The relationship of β‐glucuronidase activity in crevicular fluid to probing attachment loss in patients with adult periodontitis

Abstract

Analysis of gingival crevicular fluid (GCF) offers a non‐invasive means of studying the host response in periodontal disease, and may provide an early‐indication of the patient at risk for active periodontitis. A number of host markers have been studied for their relationship to disease activity (probing attachment loss or PAL). GCF levels of the acid glycohydrolase β‐glucuronidase (βG), a marker of primary granule release from polymorphonuclear leukocytes (PMN), have been shown to identify patients with periodontitis at risk for additional PAL. In this multicenter trial, we evaluated (a) the short‐term effect of conservative periodontal therapy on βG activity in GCF, and (b) the relationship of persistently elevated βG activity to PAL in patients who were monitored for 6 months. The study population included a total of 140 patients with chronic adult periodontitis. 130 patients were on a regular recall schedule, and 10 were previously untreated. After collection of baseline clinical data at all sites, and analysis of βG in GCF from one site (mesiobuccal) per tooth, the patients received a scaling and prophylaxis. Two weeks later patients were seen for collection of GCF. If elevated enzyme activity was found at 2 weeks, the patients were seen at 3 months for a clinical exam and GCF collection. Clinical parameters were collected from all patients at 6 months. Therapy tended to reduce βG activity in GCF. Allowing ± 20 units of βG activity for sampling and assay variability. 2 weeks after therapy, 24.5% of patients had lower βG levels, 8.6% of patients had higher levels and 66.9% of patients did not change. Using patients as the unit of observation, algorithms were created relating persistently elevated βG activity to PAL. The algorithms were based on analysis of βG data at baseline, 2 weeks and 3 months, or baseline and 2 weeks. For the algorithms using βG data collected at 3 examinations, the risk ratio ranged from 7X to 14X, total predictive value ranged from 83% to 88%, sensitivity ranged from 65% to 77%, specificity ranged from 85% to 89% and the K statistic ranged from 0.41 to 0.44. If only the baseline and 2‐week GCF data was used, accuracy decreased slightly, but a strong association between elevated βG activity and PAL remained (i.e., risk ratio of 6x to 10X). These data support our earlier findings that elevated levels of BG in GCF, indicative of increased PMN activity in the crevice, is associated with PAL in patients with chronic adult periodontitis. The role of the PMN in destruction of periodontal tissue may relate to release of lysosomal enzymes, and generation of reactive oxygen metabolites in the crevicular environment.