Technical evolution and short-term outcomes of minimally invasive versus conventional open anterior cervical discectomy and fusion: a systematic review and meta-analysis Authors: Zihua Ding, Jichen Liu, Haishan Xu, Yang Liu, Shaowei Xu, Zhen Wang Published: 2026 ABSTRACT AbstractBackgroundConventional open anterior cervical discectomy and fusion (ACDF) is an established treatment for degenerative cervical disease, but approach-related morbidity remains a concern. Minimally invasive anterior cervical techniques have evolved from microscope-assisted mini-open procedures to endoscope-assisted and selected full-endoscopic approaches. This study systematically reviewed the technical evolution of minimally invasive ACDF-related procedures and compared perioperative and short-term outcomes with conventional open ACDF.MethodsThis systematic review and meta-analysis was conducted in accordance with PRISMA. PubMed, Embase, Web of Science, and the Cochrane Library were searched. Comparative studies were included in the quantitative synthesis, whereas selected non-comparative anterior endoscopic reports were retained for qualitative contextualization of technical evolution. Risk of bias in comparative nonrandomized studies was assessed with ROBINS-I. Random-effects meta-analysis was used to pool mean differences (MDs) and odds ratios (ORs). Because pain outcomes were reported using non-uniform anatomical constructs, VAS data were summarized qualitatively rather than pooled. Studies that reported only category-specific adverse events without clear patient-level totals were excluded from the pooled analysis of overall complications to avoid double-counting.ResultsEight studies were included. Five comparative studies involving 347 patients were quantitatively synthesized, and three additional reports informed the qualitative analysis. Minimally invasive ACDF was associated with lower intraoperative blood loss than open ACDF (MD -21.93 mL, 95% CI -31.30 to -12.56; P < 0.00001; I² = 74%)). No significant between-group differences were observed for postoperative dysphagia (OR 0.71, 95% CI 0.21 to 2.43; P = 0.58), overall complications (OR 0.39, 95% CI 0.14 to 1.04; P = 0.06), length of stay (MD -0.55 days, 95% CI -1.20 to 0.11; P = 0.10), JOA (MD 0.48, 95% CI -0.05 to 1.01; P = 0.08), or NDI (MD -0.39, 95% CI -1.33 to 0.56; P = 0.42). Pain outcomes improved in both groups across individual studies but were not meta-analyzed because reported pain definitions differed across studies.ConclusionsMinimally invasive ACDF may reduce intraoperative blood loss while providing short-term neurological, disability, and complication outcomes broadly comparable to conventional open ACDF. Better-standardized comparative studies, particularly with harmonized pain definitions and blood-loss reporting, are needed.Trial registrationPROSPERO CRD42026001353049, retrospectively registered on 27 March 2026.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13018-026-06929-7.Keywords: Anterior cervical discectomy and fusion, Minimally invasive surgery, Endoscopic cervical surgery, Meta-analysis, Degenerative cervical disease Background Cervical degenerative disc disease is a common cause of cervical radiculopathy and myelopathy. Conventional open anterior cervical discectomy and fusion (ACDF) remains an established treatment because it provides reliable neural decompression and fusion stability [1–5]. However, standard open exposure may be associated with soft-tissue injury, postoperative dysphagia, recurrent laryngeal nerve palsy, and other approach-related morbidity [1–5]. Advances in optical systems and surgical instrumentation have expanded the spectrum of minimally invasive anterior cervical procedures. These techniques include microscope-assisted mini-open or tubular approaches, endoscope-assisted fusion procedures, and selected full-endoscopic anterior or transcorporeal strategies [6–8]. Their proposed advantages include smaller exposure, reduced tissue disruption, and improved visualization, but their comparative clinical benefits remain uncertain. Recent syntheses of cervical endoscopic surgery have often combined heterogeneous anterior and posterior procedures, including non-fusion techniques. To improve internal consistency, the present review focused on anterior fusion-based procedures while retaining selected anterior endoscopic reports only for qualitative contextualization of technical evolution. The aim of this study was to systematically review the evolution of minimally invasive ACDF-related procedures and compare perioperative and short-term outcomes with conventional open ACDF. Methods Aim, design and reporting standards This study was a systematic review