589

A Multi-Center Clinical Trial of a New Chairside Test in Distinguishing Between Diseased and Healthy Periodontal Sites. II. Association Between Site Type and Test Outcome Before and After Therapy
Ingvar Magnussen,* Rutger G. Persson,n Roy C. Page,n Timothy A. DeRouen,tn John M. Crawford,*** Rhonna L. Cohen,** Donald A. Chambers,**** Mario E.A.F.
Alves,*** and William B. Clark*

The aim of the present study was to evaluate the association between the outcome of a chairside test measuring gingival crevicular fluid (GCF) levels of the enzyme aspartate aminotransferase (AST) and other clinical measures of disease including probing depth, severity of inflammation, and GCF flow before and after therapy. We studied 91 patients with moderate to severe Periodontitis. Eight sites with probing depths between 5 mm and 8 mm and obvious signs of inflammation were selected and designated diseased sites. Four sites with probing depth 3 mm with no or minimal signs of inflammation were selected and designated non-diseased sites in patients. Thirty healthy individuals were enrolled and four sites in each were selected and designated healthy controls. Patients were treated with scaling and root planing and control subjects with supragingival prophylaxis. Measurements including GCF volume, gingival inflammation, and probing depth were performed at screening baseline, 1 week later at pretreatment baseline, and at weeks 2 and 4 after treatment. AST content of GCF was measured using a chairside colorometric test. It was concluded that the outcome of the test is an effective objective measure distinguishing between diseased sites and non-diseased sites in patients and control subjects when evaluated both prior to and following application of therapy. Use of this simple chairside test, when combined with other standard diagnostic procedures, provides an objective measurement permitting improved capacity to distinguish between diseased and non-diseased periodontal sites, and to better assess and monitor the outcome of therapy. J Periodontol 1996;67:589-596.

Words: Periodontitis/diagnosis; notransferase.

Key

gingival

crevicular

fluid/analysis; aspartate ami-

The efficacy of non-surgical periodontal therapy in patients diagnosed with periodontal disease is well-estab*The Periodontal Research Center, School of Dentistry, University of Florida, Gainesville, FL. 'Department of Periodontics, Health Sciences Center, University of Washington, Seattle, WA.

*Regional Clinical Dental Research Center. Department of Pathology. "Department of Dental Public Health Services. 'Department of Biostatistics. 'Department of Periodontics, University of Illinois, Chicago, "Center for Molecular Biology of Oral Diseases. "Department of Biochemistry.

IL.

lished. The clinical improvement occurs within a few weeks after instrumentation12 and a long-term predictable decrease of gingival inflammation and probing depth resulting in gain of attachment has been shown to occur.3-8 Although most patients respond to therapy with clinical improvement, at some sites in some patients the disease may continue to progress.910 The criteria by which improvements have been judged include assessment of gingival inflammation using the gingival index" and measurement of probing depths. Although the cutoff levels between disease versus health have varied, probing depths of ^ 4.0 mm and a gingival index score s 1.0 have been

590

TRIAL OF A CHAIRSIDE PERIODONTAL TEST Table 1.

J Periodontol June 1996

used as reasonable criteria to distinguish between established periodontal lesions or diseased sites and healthy or asymptomatic sites. New concepts of the progression of periodontal disease require new diagnostic tests that can accurately identify disease-active sites and predict disease activity.12-19 At present, routine clinical measures, including gingival index, bleeding on probing, assessment of attachment levels, probing depth, and bone height measurements made on radiographs are the most commonly used measures to define a diagnosis. All of these measures have low predictive value.5'6 20 New objective tests that aid in detecting actively deteriorating sites and in monitoring the outcome of treatment and stability of periodontal. health are needed. A test that is used for diagnostic purposes can, of course, also be used to assess the outcome of therapy. A test that can provide'objective information about treatment outcome would be a significant contribution, especially when therapy is provided in stages or phases, or when the clinical response dictates a change from one type of treatment to another. A monitoring test should conform to the criteria established for a diagnostic test. If there are no changes of disease status and treatments have not been performed, there should be little or no change in test results. However, if treatment of known efficacy has been performed and improvement in the clinical conditions occurs, then the test should reflect this. Disease prevalence must also be taken into account. Hence, if the prevalence of disease is low, the test must have high specificity to prevent false-positive results, whereas the test specificity is of less significance in the event the disease is very common and the risk of falsenegatives is minimized. To make claims of monitoring value, studies must be performed prior to and following treatment to determine if the test accurately describes changes of clinical status. We have performed a multi-center clinical trial aimed at determining the potential of an objective chairside test** based on the presence and levels of the enzyme aspartate aminotransferase (AST) in gingival crevicular fluid (GCF) to distinguish between diseased and healthy periodontal sites. We used fluid samples from control sites in periodontally healthy subjects, and from diseased and nondiseased sites in Periodontitis patients. All sites were sampled and tested twice before and twice after treatment by a therapy known to be effective. The study design, patient demographics, outcome of therapy, and association between test positivity and therapeutic outcome have been reported.21 The present paper reports the association between test outcome and clinical measures of disease and health of diseased and non-diseased sites in Periodontitis patients and control sites in healthy subjects before and "PerioGard, Xytronyx, Inc., San Diego, CA.

