Continuous local antibiotic perfusion for spinal infection: a systematic review of reported cases

Introduction

Spinal infections form a broad spectrum, including pyogenic spondylodiscitis, vertebral osteomyelitis, spinal epidural abscess, paraspinal and psoas abscesses, and postoperative surgical site infection (SSI). Recent population-based studies report rising incidences of pyogenic spondylodiscitis and vertebral osteomyelitis (1,2). Delayed treatment is associated with death, systemic complications, and neurological deficits (3,4). Management is often prolonged, and many cases require reoperation and extended antibiotic therapy (5,6).

Among these entities, SSI after instrumented spine surgery is particularly challenging. Removing implants risks postoperative instability and often necessitates complex reconstruction; therefore, the central goal is eradication of infection while retaining the hardware (7). To preserve implants, surgeons have used repeated debridement, systemic and local antimicrobial therapy, negative-pressure wound therapy (NPWT), and irrigation or perfusion systems (8,9). Outcomes with these conventional strategies are inconsistent. In periprosthetic joint infection (PJI), the failure rate of debridement, antibiotics, and implant retention (DAIR) ranges from 0% to 84.4%, with a pooled estimate of 35.9% [95% confidence interval (CI): 23.9–48.0%] (10). A narrative review similarly reported DAIR success rates from 11.1% to 100%, attributing heterogeneity to differences in indications, patient selection, surgical techniques, and antimicrobial regimens (11). Spinal SSI shows comparable variability: in a single-center retrospective cohort of 176 cases treated with debridement plus irrigation, the reoperation rate was 35.2% (15.9% within 90 days) and the reinfection rate was 16.4% (12).

Continuous local antibiotic perfusion (CLAP) is a recently introduced technique (first reported in 2021). It delivers high-concentration antibiotics to the infected site through dual-lumen drains while NPWT continuously removes effluent. Because eradication of biofilm-embedded bacteria may require antibiotic exposures 100–1,000 times the minimum inhibitory concentration (MIC), CLAP aims to maintain levels above the minimum biofilm eradication concentration (MBEC) (13). Early orthopaedic experience is encouraging. A recent systematic review with pooled individual participant data (IPD) across orthopaedic infections, including 81 observational studies and 256 CLAP-treated patients, reported a 30-day early response in 59.1% of evaluable patients and durable infection control in 88.4% at ≥6 months and 90.2% at ≥12 months (14). In fracture-related infection (FRI), implant retention was achieved in 35/40 cases (87.5%) and 9/9 cases (100.0%) across two series (15,16). In spinal infections, CLAP has been reported for vertebral osteomyelitis, spinal epidural abscess, and psoas abscess, with favorable outcomes (17,18). Small spinal SSI series also suggest benefit, with implant retention in all 4/4 cases and marked improvements in white blood cell count and C-reactive protein within 1 week of CLAP initiation (19).

Reflecting this growing clinical interest, the orthopaedic literature on CLAP has expanded rapidly in recent years, with successive reports refining the technique for FRI (15,20) and, more recently, synthesizing outcomes across a broader range of orthopaedic infections (14). Within this accumulating body of evidence, spinal infection is a distinct and important setting: implant retention is often clinically essential to avoid the morbidity of revision reconstruction, yet the consequences of failed infection control—including epidural extension, neurological compromise, and mechanical instability—differ substantially from those in extremity infections. Despite this, the spinal CLAP literature remains dispersed across single-center case reports and small series with heterogeneous protocols and outcome definitions, and a structured synthesis specific to the spine has not, to our knowledge, been undertaken.

However, no systematic review has yet synthesized the evidence for this approach. Therefore, this review aims to identify the reported cases of CLAP for spinal infection and to describe, in a systematic and descriptive manner, how the technique has been used and the outcomes that have been reported. We present this article in accordance with the PRISMA reporting checklist (available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0156/rc) (21).

Methods

Protocol and reporting standards

The protocol was registered on the Open Science Framework (OSF; https://osf.io/rk4g6/; registered November 24, 2025). A narrative synthesis was prespecified as the primary methodology. A study-selection flow diagram and the checklist are provided in Appendix 1, and the most recent protocol and materials are available on the OSF registration page.

Eligibility criteria

Study design: case reports and case series were eligible regardless of language or country. Conference abstracts without an accessible full text and studies with unverifiable or unavailable data were excluded.

Setting: any clinical setting.

Population: adults or children with spinal infection. Animal or in vitro studies without clinical data were excluded.

Intervention: CLAP—including intramedullary antibiotic perfusion (iMAP)/intra-soft-tissue antibiotic perfusion (iSAP) variants and continuous local antibiotic infusion (CLAI)/continuous local antibiotic administration (CLAA)—used as a primary or adjunctive treatment; concomitant use of NPWT was allowed but not required.