and meta-analysis of published literature. It was conducted and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. The PRISMA flow diagram is presented in Fig. 1, and the completed PRISMA 2020 checklist is provided as Additional file 1. Reasons for exclusion at the screening and full-text stages are reported in Fig. 1. The review was retrospectively registered in PROSPERO (CRD42026001353049) on 27 March 2026 after completion of study screening, data extraction, and primary synthesis. PRISMA flow diagram of study selection with explicit exclusion reasons Information sources and search strategy A comprehensive search of PubMed, Embase, Web of Science, and the Cochrane Library was performed using combinations of terms related to minimally invasive ACDF, endoscopic ACDF, microendoscopic or microscope-assisted ACDF, anterior transcorporeal cervical decompression, and conventional open ACDF. Eligibility criteria For quantitative synthesis, we included comparative studies (randomized or observational) enrolling adult patients with degenerative cervical disease and reporting at least one extractable perioperative or clinical outcome after minimally invasive anterior cervical procedures versus open ACDF. Selected non-comparative anterior endoscopic studies were retained only for qualitative contextualization of technical evolution. Case reports, narrative reviews, posterior-only procedures, motion-preserving non-fusion procedures, and studies without extractable or interpretable outcome data were excluded from quantitative synthesis. For this review, conventional open ACDF included standard Smith-Robinson ACDF performed under direct vision or under operating-microscope magnification. In Zhong et al. [4], microscopic ACDF referred to an open anterior approach performed under microscope magnification through a limited incision in an air medium, whereas traditional ACDF referred to standard direct-vision open ACDF through a wider conventional exposure. Because both comparator arms used an open anterior approach with standard fusion constructs rather than an endoscopic working channel, they were treated as conventional or open comparator arms in the pooled analyses. Study selection and data extraction Study selection followed PRISMA principles and is summarized in Fig. 1. Extracted variables included sample size, operative time, blood loss, length of hospital stay (LOS), pain outcomes with the anatomical site or construct of the VAS when reported, Japanese Orthopaedic Association (JOA) score, Neck Disability Index (NDI), dysphagia, and overall complications. Pain outcomes were extracted exactly as reported, including unspecified VAS, separate neck and upper-extremity VAS, neck and/or upper-limb pain VAS, and combined neck/shoulder/upper-limb pain VAS. Final or last follow-up JOA and NDI values were preferentially extracted for quantitative synthesis when available [2–5]. Quality assessment Risk of bias in comparative studies was assessed with the Risk Of Bias In Non-randomized Studies - of Interventions (ROBINS-I) framework. The seven ROBINS-I domains evaluated bias due to confounding, participant selection, intervention classification, deviations from intended interventions, missing data, outcome measurement, and selection of the reported result. The results are summarized in Table 1. Table 1ROBINS-I assessment of comparative studies included in the quantitative synthesisStudyConfoundingParticipant selectionIntervention classificationDeviations from intended interventionsMissing dataOutcome measurementReported resultOverall ROBINS-IOda [1]SeriousSeriousLowLowLowModerateModerateSeriousLi [2]ModerateModerateLowLowLowLowModerateModerateWu [3]SeriousModerateLowLowLowLowModerateSeriousZhong [4]SeriousModerateLowLowLowLowModerateSeriousTang [5]SeriousModerateLowLowLowLowModerateSeriousJudgments: Low, Moderate, Serious. ROBINS-I assessment of comparative studies included in the quantitative synthesis Judgments: Low, Moderate, Serious. Statistical analysis Random-effects meta-analysis was performed. Continuous outcomes were synthesized as mean differences (MDs) with 95% confidence intervals (CIs), and dichotomous outcomes as odds ratios (ORs) with 95% CIs. Statistical heterogeneity was assessed using the I² statistic. For multi-arm comparisons, shared intervention groups were split according to standard meta-analytic methods. Final or last follow-up JOA and NDI values were preferentially pooled. Pain outcomes were not meta-analyzed because the included studies did not use a uniform pain construct: Wu reported VAS without a clearly specified pain site, Li assessed neck and/or upper-limb pain, Tang assessed neck, shoulder, and/or upper-limb pain, and