk DW, Budimir D, Waldorff EI, et al. The effect of expandable versus static lordotic interbody implants in minimally invasive spine surgery: patient reported outcomes, sagittal alignment, and restoration of disc height and foraminal height. J Spine Surg 2023;9:39-53. [Crossref] [PubMed]
35. Daher M, Aoun M, Hurley RK, et al. Static versus expandable transforaminal lumbar interbody fusion cages: a meta-analysis of complications, clinical, and radiographic outcomes. Spine J 2026;26:329-40. [Crossref] [PubMed]
36. Ledesma JA, Lambrechts MJ, Dees A, et al. Static versus Expandable Interbody Fusion Devices: A Comparison of 1-Year Clinical and Radiographic Outcomes in Minimally Invasive Transforaminal Lumbar Interbody Fusion. Asian Spine J 2023;17:61-74. [Crossref] [PubMed]
37. Hejrati N, Martens B, Jost B, et al. Failure of an expandable lumbar interbody spacer - a critical analysis of secondary collapse, pseudoarthrosis and revision rates after thoracolumbar fusion. Eur Spine J 2025;34:182-90. [Crossref] [PubMed]
38. Guven AE, Zadeh AT, Chiapparelli E, et al. Short-Term Outcomes of a Novel Non-Screw Based Expandable Posterior Interbody Fusion Cage. World Neurosurg 2025;194:123530. [Crossref] [PubMed]
39. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007;370:1453-7. [Crossref] [PubMed]
40. Buckland AJ, Proctor DJ. Minimally Invasive Transforaminal Lumbar Interbody Fusion with Expandable Cages. JBJS Essent Surg Tech 2023;13:e21.00062. [Crossref] [PubMed]
41. Yamagata T, Naito K, Arima H, et al. A minimum 2-year comparative study of autologous cancellous bone grafting versus beta-tricalcium phosphate in anterior cervical discectomy and fusion using a rectangular titanium stand-alone cage. Neurosurg Rev 2016;39:475-82. [Crossref] [PubMed]
42. Crawford AM, Striano BM, Bryan MR, et al. Expandable versus static transforaminal lumbar interbody fusion (TLIF) cages: comparing radiographic outcomes and complication profiles. Spine J 2025;25:237-43. [Crossref] [PubMed]
43. Lovecchio FC, Vaishnav AS, Steinhaus ME, et al. Does interbody cage lordosis impact actual segmental lordosis achieved in minimally invasive lumbar spine fusion?. Neurosurg Focus 2020;49:E17. [Crossref] [PubMed]
44. Lin GX, Kim JS, Kotheeranurak V, et al. Does the application of expandable cages in TLIF provide improved clinical and radiological results compared to static cages? A meta-analysis. Front Surg 2022;9:949938. [Crossref] [PubMed]
45. Gelfand Y, Benton J, De la Garza-Ramos R, et al. Effect of Cage Type on Short-Term Radiographic Outcomes in Transforaminal Lumbar Interbody Fusion. World Neurosurg 2020;141:e953-8. [Crossref] [PubMed]
46. Kim D, Jang HJ, Moon BJ, et al. Expandable Cage in Minimally Invasive Transforaminal Lumbar Interbody Fusion: Comparative Data with Static Cage from a Single Institution and a Single Surgeon. World Neurosurg 2025;202:124428. [Crossref] [PubMed]
47. Jitpakdee K, Sommer F, Gouveia E, et al. Expandable cages that expand both height and lordosis provide improved immediate effect on sagittal alignment and short-term clinical outcomes following minimally invasive transforaminal lumbar interbody fusion (MIS TLIF). J Spine Surg 2024;10:55-67. [Crossref] [PubMed]
48. Woodward J, Koro L, Richards D, et al. Expandable versus Static Transforaminal Lumbar Interbody Fusion Cages: 1-year Radiographic Parameters and Patient-Reported Outcomes. World Neurosurg 2022;159:e1-7. [Crossref] [PubMed]
49. Class 2 Device Recall Catalyft PL [Internet]. U.S. Food and Drug Administration [cited 2026 Feb 22]. Available online: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfres/res.cfm?id=216383