• iMAP: continuous perfusion of high-concentration antibiotic into cancellous bone via an intramedullary/bony channel using an inflow catheter and a separate outflow/drain (20).
• iSAP: continuous perfusion within infected soft tissues/compartments via catheter(s) to achieve sustained local antibiotic levels with controlled egress (16).
• CLAI/CLAA: earlier/related continuous local delivery methods using single- or dual-lumen drains and an external pump to infuse antibiotics directly into the infected bed with passive or suction-assisted outflow; inclusion was permitted when used with or without NPWT.

Outcomes and definitions

The primary outcome was the reoperation- and mortality-free proportion after CLAP, defined as the proportion of patients with no need for additional (unplanned) surgery and no death after CLAP. This is a pragmatic composite surrogate, adopted because clinically detailed outcomes such as microbiological eradication were not uniformly reported across the included reports; it does not equate to definitive microbiological or clinical cure, and it may misclassify patients with persistent infection managed non-operatively, or those lost to follow-up, as treatment successes.

Secondary outcomes included the time to reported clinical stabilization, recurrence, toxicity attributable to high-concentration antibiotics (renal and hepatic dysfunction), device-related events, and reoperation.

Information sources and search strategy

We searched MEDLINE (PubMed), CENTRAL (Cochrane Library), Embase (Dialog), International Clinical Trials Registry Platform (ICTRP), and ClinicalTrials.gov from database inception through October 1, 2025, without language or publication-status restrictions. Trial registries [ClinicalTrials.gov, World Health Organization (WHO) ICTRP] were screened for ongoing or completed but unpublished studies. Reference lists, citation indices, and gray literature (Google Scholar, ResearchGate) were scanned. When access was restricted or eligibility was unclear, study authors were contacted for clarification and additional information. Full search strategies are provided in Appendix 1.

Two independent reviewers screened titles and abstracts using ASReview, an open-source tool for active learning-based screening that prioritizes potentially relevant records based on previous screening decisions (22). The stopping point for screening followed the SAFE procedure (23), a practical and conservative heuristic for determining when to end the active learning process.

The SAFE procedure consists of three phases:

• Phase 1: initial random screening—1% of the dataset was randomly screened to estimate the proportion of relevant records. In this review, a 1% random sample (76 of the 7,675 records) was screened, of which 6 were relevant.
• Phase 2: active learning screening—screening continued until twice the number of relevant records identified in phase 1 had been retrieved; here, screening continued until 12 relevant records (twice the 6 identified in phase 1) had been identified.
• Phase 3: final deep learning phase—the original SAFE procedure does not specify a strict stopping criterion for this phase; in this protocol, screening stopped once the number of consecutive irrelevant records exceeded the greater of 25 or 1% of the dataset. In this review, screening stopped after 100 consecutive irrelevant records, exceeding this pre-specified minimum. Records identified as potentially eligible were carried forward to full-text assessment, and the remaining records ranked ineligible by the model were excluded by the automation tool (Figure 1).

Figure 1 PRISMA flow diagram. ICTRP, International Clinical Trials Registry Platform.

After the title and abstract screening, we assessed the eligibility of the remaining articles based on their full texts. We contacted original authors if relevant data were missing. Disagreements between the two reviewers were resolved by discussion, and if this failed, a third reviewer acted as an arbiter.

Study selection

Two reviewers independently screened titles/abstracts and assessed full texts against the eligibility criteria. Disagreements were resolved by consensus or adjudication by a third reviewer. Studies reporting patients treated with CLAP for spinal infection were included. For mixed cohorts (e.g., including extremity infections), only spine-eligible patients were considered at the data-extraction stage. When multiple reports appeared to describe overlapping patients or cohorts, we included the report with the most complete CLAP and outcome data, and used the others only for supplementary details. To ensure transparency, a systematic patient-overlap audit was performed across all candidate reports by center, author group, treatment period, and patient-level characteristics (age, sex, causative organism, spinal level, and CLAP protocol); where the same patient appeared in more than one article, that patient was counted only once. A study-level accounting of the patients contributed by, and excluded from, each report is provided in Table S1.

Data extraction and management

Using a piloted extraction form, two reviewers independently collected data. At the study level, we recorded country, center, year, design, sample size, and follow-up length. At the participant level, we first captured baseline characteristics: activities of daily living (ADL) at admission, smoking history, hemoglobin A1c (HbA1c), dialysis status [hemodialysis (HD)/peritoneal dialysis (PD)], rheumatoid arthritis, and use of corticosteroids or other immunosuppressants. Comorbidity burden was quantified with the updated Charlson Comorbidity Index (CCI) and the number of comorbid conditions.