Subjects*

Demographic

Characteristics of Patients and Control Gainesville

43 57 80 17 3 47.0 13.8 40 60 80 10 10 32.3 9.3

Characteristics
Patient % Male % Female % White % Black % Other Average age

Seattle
50 50 81 3 16 40.2 10.1 36 64 100 0 0 35.6 7.4
et

Chicago
53
47

47 38 15 39.6 11.
67

Controls % Male % Female % White % Black % Other Average age

*From Persson
al.2

33 83 8 9
30.9 13.:

The data are reported in terms of odds ratios. We suggest that, if the chairside test does distinguish between diseased and healthy periodontal sites, percentages of sites in Periodontitis patients that are test-positive prior to treatment will be high; following application of a therapy that is known to be effective the percentage of test-positive sites will be low, and the test outcome will be significantly and positively associated with other measures of disease including probing depth, severity of inflammation, and GCF volume both before and following therapy. The present study is concerned with the following questions: 1) is the presence and level of AST in GCF associated with the commonly used measures of periodontal disease; 2) do such associations exist at different geographic locations and at different points over time; 3) do AST levels in GCF agree with other clinical measures, and are these agreements better or worse than the agreement of other clinical measures among themselves; and 4) are longitudinal changes in AST at individual sites associated with changes in other clinical measures?

following therapy.

MATERIALS AND METHODS The details of the study design and methods have been described elsewhere.21 In summary, 91 patients approximately equally distributed between three study locations (University of Florida, Gainesville; University of Illinois, Chicago; and University of Washington, Seattle) were studied. Demographic characteristics of the study are given in Table 1. In each patient, 8 sites with probing depths between 5.0 mm and 8.0 mm and with obvious signs of gingival inflammation, and 4 sites with probing depths 3.0 mm and with no or minimal signs of gingival inflammation were selected and designated diseased and nondiseased sites, respectively. Non-surgical treatment consisting of scaling and root planing with hand instruments and ultrasonic devices as deemed necessary was per-

Volume 67 Number 6

MAGNUSSON, PERSSON, PAGE,

Percent PerioGard Positive

ET AL.

591

formed on one occasion for each tooth. Four normal control sites were studied in each of 30 periodontally normal individuals before and following coronal prophylaxis. GCF was collected prior to any other tissue manipulation, volumes measured with the Periotron 6000, and AST measured by the chairside test. Gingival inflammation was assessed" and probing depths measured with an automated probe.221111 All measurements were made at the screening baseline, one week later at the pretreatment baseline, and at 2 and 4 weeks after treatment. The strength of the association between the test outcome and probing depth, gingival index, and GCF volume was calculated for all of the diseased and non-diseased sites in patients and control sites in healthy subjects. To quantify the strength of the associations, the clinical measures were dichotomized at levels corresponding to no or low levels of disease, versus moderate or high levels of disease. Cutoffs were set at probing depths s 4.0 mm versus probing depths > 5.0 mm; gingival index scores < 1 versus scores of > 2; GCF volume < 40 Periotron units versus > 40 units. In addition, associations were calculated for sites distributed according to gingival index score as well as for GCF volumes < 20, 21 to 40, 41 to 80, and > 80 Periotron units. The odds ratio was selected as the appropriate statistical approach to test the association of the test outcome with measures of disease before and after therapy. The odds ratio is appropriate because the chairside test is dichotomous and the clinical outcomes could be dichotomized to indicate disease state, and also to indicate whether they either improved or did not improve. The ratio was 1 when there was no association, greater than 0 but less than 1 when the associations were negative, and greater than 1 when the association was positive, with a larger odds ratio indicating a stronger association. In making the calculations, the unit of observation used was the individual site. To adjust the odds ratio for potential lack of independence for sites within the same mouth, each patient was considered as a stratum and the Mantel-Haenszel and log odds methods were used to estimate a common odds ratio over several strata,23 or by using a regression model to adjust for patient effects.24 In most instances the log odds and Mantel-Haenszel methods gave the same estimates and these are reported in Tables 2, 3, and 4. In the one instance (with sparse data) when the two methods differed, both are given. The one exception to the use of these methods was the comparison of the proportion of test-positive diseased sites in patients, and control sites in healthy subjects. Because these types of sites did not occur together in the same persons, a stratified analysis was not possible. Instead, the average percentage of test-posHPeriotron, Amityville, NY. Florida Probe Company, Gainesville, FL.

