Tissue reaction to implant corrosion in 38 internal fixation devices.

The corrosion characteristics, metallurgical properties, and clinical performance of 38 retrieved internal fixation devices were correlated with the tissue reaction to these devices. Metallurgical parameters included thin and heavy inclusion content, Rockwell hardness, and grain size. The excised fibrous tissue strip was directly overlying each plate at removal and sectioned between screw-hole sites. The material studied from the 38 plates consisted of 201 screw-hole junctions with associated tissue biopsy sites. Clinical histories were obtained on all 38 patients with hardware removal. The average age at the time of plate insertion was 35.6 years (range, 4 to 75 years). Insertion diagnoses included acute trauma (35 patients), joint dislocations (two patients), and fracture nonunion (one patient). The devices included seven upper extremity bone plates, 19 lower extremity bone plates, and 12 hip screw plates. The devices remained in situ an average of 20.4 months (range, 3 to 60 months). Routine asymptomatic removals were performed on 17 of the patients, while the remaining 21 patients were symptomatic at the time of removal. Included in the reasons for symptomatic removal were pain associated with the implant (eight patients), nonunion (four patients), bursae prominence (three patients), and implant breakage (two patients). Significant correlations were found between average tissue reaction scores and average crevice corrosion scores; a trend of increasing average tissue reaction scores with increasing average screw surface corrosion scores also was observed for the 38 devices, although this relationship was not significant. Average crevice corrosion scores and average screw surface corrosion scores were highly correlated for all removals, and for the asymptomatic and symptomatic removal groups. The metallurgical parameters of thin inclusion content and heavy inclusion content also were significantly correlated for all removals, as well as for symptomatic removals. Similarly, significant correlations were found between the individual tissue reaction scores and crevice corrosion scores from the 201 individual sites, again for all devices and for the asymptomatic and symptomatic removal groups. Tissue reaction scores and time in situ showed significant correlation, with tissue reaction decreasing over time. Crevice corrosion and screw surface corrosion scores were not significantly related to time in situ. The results of this study indicate that there is considerable tissue reaction to the corrosion products of this material.(ABSTRACT TRUNCATED AT 400 WORDS)