Table 3.
Biomarkers With a Statistically Significant Correlation or Between-Group Difference in Relation to an Outcome Measure in the Included Studies a
| Biomarker | First Author (Year) | N | Follow-up | Key Findings |
|---|---|---|---|---|
| Biomarkers of Chondral Metabolism | ||||
| CTX-II | Chimielewski (2012) 7 | 28 | 16 wk | u-CTX-II concentrations decreased over time and correlated with numeric pain rating scores at 4, 8, 12, and 16 wk postsurgery (r = 0.406; P = .039) Negative correlation between u-CTX-II concentrations and IKDC-SKF scores at 4, 8, 12, and 16 wk postsurgery (r = −0.402; P = .034) |
| Sullivan (2023) 60 | 23 | 4 wk | sf-CTX-II negatively correlated with VAS pain score over the first 4 wk postoperatively (r = −0.39; P = .002) | |
| Tourville (2013) 63 | 35 | 4 y | Higher u-CTX-II/s-CPII ratios positively correlated with joint-space narrowing at 4 years on weightbearing radiograph; 11 patients with joint-space narrowing had significantly higher u-CTX-II/s-CPII ratios at 4 y compared with 31 ACL-intact controls | |
| C2C | Pietrosimone (2016) 54 | 20 | 43 ± 36 mo | s-C2C levels negatively correlated with walking speed (r = −0.52; P = .02), even after accounting for variance of stance phase duration (partial r = −0.53; P = .02) |
| Pietrosimone (2017) 53 | 18 | 6 mo | s-C2C/s-CPII ratios measured within 2 wk of surgery were negatively related to vertical ground-reaction force LSI at 6 mo postoperatively (r = −0.5; P = .04), but this was not significant after controlling for walking speed (r = −0.24; P = .36) | |
| Biomarkers of Aggrecan Metabolism | ||||
| sGAG | Amano (2018) 3 | 26 | 3 y | Patient group characterized by high sf-sGAG and low sf-IL-6, IL-8, IL-10, TNF-α, MMP-1, and MMP-3 positively correlated with higher T1ρ relaxation times (medial tibia: β = 3.29, P = .001; patella: β = 2.46, P = .007) and T2 relaxation times (medial tibia: β = 1.48, P = .32; patella: β = 1.74, P = .37) in the medial tibia and patella |
| C6S | Sobue (2017) 56 | 62 | 2 y | Median baseline sf-Δdi-C6S levels at the time of surgery were significantly higher in the group who showed an increase in the number of high-grade cartilage lesions at 2-y follow-up compared with the group who did not show an increase in the number of high-grade cartilage lesions (53.4 vs 73.5; P = .004); high-grade cartilage lesions were defined as an increase in Outerbridge grade 3 or 4 lesions arthroscopically |
| KS | Sobue (2017) 56 | 62 | 2 y | Lower sf-KS levels at the time of surgery were associated with an increase in the number of high-grade cartilage lesions at 2 years, as defined by increased Outerbridge grade 3 or 4 lesions arthroscopically (P = .021) |
| C6S/C4S | Sobue (2017) 56 | 62 | 2 y | Lower sf-C6S/C4S levels at the time of surgery were associated with an increase in the number of high-grade cartilage lesions at 2 years, as defined by increased Outerbridge grade 3 or 4 lesions arthroscopically (P = .028) |
| Biomarkers of Noncollagenous Proteins | ||||
| Serum biochemical profile of increasing COMP and MCP-1 | Lisee (2021) 40 | 24 | 12 mo | Patients who had a serum biochemical profile of increasing s-COMP and increasing s-MCP-1 between the preoperative and 6-mo postoperative time points were associated with greater lateral femoral (β = 12.71; P = .04) and lateral tibial (β = 3.88; P = .001) MRI T1p relaxation times at 12 mo postoperatively; a k-means cluster analysis was used to create the different biomarker profile groups based on biomarker changes with time |
| MMP-3 | Pietrosimone (2017) 53 | 18 | 6 mo | Higher s-MMP-3 measured within the first 2 wk postsurgery correlated with reduced knee adduction moment LSI at 6-mo follow-up (r = −0.64; P = .01) Higher s-MMP-3 measured at 6 mo postoperatively correlated with reduced knee adduction moment LSI (r = −0.67; P = .01) and reduced vertical ground-reaction force loading rate LSI (r = −0.6; P = .01) at 6-mo follow-up |
| Evans-Pickett (2021) 15 | 38 | 6 mo | High sf-MMP-3 collected day 7 postinjury correlated with aberrant biomechanics at 6 mo postoperatively, including underloading and a stiffened knee gait strategy | |
