Current Pathologic Scoring Systems for Metal-on-metal THA Revisions are not Reproducible

Abstract

BackgroundThe aseptic lymphocyte vasculitis-associated lesion (ALVAL) score and the modified Oxford ALVAL score are frequently used scoring methods to evaluate the morphologic features of periprosthetic tissues around metal-on-metal (MoM) hip implants. Except for the initial studies of these two morphology scoring methods, to our knowledge, no other studies have reported on intraclass correlation coefficient (ICC) values for interobserver reliability of these scoring methods.Questions/purposesAre the ALVAL and Oxford ALVAL scores reproducible?MethodsThe periprosthetic tissue of 37 revisions of 36 patients with failed MoM THAs were independently scored by three experienced pathologists using ALVAL and Oxford ALVAL scoring methods. All patients were included who underwent revision surgery in our hospital until January 2013, with a large-head MoM prosthesis and also met the criteria: blood serum cobalt levels, available MRI scan, and intraarticular cobalt levels. The population included 26 patients with pseudotumors diagnosed by two radiologists using the method described by Matthies et al. The ALVAL describes morphologic features of the synovial lining, tissue organization, and inflammatory cell infiltrate in periprosthetic tissues. The Oxford-ALVAL score uses a semiquantitative measure of the immune response which should be easier to score.ResultsThe ALVAL score showed an ICC of 0.38 (95% CI, 0.18–0.58) (fair) for the sum score and this improved up to 0.50 (95% CI, 0.31–0.68) (moderate) using the modified Oxford ALVAL score. The individual parameters of the ALVAL score showed an ICC for the scoring of inflammatory infiltrate of 0.37 (95% CI, 0.17–0.57), an ICC of 0.32 (95% CI, 0.12–0.53) for the scoring of tissue organization, and an ICC of 0.14 (95% CI, −0.04 to 0.34) for synovial lining.ConclusionsScoring morphologic features of MoM tissue is not reproducible using the ALVAL score or the Oxford ALVAL score. This may reflect heterogeneous morphologic features in tumor tissue and between different tumor tissue samples that cannot be reliably quantified by pathologists using the parameters of these two scoring methods. An alternative, simplified scoring system should be developed to improve the interrater agreement.Level of EvidenceLevel III, diagnostic study.