The infection profile was defined with harmonized criteria: index diagnosis (pyogenic spondylodiscitis, vertebral osteomyelitis and SSI with or without implants), involved spinal region (cervical, thoracic, lumbar, sacral, cervicothoracic, thoracolumbar), single- vs. multilevel disease, anatomic extent of infection (anterior, posterior, or circumferential), epidural abscess presence, paraspinal/psoas abscess presence, depth (superficial or deep), and timing of onset (acute ≤30 days, chronic >30 days) (24,25). Microbiology included the site(s) of positive culture (local and/or blood) and identified pathogens [methicillin-susceptible Staphylococcus aureus (MSSA)/methicillin-resistant Staphylococcus aureus (MRSA), gram-negative rods, anaerobes, fungi, Mycobacterium tuberculosis, polymicrobial, or others].

Initial treatment variables included source control (none, percutaneous drainage/aspiration, open debridement/drainage, decompression for epidural abscess) and approach (none, anterior/anterolateral, posterior, circumferential), baseline stabilization (none, posterior instrumentation/fusion, anterior reconstruction/fusion, circumferential reconstruction/fusion), management of existing instrumentation (no pre-existing instrumentation, complete retention, partial exchange, partial removal, complete removal, extension), new stabilization performed (none, posterior instrumentation/fusion, anterior reconstruction/fusion, circumferential reconstruction/fusion), dural exposure, implant loosening, osteolysis. Systemic antibiotic therapy was recorded as use (yes/no), agents, and total duration.

For CLAP, we documented the interval from diagnosis to initiation, any surgery performed before CLAP, and use of NPWT. We standardized reporting of modality (iMAP/iSAP/CLAI/CLAA), antibiotic agent, concentration, flow rate, and treatment duration (days). Outcomes included the reoperation- and mortality-free status and the time to reported clinical stabilization, recurrence, toxicity attributable to high-concentration antibiotics (renal and hepatic dysfunction), device-related events, ADL at discharge, reoperation, and follow-up length.

Extraction prioritized explicit definitions of the primary outcome and implant preservation. Quality checks covered range and plausibility, internal temporal logic, cross-variable consistency, completeness, and reconciliation against published summaries; disagreements were resolved through discussion and, when needed, author queries. Data were stored on a secure, access-restricted drive. When corresponding authors provided de-identified IPD, we used these to verify eligibility and outcome timing. No meta-analysis was planned; findings were synthesized qualitatively.

Risk of bias assessment

Methodological quality of case reports and case series was appraised using the Joanna Briggs Institute (JBI) Critical Appraisal Tools for case reports/series (26). Risk-of-bias judgments considered clarity of case definition, ascertainment of exposure/intervention, completeness of follow-up, and objectivity of outcome assessment. For each included study, we calculated the quality score as the number of “yes” responses divided by the number of applicable items, expressed as a percentage.

Data synthesis (narrative)

We present means and standard deviations for continuous variables, and report counts and percentages (n, %) for categorical variables related to patient characteristics and outcomes. For age, the mean and standard deviation were calculated based on the methods described in the Cochrane Handbook. Given heterogeneity in case mix, CLAP protocols, and outcome definitions, no quantitative meta-analysis was conducted. This decision was prespecified and reflects established systematic-review methodology: when clinical, methodological, and outcome heterogeneity make statistical pooling inappropriate, a structured narrative synthesis—rather than meta-analysis—is the recommended approach, and we report it in accordance with the Synthesis Without Meta-Analysis (SWiM) reporting guideline (27) and the Cochrane Handbook. The absence of a meta-analysis therefore reflects the nature of the available evidence rather than a shortcoming of the review methodology.

We performed a structured narrative synthesis, grouping studies by infection phase (acute ≤30 days and chronic >30 days), spinal instrumentation, and surgical level (cervical vs. thoracolumbar). During the preparation of this manuscript, the authors used a generative artificial intelligence (AI) tool (ChatGPT, OpenAI) solely to improve the language and readability of the text. No AI tool was used for study design, data collection, analysis, or interpretation. All AI-assisted output was subsequently reviewed and edited by the authors, who take full responsibility for the content of the manuscript.

Results

Study selection and IPD obtained

The PRISMA flow diagram is shown in Figure 1. A total of 7,675 records were identified through database and register searches. After 7,663 records marked as ineligible by automation tools were removed, the remaining records were screened, and nine studies were ultimately included in the review (17,18,28-34). These comprised a total of 48 patients treated with CLAP for spinal infections. Of the included studies, 7 were case reports contributing 7 patients, and 2 were case series contributing 41 patients. We contacted all reachable corresponding authors to request IPD and obtained previously unpublished patient-level data for 4 patients from 4 authors. A study-level accounting of every included study—the number of patients reported in the source, the number included in this review, and the reason for any difference—is provided in Table S1. In particular, one further report (Takahara et al., 2024) described the same spinal CLAP patient as Kitano et al. (29); the Kitano report, a case series that also included non-spinal CLAP cases, was adopted and the Takahara report was not counted separately, yielding 48 unique patients across 9 studies.