DNCDNCDNC

Seattle

Gainesville

Chicago

Figure 1. Percentage of AST test-positive sites by site type, when classified as diseased (D), non-diseased sites in patients (N), or control sites in healthy subjects (C) at Seattle, Gainesville, and Chicago for screening prebaseline, baseline, and 14 and 28 days post-therapy. itive sites was computed for diseased sites in each patient, and for healthy sites in control subjects. These patientlevel averages were compared for patients and controls using a t test. In the analysis for association of test results with site type at each visit, the designation of the site as diseased or non-diseased was based on its classification at baseline.

RESULTS Scaling and root planing resulted in significant decreases in probing depth, gingival index, and GCF volumes. These values have been reported previously,21 and they are in good agreement with the known outcome of scaling
and root planing.5'625 The associations of the chairside test outcome with other clinical measures of disease are illustrated in Figures 1 to 3. Notable is the relationship between classification of sites diseased or non-diseased and healthy control, and the percentage that were test-positive (Fig. 1). For prebaseline and baseline measurements, the percentages of diseased sites that were test-positive were 87% and 81% at Seattle, 92% and 86% at Gainesville, and 70% and 60% at Chicago, respectively (Fig. 1). In marked contrast, values for control sites in healthy subjects at the three locations for the single baseline measures were very low, ranging from 7% to 14% (Fig. 1). Values for non-diseased sites in Periodontitis patients were intermediate, with percentage test-positive 18% and 12% at Seattle, 54% and 33% at Gainesville, and 15% and 9% at Chicago, for prebaseline and baseline, respectively (Fig. 1). Following treatment, the percentage of test-positive diseased sites decreased substantially with values ranging from about 18% in Seattle to 47% in Gainesville for weeks 2 and 4 posttherapy (Fig. 1). Values for non-diseased sites in patients and control sites in healthy subjects were even lower

(Fig. 1).

592

TRIAL OF A CHAIRSIDE PERIODONTAL TEST

J Periodontol June 1996

Percent PerioGard Positive

Table 2. Association of Test With Site diseased in patients and controls)

Type (diseased

versus non-

Seattle
Time Prebaseline Pretreatment baseline Week 2 Week 4

Gainesville Odds Ratio 11.2 14.4 7.7 7.9

Value

Chicago
Odds Ratio Value <0.0001

120

Odds Ratio
59.5

Value

<0.0001 <0.0001 <0.0002 <0.0001

<0.0001 <0.0001 <0.0002 <0.0001

16.3
17.4 3.9 6.4

37.9 4.6 7.0

<0.0001 <0.0001 <0.0001

Prebaseline

Table 3. Association of Test With and Controls Seattle Time Prebaseline Baseline Day 14 Day 28 Odds Ratio 25.8 13.5 7.4 1.9* 5.5' Value

Gingival

Index

Among Patients
012301230123

Baseline

Days 28 Days
Seattle

Gainesville Odds ratio 11.2 15.0 2.6 4.4

Value

Chicago
Odds Ratio 6.3 5.5 4.8 4.0
value

Gainesville

Chicago

Figure
on a

<0.0001 <0.0001 <0.0002 <0.53 <0.05

<0.0001 <0.0021 <0.0014 <0.0001

<0.0001 <0.0001 <0.0001 <0.003

2. Percentage of AST test-positive sites by gingival index score scale 0 to 3 for prebaseline, baseline, and week 2 and week 4 posttherapy for patients and control subjects at the 3 locations.
Percent PerioGard Positive

'Log

*Mantel-Haenszel method. odds method.

