Blood Cell Ratios as Predictors of Response to Platelet-Rich Plasma in Knee Osteoarthritis

Background and Rationale: Intra-articular platelet-rich plasma (PRP) is increasingly used for symptomatic knee osteoarthritis, but the clinical response is highly variable and no inexpensive, routinely available biomarker currently guides patient selection. Recent work indicates that the peripheral-blood inflammatory phenotype is associated with PRP response (Tonutti et al., 2025) and that the absolute neutrophil count is an independent early-response predictor (De Luca et al., 2025). However, simple complete-blood-count-derived ratios such as the neutrophil-to-lymphocyte ratio (NLR) have not been tested prospectively for this purpose. This study addresses that gap with a pragmatic, low-cost translational biomarker design.

Design: Prospective, single-arm, observational cohort study. Outcome assessors are blinded to patients' blood-count values; patients are not informed of their NLR; statistical analysis is conducted blinded.

Population and Setting: 120 patients aged 40 to 60 years with symptomatic Kellgren-Lawrence grade 2 to 3 knee osteoarthritis and a body mass index below 40, recruited over a 4-month window (target 30 enrollments per month) from the Department of Orthopaedics and Traumatology, University of Kyrenia, Dr. Suat Gunsel Hospital. Approximately 180 patients are anticipated to be screened, with a target of 102 completed, analyzable participants after an estimated 15 percent attrition.

Intervention (uniform, per protocol): Each participant receives three leukocyte-poor PRP injections at one-week intervals. PRP is prepared by manual double-spin centrifugation (soft spin 1500 rpm for 5 minutes, then hard spin 3300 rpm for 10 minutes) from 20 mL of whole blood per session (total 60 mL per patient across the three sessions), anticoagulated with 8.5 percent ACD-A, without exogenous activation, targeting a 4 to 6 fold platelet concentration. Injections are delivered under ultrasound guidance via a superolateral approach with skin infiltration of 1 percent lidocaine. A 48-hour restriction on non-steroidal anti-inflammatory drugs is applied after each injection.

Biomarkers: The primary predictor is the baseline NLR. Secondary predictors are PLR, SII, and MLR; tertiary predictors are C-reactive protein and erythrocyte sedimentation rate. All are derived from a routine baseline complete blood count.

Outcomes: The primary clinical outcome is the change in WOMAC total score at 6 months. Treatment response is defined a priori as at least 30 percent improvement in WOMAC together with fulfilment of the OMERACT-OARSI responder criteria. Clinical assessments occur at baseline and at 1, 3, and 6 months.

PRP Product Characterization: Quality control is performed in two tiers. Tier 1 (all 120 patients, all 360 batches) records the final platelet concentration, the total leukocyte count, and the final volume. Tier 2, a pre-specified validation subset of approximately 40 patients (the first 20 consecutive patients plus every fifth subsequent patient, approximately 120 batches), additionally records the leukocyte differential and erythrocyte contamination.

Statistical Analysis and Sample Size: The primary analysis is a multivariable linear regression of 6-month WOMAC change on baseline NLR, adjusted for age, body mass index, Kellgren-Lawrence grade, and baseline WOMAC. A secondary analysis evaluates the discriminative performance of baseline NLR for responder status by receiver operating characteristic analysis. For the primary regression (alpha 0.05, power 0.80, one numerator degree of freedom, five predictors, effect size f-squared 0.10), 82 participants are required; the secondary analysis requires 95. To satisfy both, the enrollment target is 120, anticipating 102 completers.

Ethics and Reporting: Ethics approval is sought from the Girne University Clinical Research Ethics Committee. Reporting follows the STROBE and REMARK recommendations, and the protocol follows SPIRIT guidance.