The superiority of immune-inflammation summary index for diagnosing periprosthetic joint infection

The Application of artificial intelligence in periprosthetic joint infection

Periprosthetic joint infection (PJI) represents one of the most devastating complications following total joint arthroplasty, often necessitating additional surgeries and antimicrobial therapy, and potentially leading to disability. This significantly increases the burden on both patients and the healthcare system. Given the considerable suffering caused by PJI, its prevention and treatment have long been focal points of concern. However, challenges remain in accurately assessing individual risk, preventing the infection, improving diagnostic methods, and enhancing treatment outcomes. The development and application of artificial intelligence (AI) technologies have introduced new, more efficient possibilities for the management of many diseases. In this article, we review the applications of AI in the prevention, diagnosis, and treatment of PJI, and explore how AI methodologies might achieve individualized risk prediction, improve diagnostic algorithms through biomarkers and pathology, and enhance the efficacy of antimicrobial and surgical treatments. We hope that through multimodal AI applications, intelligent management of PJI can be realized in the future.

Hypoalbuminemia is Highly Prevalent in Patients with Periprosthetic Joint Infection and Strongly Associated with Treatment Failure

OBJECTIVE

The role of hypoalbuminemia throughout the course of chronic periprosthetic joint infection (PJI) remains poorly understood. This study aimed to determine the prevalence and risk factors of hypoalbuminemia in periprosthetic joint infection (PJI) patients and to explore the association between hypoalbuminemia and treatment outcomes.

METHODS

This retrospective cohort study included 387 PJI cases who underwent two-stage exchange arthroplasty between January 2007 and August 2020, of which 342 were reimplanted. The mean follow-up period was 7.9 years. Multivariate logistic regression analyses were performed to identify risk factors for hypoalbuminemia and to assess the effect of hypoalbuminemia at 1st- and 2nd-stage exchange on the treatment outcome. Furthermore, the impact of dynamic changes in hypoalbuminemia was investigated.

RESULTS

The prevalence of hypoalbuminemia at 1st- and 2nd-stage exchange was 22.2% and 4.7%, respectively. Patients with age ≥ 68 years and those with isolation of Staphylococcus aureus, Streptococcus, or Gram-negative bacteria exhibited a higher risk of hypoalbuminemia. Hypoalbuminemia at 1st-stage was significantly related to treatment failure (OR = 3.3), while hypoalbuminemia at 2nd-stage raised the OR to 10.0. Patients with persistent hypoalbuminemia at both the 1st- and 2nd-stage exchanges had a significantly higher rate of treatment failure than patients with hypoalbuminemia at the 1st-stage but normal albumin levels at the 2nd-stage exchange (55.6% vs 20.0%, p = 0.036).

CONCLUSION

One in five patients with chronic PJI exhibits hypoalbuminemia. Hypoalbuminemia is more likely to develop in patients of advanced age and those infected by specific highly virulent organisms. Also, our results highlight the close association between hypoalbuminemia and treatment outcomes.

Preoperative Laboratory Values Predicting Periprosthetic Joint Infection in Morbidly Obese Patients Undergoing Total Hip or Knee Arthroplasty

BACKGROUND

Morbidly obese patients are an ever-growing high-risk population undergoing total hip arthroplasty (THA) and total knee arthroplasty (TKA) for end-stage osteoarthritis. This study sought to identify preoperative laboratory values that may serve as predictors of periprosthetic joint infection (PJI) in morbidly obese patients undergoing THA or TKA.

METHODS

All morbidly obese patients with preoperative laboratory data before undergoing primary elective TKA or THA were identified using the Premier Healthcare Database. Patients who developed PJI within 90 days after surgery were compared with patients without PJI. Laboratory value thresholds were defined by clinical guidelines or primary literature. Univariate and multivariable regression analyses were utilized to assess the association between PJI and preoperative laboratory values, including total lymphocyte count, neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR), monocyte-lymphocyte ratio (MLR), systemic immune-inflammation index (SII), albumin level, platelet count, albumin-globulin ratio, hemoglobin level, and hemoglobin A1c.