Study and baseline characteristics

The mean age was 69.8±11.9 years in 48 patients, and 33 of 47 patients (70.2%) were female (Table 1). Diabetes mellitus was reported in 21 of 48 patients (43.8%). The most common infection type was SSI with instrumentation (45/48, 93.8%), followed by vertebral osteomyelitis (2/48, 4.2%) and pyogenic spondylodiscitis (1/48, 2.1%). The thoracolumbar region was the most frequently involved site (33/48, 68.8%). All patients had multilevel involvement. Acute infection was reported in 28 of 48 patients (58.3%), and chronic infection in 20 of 48 (41.7%). Among 39 patients with available microbiological data, MSSA/MRSA was the most common pathogen group (17/39, 43.6%), followed by others (15/39, 38.5%). Microbiological data were missing for 9 patients (9/48, 18.8%).

Risk of bias within studies

Risk of bias was assessed using the JBI Critical Appraisal Checklists for case reports and case series (Tables 2,3). Overall, the methodological quality of the included studies was moderate to high. Items most frequently judged as unclear were reporting of adverse events in the case reports and consecutive inclusion, complete inclusion, and reporting of presenting site(s)/clinic(s) in the case series.

Treatment characteristics and clinical outcomes

Among the 8 patients with procedure-level data, source control consisted of open debridement/drainage in 6/8 (75.0%) and decompression for epidural abscess in 2/8 (25.0%) (Table 4). The surgical approach was posterior in 6/8 (75.0%). Baseline stabilization was absent in 4/48 patients (8.3%), whereas 32/48 (66.7%) had posterior instrumentation/fusion and 12/48 (25.0%) had circumferential reconstruction/fusion.

Among 45 patients with available data on management of existing instrumentation, complete retention was achieved in 23/45 (51.1%) and complete removal in 16/45 (35.6%); partial removal and extension were each reported in 1/45 (2.2%), and 4/45 (8.9%) had no pre-existing instrumentation (Table 4). Among 8 patients with available data, no new stabilization procedure was performed in 5/8 (62.5%), whereas circumferential reconstruction/fusion was performed in 3/8 (37.5%). Dural exposure, implant loosening, and osteolysis were infrequently reported (1/5, 20.0%; 1/4, 25.0%; and 1/3, 33.3%, respectively).

All 48 patients received systemic antibiotic therapy. CLAP was delivered as iSAP in all cases, and gentamicin was the perfused antibiotic in all 48 patients. Reported gentamicin concentration was 1.2 mg/mL in 45/46 cases (97.8%) and 0.4 mg/mL in 1/46 cases (2.2%). The reported flow rate was 2 mL/h in 42/43 cases (97.7%) and 1 mL/h in 1/43 cases (2.3%). The mean duration of CLAP treatment was 16.1±9.3 days.

Clinical outcomes are summarized in Table 5. The mean interval from diagnosis to CLAP initiation was 47.9±14.9 days. Any surgery before CLAP was reported in 6/8 patients (75.0%), and adjunctive NPWT was used in 46/48 patients (95.8%). No patients were treated with NPWT alone without concomitant CLAP.

Primary and secondary outcomes

Using our prespecified pragmatic definition, a reoperation- and mortality-free course was observed in 36/48 patients (75.0%) (Table 5). Among 7 patients with available data, the mean time to reported clinical stabilization was 28.5±18.7 days. Recurrence data were available for 8 patients, among whom recurrence was reported in 1 patient (12.5%). No toxicity attributable to high-concentration antibiotics, including renal or hepatic dysfunction, was reported among the 8 patients with available safety data, and no device-related events were reported. Reoperation after CLAP was reported in 12/48 patients (25.0%). The mean follow-up length in the studies reporting this outcome was only 7.5±4.3 months; given this short follow-up and the surrogate nature of the primary outcome, the 75.0% figure should be interpreted with caution.

Outcomes by clinical stratum (descriptive)

Because of the small number of patients and the heterogeneity of the included reports, we did not perform inferential subgroup testing. Instead, outcomes are summarized descriptively according to infection phase, presence of spinal instrumentation, and surgical level; no clear differences were apparent across these strata, but the numbers were too small to support any statistical inference (Table 6).

Discussion

Summary of findings

In this systematic review of 9 studies including 48 patients, CLAP was used mainly for postoperative spinal SSI with instrumentation and was applied as part of multimodal treatment, usually together with systemic antibiotics, surgical source control, and NPWT. A reoperation- and mortality-free course was recorded in 75.0% of patients, and reoperation after CLAP was reported in 25.0%. Among cases with available data, complete implant retention was achieved in 51.1%, whereas complete implant removal was required in 35.6%. No treatment-related toxicity or device-related adverse events were reported in the limited safety data available.