Zhong reported separate neck pain and upper-extremity pain values [2–5]. Li et al. reported JOA as median (interquartile range) and was therefore excluded from the pooled JOA mean-difference analysis [2]. Oda et al. reported category-specific adverse events that could not be safely reconciled to patient-level overall complication counts and was therefore excluded from the pooled analysis of overall complications [1]. Because blood loss was the key differentiating perioperative outcome and heterogeneity was substantial, a post hoc leave-one-out sensitivity analysis was performed for that endpoint. Results Study characteristics and quality assessment Eight studies were included in the systematic review. Five comparative studies involving 347 patients contributed to the quantitative synthesis, whereas three additional reports informed the qualitative description of technical evolution and broader minimally invasive anterior cervical procedures. Reported follow-up ranged from immediate postoperative assessment to approximately 20.5 months. The screening process is summarized in Fig. 1. Techniques represented in the included literature comprised microendoscopic or microscope-assisted ACDF, endoscope-assisted ACDF, mini-open endoscope-assisted ACDF, selected full-endoscopic anterior or transcorporeal procedures, and conventional open ACDF. The characteristics of the comparative studies are summarized in Table 2. ROBINS-I judgments indicated overall serious risk of bias in four studies and moderate risk in one propensity score-matched study (Table 1). Table 2Characteristics of comparative studies included in the quantitative synthesisStudyDesignPopulation / diagnosisProcedure comparisonSample size / follow-upRole, key outcomes, and pain definitionOda [1]Retrospective observational cohortSingle-level degenerative cervical disease treated with ACDFMicroendoscopic ACDF vs. conventional open ACDF8 vs. 15; postoperative days 1, 2, and 7Quantitative: blood loss, LOS, dysphagia, complicationsLi [2]Retrospective propensity score-matched cohortSingle-segment cervical spondylotic myelopathyEndoscopy-assisted ACDF vs. open ACDF54 vs. 54; 24.2 months (15–38)Quantitative: blood loss, LOS, dysphagia, complications, JOA, NDI; pain outcome reported as VAS for neck and/or upper-limb pain (qualitative only)Wu [3]Retrospective comparative cohortSingle-segment cervical spondylotic myelopathyEndoscope-assisted ACDF vs. open ACDF28 vs. 24; 13.0 months (12–17)Quantitative: blood loss, LOS, dysphagia, complications, JOA; VAS reported without clearly specified pain site (qualitative only)Zhong [4]Retrospective three-arm comparative cohortCervical spondylotic myelopathy with posterior vertebral osteophyte or OPLLMini-open endoscope-assisted ACDF vs. microscopic ACDF and traditional ACDF11 vs. 11 vs. 11; approximately 20.3 monthsQuantitative: LOS, dysphagia, complications, JOA, NDI; separate neck and upper-extremity VAS values summarized qualitatively onlyTang [5]Retrospective comparative cohortCervical spondylotic myelopathyMini-open endoscopy-assisted ACDF vs. conventional ACDF61 vs. 70; 13.2 vs. 14.8 monthsQuantitative: JOA, NDI; VAS for neck, shoulder, and/or upper-limb pain summarized qualitatively only; LOS not pooled Characteristics of comparative studies included in the quantitative synthesis Technical evolution of minimally invasive techniques The technical evolution of minimally invasive anterior cervical surgery can be viewed in three broad phases. Phase I comprised microscope-assisted or tubular ACDF techniques that reduced exposure while preserving standard fusion constructs. Phase II included endoscope-assisted fusion procedures, which improved illumination and magnification within less extensive soft-tissue dissection. Phase III extended minimal-access decompression toward selected full-endoscopic anterior and transcorporeal strategies that emphasized structural preservation. Overall, this evolution reflects a shift from simple exposure reduction toward more refined tissue preservation and biomechanical consideration. Examples of these phases are illustrated by Vergara 2018, Morgenstern 2024, and Ma 2022 [6–8], and parallel broader innovations in anterior cervical access, implants, and perioperative management [9]. Perioperative outcomes Intraoperative blood loss. Pooled analysis of three studies showed that minimally invasive ACDF was associated with significantly lower intraoperative blood loss than open ACDF (MD = − 21.93 mL, 95% CI − 31.30 to − 12.56; P < 0.00001; I²= 74%; Fig. 2). In a post hoc leave-one-out sensitivity analysis, exclusion of Oda 2025 eliminated heterogeneity and preserved the direction of effect (MD = − 25.38 mL, 95% CI − 28.00 to − 22.75; I²= 0%), suggesting that Oda 2025 was the main source of between-study inconsistency. Forest plot comparing intraoperative blood loss between minimally invasive ACDF and conventional open ACDF Length of hospital stay. Pooled analysis of four comparisons showed no statistically significant difference in length of stay between minimally invasive ACDF and open ACDF (MD = − 0.55 days, 95% CI − 1.20 to 0.11; P = 0.10; I² = 43%; Fig. 3). Forest plot comparing length of stay between minimally invasive ACDF and conventional open ACDF Clinical and functional outcomes Pain outcomes. Pain outcomes were summarized qualitatively only because the included studies used non-uniform constructs. Wu 2024 reported an unspecified VAS, Li 2025 assessed neck and/or upper-limb pain, Tang 2024 assessed neck, shoulder, and/or upper-limb pain, and Zhong 2025 reported separate neck and upper-extremity VAS values. Across these studies, both minimally invasive and open ACDF improved pain from baseline, and some studies reported earlier postoperative pain improvement in the minimally invasive group [2–5]. JOA. Pooled analysis of four comparisons showed no statistically significant difference in JOA score, although the pooled estimate numerically favored minimally invasive ACDF (MD = 0.48, 95% CI -0.05 to 1.01; P = 0.08; I²= 80%; Fig. 4B). Individual studies by Wu 2024 and Tang 2024 reported better postoperative JOA scores in the minimally invasive group, whereas Zhong 2025 showed similar final scores between groups; Li 2025 also reported greater early postoperative JOA improvement in the endoscopy-assisted group but could not be pooled because the outcome was reported as median (interquartile range) [2–5]. Pain and functional outcomes after minimally invasive ACDF versus conventional open ACDF. (A) Pain outcomes were not pooled because VAS definitions were clinically heterogeneous across studies. (B) Japanese Orthopaedic Association (JOA) score; (C) Neck Disability Index (NDI) NDI. Pooled analysis of four comparisons showed no significant difference in NDI between minimally invasive ACDF and open ACDF (MD = -0.39, 95% CI -1.33 to 0.56; P = 0.42; I²= 0%; Fig. 4C). Complications Postoperative dysphagia. Pooled analysis of five comparisons showed no significant difference in postoperative dysphagia between minimally invasive ACDF and open ACDF (OR = 0.71, 95% CI 0.21 to 2.43; P = 0.58; I² = 0%; Fig. 5A), although the point estimate numerically favored minimally invasive ACDF. Forest plots comparing postoperative dysphagia (A) and overall complications (B) between minimally invasive ACDF and conventional open ACDF Overall complications. For overall complications, Oda 2025 was excluded from the pooled analysis to avoid potential double-counting of category-specific adverse events. After exclusion, no statistically significant difference was found between minimally invasive ACDF and open ACDF, but the pooled estimate numerically favored minimally invasive ACDF (OR = 0.39, 95% CI 0.14 to 1.04; P = 0.06; I² = 0%; Fig. 5B). Neurological injury and reoperation. Neurological injury and reoperation were infrequently reported and were summarized qualitatively only. Summary of pooled findings A summary of pooled quantitative findings is provided in Table 3. Table 3Summary of pooled quantitative findingsOutcomeComparisonsEffect measurePooled estimate95% CIP value I²InterpretationIntraoperative blood loss3MD-21.93 mL-31.30 to -12.56< 0.0000174%Significantly lower in minimally invasive ACDFPostoperative dysphagia5OR0.710.21 to 2.430.580%No significant difference; point estimate favored minimally invasive ACDFOverall complications*4OR0.390.14 to 1.040.060%No significant difference; numerical trend favoring minimally invasive ACDFLength of stay4MD-0.55 days-1.20 to 0.110.1043%No significant differenceJOA4MD0.48-0.05 to 1.010.0880%No significant difference; numerical trend favoring minimally invasive ACDFNDI4MD-0.39-1.33 to 0.560.420%No significant difference*Oda 2025 was excluded from the pooled analysis of overall complications to avoid potential double-counting of category-specific adverse events. Summary of pooled quantitative findings *Oda 2025 was excluded from the pooled analysis of overall complications to avoid potential double-counting of category-specific adverse events. Discussion Principal findings The principal finding of this review is that minimally invasive ACDF was associated with significantly lower intraoperative blood loss than conventional open ACDF. By contrast, pooled analyses did not demonstrate statistically significant differences in postoperative dysphagia, overall complications, length of stay, JOA, or NDI. The pooled estimates for JOA and overall