Odds ratios for the associations between percentages of test-positive and site type before and following therapy are shown in Table 2. Among the three locations, odds ratios were very high, ranging from 59.5 to 16.3 for baseline measures, and from 7.9 to 3.9 for post-treatment measures. values were < 0.0001 to < 0.0002, showing highly significant associations. The prebaseline associations of the chairside test outcome and probing depth were identical to those presented in Figure 1, since disease status was defined by probing depth at baseline. Probing depth decreased following therapy, although a significant positive association remained between probing depths > 4 mm, and percentage of testpositive sites with odds ratios of 2.5 and greater. The only exception was for Seattle patients at week 4, where values for sites greater or less than 4 mm did not differ. This exception probably resulted from the fact that only 28% of the sites with probing depths greater than 4 mm at baseline persisted in being test-positive at week 4 in Seattle patients. The relationship between chairside test outcome and gingival index is shown in Table 3 and Figure 2. With minor exceptions, the percentage of test-positive sites consistently increased with increasing gingival index scores at the same visit, and consistently decreased over time within the same gingival index category as a result of treatment (Fig. 2). The few exceptions occurred where the numbers were very small or where the preceding values were high. The data were dichotomized for low values (Gl 0 or 1) versus high values (Gl 2 or 3), and odds ratios calculated (Table 3). Values were highest in the
= =

Prebaseline aseline
28

Days Days

Figure 3. Percentage of AST test-positive sites by GCF volume with sites classified in terms of Periotron units as < 20 (A), 21 to 40 (B), 41 to 80 (C), and > 80 (D) for patients and control subjects at the 3 locations.

pretreatment baseline visits where they ranged from 25.8

to 5.5 with

values < 0.0001 to 0.0021. Values decreased after treatment, but remained positive and significant, ranging from 7.4 to 2.6 with values < 0.0001 to 0.003 for all locations except Seattle, where the odds ratio was greater than 1.0, but it was significant (P < 0.05) by one method of calculation and not by another method. The associations of chairside test outcome with GCF volumes are shown in Figure 3 and Table 4. In general, the percentage of sites that were test-positive increased with increasing GCF volume. When GCF volume was dichotomized with a cutoff of 40 Periotron units, the odds ratio for test-positivity for sites > 40 was consistently and significantly elevated for all visits at all locations (Table 4). The odds ratios ranged from 6.3 to 21.9, except for Seattle patients week 2 where the odds ratio was 3.8. All values were < 0.0001, except for a Gainesville patient at baseline, where the value was < 0.0021, and Seattle

Volume 67 Number 6

MAGNUSSON, PERSSON, PAGE, Among

Percent PerioGard Positive

ET AL.

593

Table 4. Association of Test With GCF Periotron Units Patients and Controls
Seattle Time Odds Ratio Value Gainesville

Chicago
Odds Ratio
15.7 11.3 10.9 16.3 Value
100

Odds Ratio
21.9 14.8 11.6 6.3

Value <0.0001 <0.0021 <0.0001 <0.0001

Prebaseline Baseline Day 14 Day 28

14.0 12.1 3.8 8.1

<0.0001 <0.0001 <0.02 <0.004

<0.0001 <0.0001 <0.0001 <0.0001

100

Percent Agreement

Screening
Prebaseline PG/PD

C.3 (ZD HB

Figure
were

5. The combined clinical assessment index. Measures

PG/GI PG/GCF PD/GI PD/GCF GI/GCF

of disease

dichotomized as positive or negative, using cutoffs or probing depth a 5.0, gingival index a 2.0, and GCF a 40 Periotron units. If none of the measures was positive, the index value was 0. When 1, 2, or 3 were positive, the index value was I, 2, or 3.