Interpretation in the context of previous studies

Previous studies on postoperative spinal SSI and implant-related infection have reported persistent challenges in achieving infection control while preserving instrumentation, together with substantial variability in treatment outcomes (7,35-37). In a systematic review, Hersh et al. found marked heterogeneity in the literature regarding implant management in postoperative spine infection, with some studies supporting implant retention, whereas others favored implant removal in late-onset infection (7). Similarly, Karczewski et al. emphasized that the available evidence on postoperative spinal implant infection is limited by inconsistent definitions, heterogeneous patient populations, and the lack of standardized treatment strategies (35). Taken together, these findings support the view that infection management after instrumented spinal surgery remains highly individualized and generally requires multimodal treatment rather than any single definitive intervention (7,35).

In the present review, a reoperation- and mortality-free course (the pragmatically defined primary outcome) was reported in 75.0% of patients; however, because all patients received multimodal treatment and no comparator group was available, these uncontrolled data cannot establish whether CLAP itself contributed to this outcome. Implant preservation remained difficult: complete retention was achieved in only 51.1% of cases with available data, whereas complete implant removal was still required in 35.6%. These observations are broadly consistent with prior reports indicating that implant retention may be feasible in some patients with early postoperative infection but often remains difficult in advanced or chronic cases (7,35,36).

At the same time, favorable outcomes have also been reported with conventional multimodal treatment strategies without CLAP. Zhang et al. reported that many patients with SSI after posterior thoracolumbar and lumbar instrumentation could be successfully managed with revision surgery, irrigation and debridement, implant retention, and prolonged antibiotic therapy; instrumentation was retained in 175 of 191 cases, with no recurrence during follow-up (38). Likewise, Dubée et al. reported that debridement with implant retention followed by 3 months of combination antibiotic therapy appeared safe and effective for early-onset postoperative spinal implant infection (39). In the present review as well, debridement before CLAP was reported in 75.0% of patients with available data, systemic antibiotic therapy was used in all 48 patients, and NPWT was used in 46 of 48 patients. Collectively, these findings suggest that the potential role of CLAP is not to replace established treatments, but rather to enhance local antimicrobial delivery within an existing multimodal treatment framework. However, the current evidence remains insufficient to determine whether CLAP provides incremental benefit beyond established strategies such as debridement, systemic antibiotics, and NPWT (36-38).

Nevertheless, an important point in the existing literature is that CLAP has a plausible biological rationale. In implant-related infection, bacterial biofilm formation may markedly reduce the effectiveness of systemic antibiotic therapy alone (40). CLAP may therefore offer a theoretical advantage by maintaining a high local antibiotic concentration at the site of infection while systemic therapy addresses surrounding tissue involvement and hematogenous spread (41). This concept is analogous to implant-retention strategies used in other orthopaedic fields, such as DAIR for PJI, where success rates vary widely depending on patient selection, chronicity of infection, causative organisms, and surgical technique (7,35). Therefore, the findings of the present review should be interpreted not as evidence that CLAP is superior to existing treatments, but rather as preliminary support for its potential role as a useful adjunct within a broader multimodal treatment strategy.

Our review should also be distinguished from the recent systematic review and individual-participant-data meta-analysis of CLAP for orthopaedic infections by Ng et al. (14). That review addressed CLAP across all orthopaedic sites—predominantly PJI, FRI, and extremity osteomyelitis—and analyzed outcomes by infection type rather than by anatomical location; spinal cases formed only a small subset and were not examined as a spine-specific group. The present review is not simply a subset of that cohort. First, it is restricted to spinal infection and harmonizes spine-specific variables (infection phase, instrumentation status, implant management, spinal level, and epidural or paraspinal/psoas involvement) that are essential for spinal decision-making but are not captured in an analysis pooled across anatomical sites. Second, it incorporates additional spine-specific reports that are not part of that cohort—including several more recent single-patient reports—together with previously unpublished IPD obtained from corresponding authors for 4 patients, while consolidating earlier same-group reports into the most complete multicenter series to avoid double counting. The two reviews therefore overlap only partially at the level of individual spinal reports, and the present work provides the first spine-focused synthesis, risk-of-bias appraisal, and outcome framing for CLAP in spinal infection.

Strengths

First, to our knowledge, this is the first systematic review to specifically synthesize the evidence on CLAP for spinal infections. The review was conducted and reported in accordance with the PRISMA 2020 statement. It consolidates the currently available evidence and identifies key gaps to be addressed in future research.

Second, this review adopted a broad scope by including spinal infections as a whole, rather than limiting the analysis to postoperative spinal SSI alone. This approach allowed us to capture a wider range of clinical settings in which CLAP has been applied, including vertebral osteomyelitis and pyogenic spondylodiscitis.