complications numerically favored minimally invasive ACDF, but neither reached statistical significance. Pain outcomes were reported inconsistently and were therefore summarized qualitatively rather than pooled. Taken together, these findings suggest that the most consistent advantage of minimally invasive anterior cervical techniques lies in reduced operative exposure rather than clear superiority across short-term clinical outcomes. Technical context and interpretation The included literature illustrates a stepwise evolution from microscope-assisted or mini-open exposure-reduction strategies toward endoscope-assisted fusion and, ultimately, selected full-endoscopic anterior or transcorporeal procedures [6–8]. This progression mirrors broader ACDF innovation, in which smaller access corridors, improved optical platforms, and refined implant and hemostatic strategies have been developed to reduce approach-related morbidity while preserving decompression quality [9]. Earlier techniques largely preserved the conventional fusion construct while attempting to reduce soft-tissue disruption; later approaches pursued progressively smaller working corridors and greater structural preservation. Although full-endoscopic and transcorporeal strategies are conceptually attractive, the comparative evidence base remains limited, heterogeneous, and dominated by short-term observational studies [1–9]. The reduction in blood loss is biologically plausible and likely reflects less soft-tissue dissection and a narrower surgical corridor. However, heterogeneity for this outcome was substantial and appears to be driven largely by Oda 2025, in which absolute blood loss was very low in both study arms (19.2 vs. 22.5 mL) and the between-group difference was small. By contrast, Li 2025 and Wu 2024 reported larger between-group differences. Several factors may explain this pattern. First, blood-loss ascertainment was not uniform: Li 2025 estimated blood loss from perioperative hematocrit change, whereas Oda 2025 and Wu 2024 reported directly observed intraoperative values. Second, the operative strategies were not identical: Oda evaluated a microendoscopic single-level technique developed specifically to minimize retraction manipulation and postoperative prevertebral swelling, whereas Li and Wu used endoscope-assisted ACDF through a mini-open corridor. Third, perioperative management details that can influence apparent blood loss, such as hemostatic adjuncts, drain use, and the handling of hidden blood loss, were not reported consistently across studies [9]. Reassuringly, the leave-one-out analysis showed that omitting Oda 2025 eliminated heterogeneity while preserving the direction and statistical significance of the pooled effect, suggesting that the blood-loss advantage is directionally robust but sensitive to study-specific measurement and technique. Functional and safety outcomes were otherwise broadly comparable. NDI showed no between-group difference and no meaningful heterogeneity. JOA numerically favored minimally invasive ACDF, and this signal was consistent with the individual findings of Wu 2024 and Tang 2024, while Zhong 2025 showed comparable final JOA scores across groups and Li 2025 reported earlier postoperative improvement that could not be pooled because JOA was reported as median (interquartile range) [2–5]. This pattern suggests that minimally invasive ACDF may facilitate earlier neurological recovery in some cohorts, but the current comparative evidence is too small and heterogeneous to support a definitive advantage at final follow-up. For complications, the pooled OR for overall adverse events also numerically favored minimally invasive ACDF, although confidence intervals remained wide because event counts were low. Dysphagia showed no significant between-group difference, which is not surprising given that postoperative dysphagia after ACDF is influenced by retraction, implant profile, prevertebral swelling, number of levels, and, importantly, by how the outcome is measured and at what time point it is assessed [1, 10, 11]. Broader ACDF literature likewise shows substantial variation in dysphagia estimates across studies, underscoring the need for standardized complication definitions and assessment schedules [10, 11]. Pain location also matters clinically: neck pain and arm pain reflect partly different mechanisms, and the included studies variably reported an unspecified VAS, combined neck and/or upper-limb pain, combined neck/shoulder/upper-limb pain, or separate neck and upper-extremity VAS scores. For that reason, we considered a pooled pain estimate potentially misleading and retained these data only for qualitative interpretation. Learning