The PJI-TNM Classification as Predictor for Revision-Free Implant Survival Rates in Patients with Periprosthetic Joint Infection of the Hip or Knee Joint.

BACKGROUND

Periprosthetic joint infections (PJIs) remain a major challenge in arthroplasty. This study tries to evaluate the PJI-TNM classification as predictor for the revision-free implant survival in patients with PJI of the hip or knee joint.

METHODS

To this end, we perform a retrospective study of all consecutive patients with PJI of an inlying hip or knee arthroplasty between January 2015 and December 2019.

RESULTS

A total of 443 cases (hip: = 247; knee = 196) were identified. In total, 439 patients underwent surgery (DAIR: = 138 cases (31%), explantation: = 272 (61%), irrigation with debridement without exchange of implant components: = 29 (6.5%)). Four patients refused surgical treatment and 39.5% were lost to follow-up. In total, 78 patients died during follow-up and 27 deaths were directly related to PJI/complications during treatment. Patients with inlying "standard"-implants ( < 0.001) and without previous history of PJI ( = 0.002) displayed a significantly higher postoperative revision-free implant survival. In terms of the PJI-TNM subclassification, patients with loosened implants but without soft-tissue defects (T1) displayed the highest revision-free implant survival. In contrast, patients classified as M3 (no surgical treatment possible) displayed an inferior outcome compared to M0, M1, or M2. Patients with different N-subclassifications ("non-human cells"/causative pathogen) did not display differences in revision-free implant survival.

CONCLUSIONS

The PJI-TNM classification is well suited to classify PJIs. Its complexity allows for more than 500 different combinations of classifications. Further validation data are needed, but to us, the PJI-TNM classification seems to offer the possibility of comparing patients with PJIs. It may, therefore, be a very valuable tool in order to compare cohorts with PJIs and provide individual data for patient specific outcomes.