Limitations

Most importantly, the current evidence base is overwhelmingly dominated by a single retrospective multicenter case series, which contributed 40 of the 48 patients (83.3%); the remaining evidence comprised one small case series (1 patient) and seven single-patient case reports. Consequently, the findings of this review largely reflect the experience of a single cohort rather than independent validation across multiple studies, and all results and conclusions should be interpreted in this light. First, the available evidence mainly consisted of cases from case series of postoperative spinal SSI. These studies are inherently prone to selection, reporting, and publication biases. The included evidence was dominated by postoperative spinal SSI, which limits the generalizability of the findings to other spinal infections such as primary vertebral osteomyelitis or pyogenic spondylodiscitis. Second, CLAP was almost never used as a stand-alone treatment: all patients also received systemic antibiotics, 95.8% received NPWT, and most underwent surgical debridement or other source control. Because no control group was available, these concurrent interventions represent major confounders, and it was not possible to determine whether the observed outcomes were attributable to CLAP itself or to the co-interventions. Third, reporting was inconsistent across the included studies, and missing data were common for several clinically important variables. Outcome definitions and completeness of reporting varied substantially, which limited comparability between reports. Fourth, the included evidence showed marked clinical and geographic homogeneity: all studies originated from Japanese centers and used highly similar protocols, including iSAP, gentamicin at 1.2 mg/mL, and an infusion rate of 2 mL/h. Although this uniformity reduces protocol-related heterogeneity, it also means that the findings reflect a single national practice pattern and a single antibiotic regimen, so their external validity is limited; it remains unknown whether comparable outcomes would be obtained in other healthcare settings, with different causative organisms and resistance patterns, or with alternative antibiotics, concentrations, infusion rates, or delivery modalities (e.g., iMAP or CLAI/CLAA). These results should therefore not be generalized beyond the specific population, setting, and protocol represented here without further multinational, multi-protocol studies. The definition of the primary outcome was also constrained by this heterogeneity: because clinically detailed outcomes were not reported uniformly, we used the reoperation- and mortality-free proportion (the absence of any additional surgery or death after CLAP) as a pragmatic outcome that could be applied across reports, recognizing that this may misclassify patients with persistent infection managed non-operatively, or those lost to follow-up, as treatment successes.

However, this pragmatic outcome may not fully reflect true microbiological or clinical resolution of infection. These issues also precluded quantitative meta-analysis. Finally, the subgroup findings were summarized descriptively and were not statistically tested; given the small numbers, this review cannot exclude the possibility that CLAP may be more beneficial in selected clinical subgroups. Our search also had limitations of coverage and screening validation. We searched major international databases (MEDLINE, Embase, CENTRAL, ICTRP, and ClinicalTrials.gov) but did not include Japanese-language databases (e.g., Ichushi Web); because a substantial part of the CLAP literature is published in Japanese, some Japanese-language reports may not have been captured. In addition, the active-learning screening (ASReview) was not subjected to a formal recall validation against a labeled reference set. We mitigated these risks through duplicate screening, the conservative SAFE stopping rule, and backward and forward citation and gray-literature searching; nonetheless, the possibility that a small number of eligible reports were missed cannot be excluded.

Conclusions

In conclusion, the currently available evidence on CLAP for spinal infection is limited to uncontrolled case reports and case series and is dominated by a single cohort. This review therefore describes how CLAP has been applied and the outcomes that have been reported, rather than demonstrating its effectiveness. In the reported cases, a reoperation- and mortality-free course (the pragmatically defined primary outcome) was documented in 75.0% of patients, and no major treatment-related toxicity was reported in the limited safety data available; however, these outcomes reflect multimodal treatment rather than CLAP alone and were assessed over a short mean follow-up of 7.5 months. Because every patient also received systemic antibiotics, almost all received NPWT, and most underwent surgical debridement, these concurrent interventions are major confounders, and the independent contribution of CLAP cannot be determined from the available uncontrolled data.

However, the current evidence remains limited by small sample size, study design, and treatment heterogeneity. Further multicenter prospective studies with standardized definitions of infection control, recurrence, safety outcomes, and implant management are needed to better define the role of CLAP and to identify which patients are most likely to benefit from this approach.

Acknowledgments

AI-assisted language editing (e.g., ChatGPT) was used to improve clarity and readability of the manuscript. All content was reviewed and approved by the authors.

Footnote

Reporting Checklist: The authors have completed the PRISMA reporting checklist. Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0156/rc

Peer Review File: Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-0156/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-2026-0156/coif). The 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.