Seattle

Gainesville

Chicago

Percent Agreement

am PG/PD

PG/GI
PG/GCF

E33 PD/GI

Cm

PD/GCF GI/GCF

Seattle

Gainesville

Chicago

Figure of disease, including probing depth (PD), gingival inflammation {Gl), and gingival fluid flow (GCF), and between the disease measures themselves; top: data from prebaseline; bottom: data from day 28.

4. The agreement between AST test outcome and other measures

patients
<

at

weeks 2 and 4 where values

were

<

0.02 and

assess the extent of agreement of the chairside test outcome with other clinical measures of disease, all of the clinical measures including the test outcome were paired and the extent of agreement calculated. Results for

To

0.004, respectively.

the initial baseline and 28-day data are presented in Figure 4, which are typical of the agreements observed. Although the extent of agreement between various clinical

varied from one study location and patient population to another, and from visit to visit as the patient's clinical condition changed, it is notable that the agreement between the test outcome and other clinical measures at each location and visit are generally in the same range as the pairwise agreements among the other clinical measures. In fact, agreement of the chairside test was within or above that range 20 of 36 times, and those values that were exceptions were just slightly below that range. Because the best measure of clinical manifestations of disease before and following therapy is a combination of all of the clinical outcomes, we devised a combined clinical assessment index. Briefly, the dichotomized probing depth, gingival crevicular fluid volume, and gingival index scores were included to create the combined clinical assessment index. Index values extended from zero where all three measures were negative, to "1" when one measure was positive, to "2" when two were positive, and to "3" when all three were positive. The outcome of plotting the percentage of test-positive outcomes for all sites in the study on this basis is shown in Figure 5. Clearly, the percentage of test-positive outcomes increased for all sites and all visits with increasing combined clinical assessment scores, and this was true for all three study locations. At the initial baseline assessment, more than 90% of sites scoring 3 were chairside test-positive, while fewer than 15% scoring 0 were test-positive. Likewise, following treatment at 28 days, far fewer sites scored 3 and of those that did, approximately 55% were test-positive and just over 10% scoring 0 at that timepoint were test-positive. Finally, since the previous analyses looked at associations between outcome of the chairside test and clinical parameters in a cross-sectional manner, although at difmeasures

594

TRIAL OF A CHAIRSIDE PERIODONTAL TEST

J Periodontol June 1996

ferent times prior to and after treatment, they did not provide direct evidence that change in AST at a site was associated with change in other clinical parameters. In order to address that issue directly, we compared the longitudinal change from baseline to day 28 in clinical parameters at an individual site with longitudinal changes in the test outcome at that site. Sites were categorized as improved if they were chairside test-positive at baseline but negative at day 28, and they were categorized as not improved if their chairside test-outcome did not change or went from negative to positive. The average decreases (i.e., improvements) in probing depth, gingival inflammation, and GCF are shown in Figure 6 for sites that improved as determined by the chairside-test versus sites that did not improve. For all three clinical parameters and in all three study centers, sites that went from positive to negative on the chairside test consistently improved more in the other clinical indices, usually by statistically significant amounts, than sites that were not improved as judged by the test.

Average

Decrease In

Probing Depth

Seattle
.0001

Gainesville
-

Chicago
<0.003

0.085

Average

Decrease In

Gingival Index

DISCUSSION

Previously published studies have demonstrated a very high level of association between elevated levels of the enzyme aspartate aminotransferase and active periodontal destruction as manifested by documented loss of peri-

odontal attachment.26-27 Levels of AST also increase with increasing severity of gingival inflammation.28 The present study demonstrates a high degree of association between elevated levels of AST in GCF and various measures of Periodontitis including increasing GCF volume, probing depth, and gingival index score, when compared individually or combined. The strong associations were observed both before and following treatment by scaling and root planing, a known effective form of therapy. Before and after therapy, most sites that were positive for all these measures of disease; i.e., probing depth s 5.0 mm, GCF volume & 40 units, and gingival index score > 2.0, were also test-positive, whereas those sites that were negative for all these manifestations were almost all chairside test-negative. There was a very strong association between combined clinical assessment score and percentage of chairside test-positivity for all timepoints measured, both before and after treatment. Test-positivity was associated as strongly with other measures of Periodontitis as the measures were with each other. Thus the chairside test appears to provide an objective measure of periodontal disease that is a valuable adjunct to other traditional measures, including probing depth and extent of gingival inflammation. While gingival fluid has numerous enzymes and other components that could associate with active tissue destruction,29 assessment of AST activity seems to be particularly promising. AST levels in serum, cerebrospinal fluid, and arthritic joint fluids have been used for many