RESULTS

Of the 6,780 patients identified (TKA: 76.67%; THA: 23.33%), 47 (0.69%) developed PJI within 90 days after surgery. The rate of PJI was 1.69% for patients with a hemoglobin level of <12 g/dL (for females) or <13 g/dL (for males), 2.14% for those with a platelet count of <142,000/µL or >417,000/µL, 1.11% for those with an NLR of >3.31, 1.69% for those with a PLR of >182.3, and 1.05% for those with an SII of >776.2. After accounting for potential confounding factors, we observed an association between PJI and an abnormal preoperative NLR (adjusted odds ratio [aOR]: 2.38, 95% confidence interval [CI]: 1.04 to 5.44, p = 0.039), PLR (aOR: 4.86, 95% CI: 2.15 to 10.95, p < 0.001), SII (aOR: 2.44, 95% CI: 1.09 to 5.44, p = 0.029), platelet count (aOR: 3.50, 95% CI: 1.11 to 10.99, p = 0.032), and hemoglobin level (aOR: 2.62, 95% CI: 1.06 to 6.50, p = 0.038).

CONCLUSIONS

This study identified preoperative anemia, abnormal platelet count, and elevated NLR, PLR, and SII to be associated with an increased risk of PJI among patients with a body mass index of ≥40 kg/m 2 . These findings may help surgeons risk-stratify this high-risk patient population.

LEVEL OF EVIDENCE

Prognostic Level III . See Instructions for Authors for a complete description of levels of evidence.

Identification of immune infiltration and immune-related biomarkers of periprosthetic joint infection

BACKGROUND

The immune response associated with periprosthetic joint infection (PJI) is an emerging but relatively unexplored topic. The aim of this study was to investigate immune cell infiltration in periprosthetic tissues and identify potential immune-related biomarkers.

METHODS

The GSE7103 dataset from the GEO database was selected as the data source. Differentially expressed genes (DEGs) and significant modular genes in weighted correlation network analysis (WGCNA) were identified. Functional enrichment analysis and transcription factor prediction were performed on the overlapping genes. Next, immune-related genes from the ImmPort database were matched. The protein-protein interaction (PPI) analysis was performed to identify hub genes. CIBERSORTx was used to evaluate the immune cell infiltration pattern. Spearman correlation analysis was used to evaluate the relationship between hub genes and immune cells.

RESULTS

A total of 667 DEGs were identified between PJI and control samples, and 1847 PJI-related module genes were obtained in WGCNA. Enrichment analysis revealed that the common genes were mainly enriched in immune and host defense-related terms. TFEC, SPI1, and TWIST2 were the top three transcription factors. Three hub genes, SDC1, MMP9, and IGF1, were identified in the immune-related PPI network. Higher levels of plasma cells, CD4+ memory resting T cells, follicular helper T cells, resting mast cells, and neutrophils were found in the PJI group, while levels of M0 macrophages were lower. Notably, the expression of all three hub genes correlated with the infiltration levels of seven types of immune cells.

CONCLUSION

The present study revealed immune infiltration signatures in the periprosthetic tissues of PJI patients. SDC1, MMP9, and IGF1 were potential immune-related biomarkers for PJI.

A machine learning-based model for "In-time" prediction of periprosthetic joint infection

Background

Previous criteria had limited value in early diagnosis of periprosthetic joint infection (PJI). Here, we constructed a novel machine learning (ML)-derived, "in-time" diagnostic system for PJI and proved its validity.

Methods

We filtered "in-time" diagnostic indicators reported in the literature based on our continuous retrospective cohort of PJI and aseptic prosthetic loosening patients. With the indicators, we developed a two-level ML model with six base learners including Elastic Net, Linear Support Vector Machine, Kernel Support Vector Machine, Extra Trees, Light Gradient Boosting Machine and Multilayer Perceptron), and one meta-learner, Ensemble Learning of Weighted Voting. The prediction performance of this model was compared with those of previous diagnostic criteria (International Consensus Meeting in 2018 (ICM 2018), etc.). Another prospective cohort was used for internal validation. Based on our ML model, a user-friendly web tool was developed for swift PJI diagnosis in clinical practice.