| MMP-9 | Latterman (2018) 36 | 22 | 2 y | Patients who failed to meet the Patient Acceptable Symptom state for the KOOS–Quality of Life had significantly higher sf-MMP-9 on the day of surgery compared with those did meet this (mean ± SD, 30.99 ± 35.96 vs 6.94 ± 10.30 ng/mL; P = .01; Cohen d = 1.07) |
| Biomarkers of Other Processes (ie, inflammation) | ||||
| IL-6 | Pietrosimone (2017) 53 | 18 | 6 mo | Higher s-IL-6 measured at the 6-mo postoperative mark correlated with reduced knee adduction moment LSI (r = −0.60; P = .02) after controlling for walking speed |
| Evans-Pickett (2021) 15 | 38 | 6 mo | High sf-IL-6 collected day 7 postinjury correlated with aberrant biomechanics at 6 mo postoperatively, including underloading and a stiffened knee gait strategy | |
| Gupta (2021) 19 | 59 | 12 mo | Preoperative sf-IL-6 was associated with VAS scores, KT-1000 arthrometer testing, Lysholm knee scores, and Tegner scores at 12 mo; higher IL-6 preoperatively was associated with increased pain scores, decreased mechanical stability, and poorer Lysholm and Tegner scores | |
| Sullivan (2023) 60 | 23 | 4 wk | sf-IL-6 was correlated with VAS pain scores preoperatively and over the first 4 wk postoperatively (r = 0.52; P < .001) | |
| IL-1b | Inoue (2016) 25 | 79 | 3 mo | Higher sf-IL-1b taken at 3-4 days postoperatively was seen in patients with a delayed recovery according to the authors’ 5-point ordinal recovery grading system (P = .03) |
| IL-1a | Latterman (2018) 36 | 22 | 2 y | Patients who failed to meet the Patient Acceptable Symptom State for the KOOS–Quality of Life had significantly higher sf-IL-1a on the day of surgery (P = .004) Patients who failed to meet the Patient Acceptable Symptom State for the IKDC had significantly higher sf-IL-1a on the day of surgery (P = .02) |
| IL-1Ra | Latterman (2018) 36 | 22 | 2 y | Patients who failed to meet the Patient Acceptable Symptom State for the KOOS–Quality of Life had significantly higher sf-IL-1Ra on the day of surgery (P = .03) |
| Markus (2023) 44 | 18 | 7.8 y | Lateral tibial plateau chondral lesion size as assessed by modified Outerbridge classification on 3-T MRI correlated with sf-IL-1Ra levels at the time of surgery (R2 = 0.271; P = .032) | |
| MCP-1 | Markus (2023) 44 | 18 | 7.8 y | sf-MCP-1 levels at the time of surgery correlated with lateral femoral condyle chondral lesion depth (R2 = 0.362; P = .01) and size (R2 = 0.292; P = .025) as assessed by the modified Outerbridge classification on 3-T MRI |
| VEGF | Markus (2023) 44 | 18 | 7.8 y | sf-VEGF levels at the time of surgery correlated with patellar chondral lesion depth (R2 = 0.606; P = .001) and size (R2 = 0.403; P = .006) as well as trochlear lesion size (R2 = 0.57; P = .001) as assessed by the modified Outerbridge classification on 3-T MRI |
| RANTES | Avila (2022) 4 | 32 | 92 days b | sf-RANTES taken before surgical incision was significantly higher in the ACLR group with stiffness requiring manipulation under anesthesia or arthroscopic arthrolysis as compared with matched controls (OR, 2.28; P = .019) |
| bFGF | Avila (2022) 4 | 32 | 92 days b | sf-bFGF taken before surgical incision was significantly higher in the ACLR group with stiffness requiring manipulation under anesthesia or arthroscopic arthrolysis as compared with matched controls (OR, 1.91; P = .047) |
| TTT | Kodama (2018) 32 | 120 | 12 mo | Patients with cyclops nodule formation 12 mo after double-bundle hamstring ACLR had significantly higher TTT as measured at the time of surgery (OR, 9.34; 95% CI, 1.94-90.3; P = .002) |
a
Table stratified according to biomarker classification described by Lotz et al. 42 See Appendix Table A2 for a list of biomarker abbreviations. Biomarker prefixes: s-, serum; sf-, synovial fluid; u-, urinary. ACLR, anterior cruciate ligament reconstruction; KOOS, Knee injury and Osteoarthritis Outcome Score; LSI, Limb Symmetry Index; MRI, magnetic resonance imaging; VAS, visual analog scale.
b
Time to diagnosis of stiffness; actual follow-up time not stated.