Seattle
<0.0001

Gainesville
0.085

Chicago
.0001

Average Decrease In GCF

Seattle
<0.0001

Gainesville
"

Chicago
<0.0019

0.004

Figure 6. Improvement in clinical parameters (baseline to day 28) for sites that improved over time on the AST test (positive to negative) versus sites that did not improve (no change, or negative to positive) in patients. Top probing depth; Middle gingival index; and Bottom GCF volume.
= = =

Volume 67 Number 6

MAGNUSSON, PERSSON, PAGE,

ET AL.

595

years in assessing the progress of tissue destruction in various organs and tissues.30 This enzyme is usually confined to cell cytoplasm and is released only upon cell death, and the amount of activity observed generally reflects the extent of cell death and tissue destruction. Sufficient amounts of AST are present in GCF for measurement, and the methods needed for sample collection and activity measurement are simple and reproducible. Although many components in GCF, including AST, are also present in serum, "contamination" of GCF AST with that from serum is sufficiently small that it does not affect the outcome of the measurement significantly. We studied diseased and non-diseased sites in Periodontitis patients, and control sites in periodontally healthy subjects. This was done because non-diseased sites in patients could differ from control sites in periodontally healthy individuals, and the non-diseased sites were subjected to 'the same treatment as diseased sites, while the other control sites were treated by prophylaxis. Differences were observed in that generally the percentage of chairside test-positive sites was greater in nondiseased sites in patients than control sites in healthy subjects. We selected probing depth, severity of gingival inflammation, and GCF volume as measures of periodontal status, in contrast to attachment level and alveolar bone status as manifested on radiographs. These measures were selected because changes in their magnitude do reflect changes in periodontal conditions before and after therapy, while improvement in attachment level and alveolar bone gain would not be expected to change significantly within the 4 weeks post-therapy used in the protocol. The rules we followed for the selecting test teeth and test sites around teeth minimized possible contamination of GCF samples among sites, and ensured that the outcome variables measured were independent from one site to another, except for patient effects. Furthermore, we required probing depth of diseased test sites to be > 5.0 mm to ensure inclusion of true periodontal pockets in contrast to deepened sulci, and ^ 8.0 mm to ensure maximum uniformity from site to site in the thoroughness of scaling and root planing. We used an automated probe to assess probing depth because it employs a controlled probing force and is more sensitive and reproducible than manual probing.22 The age, gender, and race of our combined population were typical of the expected target population, although there were differences especially in race and age across the three geographical samples. The study was designed to evaluate the utility of the test in three different clinical settings in three geographical locations. Although the study design included instruction and standardization in reading the chairside test, there was no attempt to calibrate or standardize the clinical measurements or therapy other than providing the general description of measures to use and therapy to provide. The idea was to evaluate

the test in different clinical settings that included the variations seen in actual clinical practice. In enrolling patients at each location, beyond requiring that they met the inclusion requirements as far as number of diseased sites, there was little effort to standardize the type of patient included. Thus, there was likely more variation across geographical locations in the enrolled patient samples, the clinical measures, and the clinical therapy provided than would ordinarily be the case for a multi-center study of treatment efficacy. This variation across geographical locations is manifested in the proportions of the sites that are test-positive at baseline and in the magnitudes of the associations of the chairside test with other clinical indices. Nevertheless, there are consistencies across geographical locations in the decrease in the population of test-positive sites after treatment, and in the statistically significant positive associations with clinical indices of disease at all points in time and in all locations. To our knowledge, a combined clinical assessment index of the sort we devised and used here has not been used previously in periodontal clinical trials. Our approach was based on the idea that the best assessment of periodontal status would be expected to result from concurrent consideration of all of the clinical data in assessing the association with the objective chairside test values. Indeed, this combination approach is that traditionally used by clinicians in arriving at a diagnosis and treatment plan. It is notable that a very strong association between percentage of chairside test positivity and combined clinical assessment score was observed at timepoints both before and following treatment when all diseased, non-diseased, and control sites were included, while similar correlation also existed when the outcome variables of probing depth, GCF flow, and severity of inflammation were each compared separately with percentage of AST positivity. In addition, it is important to note that sites that went from positive to negative in AST demonstrated significantly greater improvement in other clinical measures over the course of the trial. In summary, the AST test appears to be an effective objective measure which helps distinguish between diseased and non-diseased or healthy periodontal sites in periodontally normal subjects and Periodontitis patients evaluated both prior to and following application of effective therapy. Use of this simple chairside test, when combined with other standard diagnostic procedures, provides an objective measure permitting improved capacity to distinguish between diseased and non-diseased sites, and to better assess and monitor the outcome of therapy.