Results

A total of 254 patients (199 for development and 55 for validation cohort) were included in this study with 38.2% of them diagnosed as PJI. We included 21 widely accessible features including imaging indicators (X-ray and CT) in the model. The sensitivity and accuracy of our ML model were significantly higher than ICM 2018 in development cohort (90.6% vs. 76.1%, P = 0.032; 94.5% vs. 86.7%, P = 0.020), which was supported by internal validation cohort (84.2% vs. 78.6%; 94.6% vs. 81.8%).

Conclusions

Our novel ML-derived PJI "in-time" diagnostic system demonstrated significantly improved diagnostic potency for surgical decision-making compared with the commonly used criteria. Moreover, our web-based tool greatly assisted surgeons in distinguishing PJI patients comprehensively.

Level of evidence

Diagnostic Level III.

Accuracy of Albumin, Globulin, and Albumin-Globulin Ratio for Diagnosing Periprosthetic Joint Infection: A Systematic Review and Meta-Analysis

Periprosthetic joint infection (PJI) is one of the most intractable orthopedic diseases, partly because of the difficulty in differentiating septic from aseptic conditions. We aimed to evaluate and consolidate the diagnostic accuracy of the quantitative assessment of serum albumin (Alb), globulin (Glb), and albumin-globulin ratio (AGR), alone or in combination with the inflammatory marker, C-reactive protein (CRP), for PJI. We searched the PubMed, CINAHL, and Cochrane Library databases for studies that quantitatively measured Alb, Glb, or AGR for the diagnosis of PJI up until the 30 April 2023. A total of 2339 patients were included from 10 studies, including 845 patients with a definitive diagnosis of PJI and 1494 with non-PJI. The pooled sensitivity, specificity, and area under the curve (AUC) in the summary receiver-operating characteristic curve were as follows: 0.625, 0.732, and 0.715 for Alb; 0.815, 0.857, and 0.887 for Glb; 0.753, 0.757, and 0.875 for AGR; 0.788, 0.837, and 0.876 for CRP; 0.879, 0.890, and 0.917 for the CRP-Alb ratio; and 0.845, 0.855, and 0.908 for the CRP-AGR ratio. Serum Alb, Glb, and AGR levels are feasible and accurate diagnostic markers for PJI, and the combination of these markers with CRP levels may potentially improve preoperative serum diagnostic accuracy. Future prospective studies are required to verify these findings because of the small numbers of included studies.

Fibrinogen, albumin-to-globulin ratio, and fibrinogen to albumin-to-globulin ratio may be potential diagnostic biomarkers for infected tibial nonunion

AIM

The accurate preoperative diagnosis of infected tibial nonunion remains challenging. Hence, we evaluated the diagnostic potential of novel biomarkers for infected tibial nonunion.

METHODS

This single-center retrospective study was conducted in 252 patients divided into two groups: infected tibial nonunion (67 patients) and aseptic tibial nonunion (185 patients). The preoperative clinical biomarkers included D-dimer, fibrinogen, albumin, globulin, total protein, and C-reactive protein (CRP) levels; albumin-to-globulin ratio (AGR); erythrocyte sedimentation rate (ESR); and white blood cell (WBC) count. Receiver operating characteristic (ROC) curves, sensitivity, and specificity were utilized to compare the biomarkers' diagnostic potential.

RESULTS

The area under the curve (AUC) values for fibrinogen and AGR were 0.829 and 0.821, respectively, suggesting similarly good diagnostic potentials for infected tibial nonunion. Fibrinogen and AGR were better diagnostic biomarkers for infected tibial nonunion than the WBC count; ESR; D-dimer, albumin, globulin, CRP, and total protein levels, whose AUC values were 0.623, 0.684, 0.741, 0.797, 0.765, 0.715, and 0.554, respectively. The sensitivity and specificity of fibrinogen with a cut-off value of 3.35 g/L were 71.64% and 84.86%, respectively. The corresponding values for AGR with a cut-off value of 1.33 were 73.13% and 86.49%. Moreover, the fibrinogen-AGR (FAGR), i.e., the combination of fibrinogen and AGR, had the highest diagnostic accuracy for infected tibial nonunion (AUC = 0.906). The optimal FAGR cut-off was 2.69, with fair sensitivity (74.63%) but the highest specificity (94.59%).

CONCLUSION

Fibrinogen, AGR, and FAGR are promising biomarkers for the diagnosis of infected tibial nonunion.