STUDY DESIGN: Retrospective radiographic evaluation of the sagittal alignment of the lumbar spine in patients undergoing short-segment instrumented posterior lumbar interbody fusion with cage systems of different shape. OBJECTIVES: To determine whether rectangular and wedge-shaped cages have a different influence on the sagittal alignment of the lumbar spine in patients undergoing short-segment instrumented posterior lumbar interbody fusion. SUMMARY OF BACKGROUND DATA: Previous studies of sagittal alignment after posterior lumbar interbody fusion have focused on the impact of intraoperative patient, hip, and knee positioning, as well as instrumentation characteristics on sagittal posture. The influence of the cage shape on indexes of total and segmental sagittal alignment of the lumbar spine is yet unknown. METHODS: Forty-two patients having undergone instrumented short-segment posterior lumbar interbody fusion were reviewed retrospectively. Twenty-two patients (12 women and 10 men, 38-78 years of age) had posterior lumbar interbody fusion with rectangular cages. The fused segments were: 4 at L3-L4, 16 at L4-L5, 11 at L5-S1. Thirteen patients had single- and nine patients double-level fusion. Twenty patients (8 women and 12 men, 34-81 years of age) had posterior lumbar interbody fusion with wedge-shaped cages. The fused segments were: 4 at L3-L4, 15 at L4-L5, 11 at L5-S1. Ten patients had single- and 10 patients double-level fusion. Cages were packed with cancellous bone from the posterior iliac crest and/or bone fragments harvested by laminectomy. All patients had additional pedicle screw fixation. Pre- and postoperative standing lateral radiographs were assessed for segmental and lumbar lordosis as well lumbar and sacral tilt. Data were analyzed with repeated measures analysis of variance. RESULTS The mean follow-up period was 18 months with a minimum follow-up period of 14 months. Mean segmental lordosis of the fused segments showed significant changes between the two implant groups (P < 0.05). Segmental lordosis decreased in the rectangular cage group from 10 degrees to 2 degrees at L3-L4, from 10 degrees to 5 degrees at L4-L5, and from 9 degrees before to 6 degrees after fusion surgery at L5-S1. In the wedge-shaped cage group, segmental lordosis increased from 4 degrees to 7 degrees at L3-L4, from 2 degrees to 8 degrees at L4-L5, and from 9 degrees to 17 degrees at L5-S1. Analysis of changes in lumbar lordosis and lumbar and sacral tilt did not show significant differences though opposite trends: lumbar lordosis decreased from 55 degrees to 48 degrees in the rectangular cage group and increased from 45 degrees to 53 degrees in the wedge-shaped cage group. Lumbar tilt measured 98 degrees before and 102 degrees after surgery in the rectangular cage group and 97 degrees before and 94 degrees after surgery. Sacral tilt measured 44 degrees before and 40 degrees after surgery in the rectangular cage group and measured 42 degrees before and 45 degrees after surgery in the wedge-shaped cage group. CONCLUSIONS: The cage geometry has a significant impact on the alignment of the lumbar spine after instrumented posterior lumbar interbody fusion. With rectangular cages, lumbar lordosis and segmental lordosis of the segments fused decrease; sagittal balance is maintained by compensatory changes of the sacral tilt. Wedge-shaped cages significantly increase segmental lordosis, enhance lumbar lordosis, and therefore should be preferred for restoring sagittal alignment in instrumented posterior lumbar interbody fusion procedures.