Clinical and radiologic outcomes of expandable versus static cages in biportal endoscopic TLIF: focus on endplate injury and subsidence

BACKGROUND: Expandable cages (ECs) theoretically enhance disc height restoration in biportal endoscopic transforaminal lumbar interbody fusion (BE-TLIF), but concerns persist regarding endplate injury and early subsidence. This study compared the safety and efficacy of ECs versus static cages (SCs) in BE-TLIF. METHODS: We retrospectively analyzed 73 BE-TLIF levels (37 ECs, 36 SCs) performed by a single surgeon between May 2023 and May 2025. BE-TLIF allowed direct endplate visualization and routine bilateral facet release. SCs were banana-shaped PEEK or 3D porous titanium cages inserted in the largest feasible size and rotated into position, whereas ECs were bullet-shaped titanium cages expanding only in height. Radiographic outcomes included endplate injury on postoperative CT, subsidence at 6 weeks and 3 months, disc height change, and segmental lordosis/PI-LL. Clinical outcomes were VAS (back/leg), ODI, and FRI. RESULTS: Despite lower BMI and higher BMD in the EC group, subsidence was significantly more frequent with ECs. ECs showed higher endplate injury (18.9% vs. 2.8%, p = 0.030) and greater disc height loss at 6 weeks (- 0.96 ± 1.89 vs. - 0.18 ± 1.37 mm, p = 0.047). Clinically significant (> 2 mm) and severe (> 25% height) subsidence occurred more often with ECs (35.1% vs. 11.1%, p = 0.032; 18.9% vs. 2.8%, p = 0.030). All patients with endplate injury developed subsidence. No new subsidence appeared after 6 weeks. Although ECs achieved greater immediate disc height restoration (4.07 ± 2.28 vs. 2.65 ± 2.13 mm, p = 0.009), they yielded less pain improvement (VAS back: p = 0.004 at 6 weeks, p = 0.029 at 3 months; VAS leg: p = 0.001 at 6 weeks, p = 0.002 at 3 months), while ODI and FRI were comparable. Segmental lordosis and PI-LL correction did not differ. CONCLUSIONS: In this short-term analysis of BE-TLIF, ECs are associated with higher risks of endplate injury and early subsidence, leading to inferior early pain outcomes despite superior initial disc height restoration. These drawbacks likely stem from titanium composition, small footprint, and concentrated expansion forces on the central endplate. These findings suggest that, despite the technical advantages of BE-TLIF regarding endplate preparation and bilateral facet release, current EC designs may possess inherent limitations, underscoring the need for cautious use in osteoporotic patients and further investigation into long-term clinical durability. TRIAL REGISTRATION: Retrospectively registered.