curve and implementation considerations Endoscopic and mini-open anterior cervical procedures require advanced hand-eye coordination, familiarity with a constrained working corridor, and careful case selection [3–8]. As a result, the learning curve may attenuate some of the expected perioperative benefits in early experience. Standardized training pathways and reporting of surgeon experience would improve interpretation of future comparative studies. Clinical implications The available evidence suggests that minimally invasive ACDF can achieve short-term neurological, disability, and complication outcomes broadly comparable to conventional ACDF while reducing intraoperative blood loss. In current practice, these techniques may be most attractive when minimizing exposure-related tissue disruption is prioritized and when adequate technical expertise is available. Limitations Several limitations should be acknowledged. First, the comparative evidence base was entirely nonrandomized, and ROBINS-I identified serious overall risk of bias in four of five comparative studies. Second, substantial surgical heterogeneity remained across included studies, particularly with respect to access strategy, optical platform, and technical scope. Third, blood-loss ascertainment and perioperative haemostatic practice were not standardized across studies, which likely contributed to heterogeneity for the key pooled outcome. Fourth, pain outcomes were reported using non-uniform constructs (unspecified VAS, separate neck and upper-extremity pain VAS, neck and/or upper-limb pain VAS, or combined neck/shoulder/upper-limb pain VAS), so they could not be synthesized quantitatively. Fifth, follow-up duration was limited and inconsistently reported, restricting assessment of longer-term outcomes and biomechanical consequences. Sixth, the small number of studies contributing to each pooled outcome limited the robustness of heterogeneity assessment and precluded meaningful evaluation of publication bias. Finally, the review was retrospectively registered in PROSPERO after data extraction and synthesis, which represents a methodological limitation of the review process. Future directions Future work should emphasize higher-quality comparative designs, standardized patient-level complication reporting, harmonized blood-loss definitions, explicit separation of axial neck pain from radicular arm pain, and consistent follow-up time points for functional outcomes such as JOA and NDI. Clearer differentiation between fusion-based minimally invasive ACDF techniques and motion-preserving anterior endoscopic procedures would further improve clinical interpretation and the reliability of future pooled analyses. Conclusions Minimally invasive ACDF appears to provide short-term neurological, disability, and complication outcomes broadly comparable to conventional open ACDF while significantly reducing intraoperative blood loss. Current evidence does not demonstrate statistically significant differences in dysphagia, overall complications, length of stay, JOA, or NDI, although the pooled estimates for JOA and overall complications numerically favored minimally invasive ACDF. Because pain outcomes were reported inconsistently across studies, they were summarized qualitatively rather than meta-analyzed. Higher-quality prospective comparative studies with standardized blood-loss and pain reporting are required before stronger conclusions regarding superiority can be made. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements Not applicable. Abbreviations Anterior cervical discectomy and fusion Confidence interval Interquartile range Japanese Orthopaedic Association Length of stay Mean difference Neck Disability Index Ossification of the posterior longitudinal ligament Odds ratio Preferred Reporting Items for Systematic Reviews and Meta-Analyses Risk Of Bias In Non-randomized Studies - of Interventions Visual Analog Scale Author contributions ZD, JL, HX, YL, SX, and ZW contributed to the conception and design of the study, literature screening, data extraction, data analysis, interpretation of the findings, drafting of the manuscript, and critical revision of the manuscript. All authors read and approved the final manuscript. Funding This study was supported by the Key R&D Projects in Ningxia Hui Autonomous Region, China (No. 2022BEG03086). Data availability All data generated or analysed during this study are included in this published article and its supplementary information files. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement All data generated or analysed during this study are included in this published article and its supplementary information files.