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. Kramer A, Thavarajasingam SG, Neuhoff J, et al. Epidemiological trends of pyogenic spondylodiscitis in Germany: an EANS Spine Section Study. Sci Rep 2023;13:20225. [Crossref] [PubMed]
2. Conan Y, Laurent E, Belin Y, et al. Large increase of vertebral osteomyelitis in France: a 2010-2019 cross-sectional study. Epidemiol Infect 2021;149:e227. [Crossref] [PubMed]
3. Kilinc F, Setzer M, Gessler F, et al. Key Predictors of Treatment Failure in Conservatively Managed Spondylodiscitis: A Long-Term Retrospective Cohort Study. J Clin Med 2025;14:1973. [Crossref] [PubMed]
4. Thavarajasingam SG, Vemulapalli KV, Vishnu K S, et al. Conservative versus early surgical treatment in the management of pyogenic spondylodiscitis: a systematic review and meta-analysis. Sci Rep 2023;13:15647. [Crossref] [PubMed]
5. Karadimas EJ, Bunger C, Lindblad BE, et al. Spondylodiscitis. A retrospective study of 163 patients. Acta Orthop 2008;79:650-9. [Crossref] [PubMed]
6. Rutges JP, Kempen DH, van Dijk M, et al. Outcome of conservative and surgical treatment of pyogenic spondylodiscitis: a systematic literature review. Eur Spine J 2016;25:983-99. [Crossref] [PubMed]
7. Hersh A, Young R, Pennington Z, et al. Removal of instrumentation for postoperative spine infection: systematic review. J Neurosurg Spine 2021;35:376-88. [Crossref] [PubMed]
8. Lall RR, Wong AP, Lall RR, et al. Evidence-based management of deep wound infection after spinal instrumentation. J Clin Neurosci 2015;22:238-42. [Crossref] [PubMed]
9. Chikawa T, Sakai T, Bhatia NN, et al. Retrospective study of deep surgical site infections following spinal surgery and the effectiveness of continuous irrigation. Br J Neurosurg 2011;25:621-4. [Crossref] [PubMed]
10. Abbaszadeh A, Yilmaz MK, Izadi N, et al. Efficacy of Debridement, Antibiotics, and Implant Retention in Total Hip and Knee Arthroplasty: A Systematic Review and Meta-Analysis. J Arthroplasty 2026;41:19-38. [Crossref] [PubMed]
11. Sendi P, Trebse R, Zimmerli W. Debridement, antibiotics and implant retention: are all approaches to periprosthetic joint infection equal? Same procedures, different outcomes. Clin Microbiol Infect 2025;31:1445-50. [Crossref] [PubMed]
12. Snowden R, Garza R, Alexeev P, et al. Risk factors associated with reoperation and recurrent infection following surgical site infections treated with surgical irrigation and debridement. Spine Open 2025;1:e0011.
13. Himeno D, Matsuura Y, Maruo A, et al. A novel treatment strategy using continuous local antibiotic perfusion: A case series study of a refractory infection caused by hypervirulent Klebsiella pneumoniae. J Orthop Sci 2022;27:272-80. [Crossref] [PubMed]
14. Ng PZ, Yamamoto N, Yamamoto M, et al. Continuous local antibiotic perfusion for orthopaedic infections: A systematic review and pooled individual participant data analysis of observational reports. J Orthop Surg (Hong Kong) 2026;34:10225536261441913. [Crossref] [PubMed]
15. Maruo A, Oda T, Mineo R, et al. Continuous local antibiotic perfusion: A treatment strategy that allows implant retention in fracture-related infections. J Orthop Surg (Hong Kong) 2022;30:10225536221111902. [Crossref] [PubMed]
16. Kosugi K, Zenke Y, Sato N, et al. Potential of Continuous Local Antibiotic Perfusion Therapy for Fracture-Related Infections. Infect Dis Ther 2022;11:1741-55. [Crossref] [PubMed]
17. Nagatani S, Kato S, Yokogawa N, et al. Surgical Treatment for Emphysematous Osteomyelitis of the Lumbar Spine: A Case Report. Spine Surg Relat Res 2024;8:540-3. [Crossref] [PubMed]
18. Ohyama S, Inoue M, Toshi N, et al. Retroperitoneal Continuous Local Antibiotic Perfusion for Refractory Pyogenic Vertebral Osteomyelitis: A Case Report. Cureus 2023;15:e50636. [Crossref] [PubMed]
19. Takahashi H, Koda M, Funayama T, et al. Continuous local antibiotic perfusion for patients with surgical site infection after instrumented spinal surgery; a novel technique to retain the implants. J Clin Neurosci 2021;93:70-4. [Crossref] [PubMed]
20. Maruo A, Oda T, Miya H, et al. Intra-medullary antibiotics perfusion (iMAP) for the control of fracture-related infection early after osteosynthesis. J Orthop Surg (Hong Kong) 2021;29:23094990211051492.
21. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021;372:n71. [Crossref] [PubMed]
22. Van De Schoot R, De Bruin J, Schram R, et al. An open source machine learning framework for efficient and transparent systematic reviews. Nat Mach Intell 2021;3:125-33.
23. Boetje J, van de Schoot R. The SAFE procedure: a practical stopping heuristic for active learning-based screening in systematic reviews and meta-analyses. Syst Rev 2024;13:81. [Crossref] [PubMed]
24. Kalfas F, Severi P, Scudieri C. Infection with Spinal Instrumentation: A 20-Year, Single-Institution Experience with Review of Pathogenesis, Diagnosis, Prevention, and Management. Asian J Neurosurg 2019;14:1181-9. [Crossref] [PubMed]
25. Chahoud J, Kanafani Z, Kanj SS. Surgical site infections following spine surgery: eliminating the controversies in the diagnosis. Front Med (Lausanne) 2014;1:7. [Crossref] [PubMed]
26. Munn Z, Barker TH, Moola S, et al. Methodological quality of case series studies: an introduction to the JBI critical appraisal tool. JBI Evid Synth 2020;18:2127-33. [Crossref] [PubMed]
27. Campbell M, McKenzie JE, Sowden A, et al. Synthesis without meta-analysis (SWiM) in systematic reviews: reporting guideline. BMJ 2020;368:l6890.
28. Takahashi H, Okuyama K, Toki Y, et al. Efficacy and Limitations of Continuous Local Antibiotic Perfusion in Treating Surgical Site Infections Following Instrumented Spinal Surgery: A Retrospective Multicenter Study. Infect Dis Ther 2025;14:421-31. [Crossref] [PubMed]
29. Kitano D, Sakurai A, Kuwazuru K, et al. Intra-soft tissue and intramedullary antibiotic perfusion in combination with negative pressure wound therapy. J Wound Care 2023;32:S14-23. [Crossref] [PubMed]
30. Okada F, Ohishi H, Iwakura R, et al. A case of postoperative infection at the site of spinal instrumentation for thoracolumbar fracture successfully treated with CLAP therapy. J Jpn Soc Bone Joint Infect 2024;38:41-6.
31. Uehara K, Okazaki T, Imagama T. A case of successful treatment using a combination of CLAP and latissimus dorsi muscle flap for infection after posterior thoracic spinal fusion following radiation therapy. J Jpn Soc Bone Joint Infect 2024;38:87-91.
32. Matsumoto M, Choe H, Enomoto D, et al. A case of repeated pyogenic spondylitis unresolved with spinal fusion surgery cured by continuous local antibiotic perfusion therapy (CLAP). J Jpn Soc Bone Joint Infect 2024;38:37-40.
33. Kuniyasu A, Toki Y, Furuya T, et al. A case of decubitus infection of a spinal fusion implant that was cured by only partial removal with continuous local antibiotic perfusion. Journal of The Japan Medical Society of Spinal Cord Lesion 2024;37:90-3.
34. Kawamoto A, Sakagami A. A case of MRSA epidural abscess treated with debridement and continuous local antibiotic perfusion (CLAP). Chubu Nihon Seikei Saigai Geka Gakkai Zasshi 2025;68:77-8.
35. Karczewski D, Schnake KJ, Osterhoff G, et al. Postoperative Spinal Implant Infections (PSII)-A Systematic Review: What Do We Know So Far and What is Critical About It?. Global Spine J 2022;12:1231-46. [Crossref] [PubMed]
36. Bürger J, Palmowski Y, Pumberger M. Comprehensive treatment algorithm of postoperative spinal implant infection. J Spine Surg 2020;6:793-9. [Crossref] [PubMed]
37. Nakamura S, Nakai T, Hosozawa K, et al. Treatment strategy for surgical site infection post posterior lumbar interbody fusion: A retrospective study. J Orthop 2022;31:40-4. [Crossref] [PubMed]
38. Zhang T, Lian X, Chen Y, et al. Clinical outcome of postoperative surgical site infections in patients with posterior thoracolumbar and lumbar instrumentation. J Hosp Infect 2022;128:26-35. [Crossref] [PubMed]
39. Dubée V, Lenoir T, Leflon-Guibout V, et al. Three-month antibiotic therapy for early-onset postoperative spinal implant infections. Clin Infect Dis 2012;55:1481-7. [Crossref] [PubMed]
40. Mirza YH, Tansey R, Sukeik M, et al. Biofilm and the Role of Antibiotics in the Treatment of Periprosthetic Hip and Knee Joint Infections. Open Orthop J 2016;10:636-45. [Crossref] [PubMed]
41. Okae Y, Nishitani K, Sakamoto A, et al. Estimation of Minimum Biofilm Eradication Concentration (MBEC) on In Vivo Biofilm on Orthopedic Implants in a Rodent Femoral Infection Model. Front Cell Infect Microbiol 2022;12:896978. [Crossref] [PubMed]