Acknowledgment This study was supported by Xytronyx, Inc.,

San Diego, CA. It is with great sadness we report the death of Dr. William Clark in February 1995.

596

TRIAL OF A CHAIRSIDE PERIODONT AL TEST

J Periodontol June 1996 A predictor for the progression of periodontal disease. J Clin Periodontol 1986;13:590-596. 18. Jenkins WMM, MacFarlane TW, Gilmour WH. Longitudinal study of untreated Periodontitis. I. Clinical findings. J Clin Periodontol

REFERENCES
1.

probing.

Winkelhoff AJ, van der Veiden U, deGraff J. Microbial succession in recolonization of deep periodontal pockets after a single course of supra- and subgingival debridement. J Clin Periodontol
van

1988;15:116-122.
2. Hammerle CHE Joss A, Lang NP. Short-term effects of initial periodontal therapy (hygienic phase). J Clin Periodontol 1991;18:233239. 3. Pihlstrom BL, Oliphant , McHugh RB. Molar and nonmolar teeth compared over 6 years following two methods of periodontal therapy. J Periodontol 1984;55:499-504. 4. Badersten A, Nilveus R, Egelberg J. Effect of nonsurgical periodontal therapy. I. Moderately advanced Periodontitis. J Clin Periodontol 1981;8:57-72. 5. Badersten A, Nilveus R, Egelberg J. Effect of nonsurgical therapy VIII. Probing attachment changes related to clinical characteristics. J Clin Periodontol 1987;14:425-432. 6. Badersten A, Nilveus R, Egelberg J. 4-year's observations of basic periodontal therapy. J Clin Periodontol 1987;14:438-444. 7. Magnusson I, Lindhe J, Yoneyama T, Liljenberg B. Recolonization of a subgingival microflora in deep pockets. J Clin Periodontol

1988;15:324-330.
19.

Papapanou PN, Wennstrom JL, Grondahl K. A 10-year retrospective study of periodontal disease progression. J Clin Periodontol
1989;16:403-411.

1984;11:193-207.
8. Caffesse R, Sweeney PL, Smith BA. Scaling and root planing with and without periodontal flap surgery. J Clin Periodontol 1986,13: 205-210. 9. Hirschfeld L, Wasserman B. A long term survey of tooth loss in 600 treated periodontal patients. J Periodontol 1978;49:225-237. 10. McFall WT. Tooth loss in 100 treated patients with periodontal disease. J Periodontol 1982;53:539-549. 11. Le H, Silness J. Periodontal disease in pregnancy. I. Prevalence and severity. Acta Odonol Scand 1963;21:533-555. 12. Hancock . Determination of periodontal disease activity. / Periodontol 1981;52:492-499. 13. Lindhe J, Haffajee AD, Socransky SS. Progression of periodontal disease in adult subjects in the absence of periodontal therapy. / Clin Periodontol 1983;10:433-442. 14. Lindhe J, Okamoto H, Yoneyama T, Haffajee AD, Socransky SS. Longitudinal changes in periodontal disease in untreated subjects. J Clin Periodontol 1989;16:662-670. 15. Buckley LA, Crowley MJ. A longitudinal study of untreated periodontal disease. J Clin Periodontol 1984;11:523-530. 16. Socransky SS, Haffajee AD, Goodson JM, Lindhe J. New concepts of destructive periodontal disease. J Clin Periodontol 1984;11:21 32. 17. Lang NR Joss A, Organic T, Gusberti FA, Siegrist BE. Bleeding on

20. Halazonetis TD, Haffajee AD, Socransky SS. Relationship of clinical parameters to attachment loss in subsets of subjects with destructive periodontal diseases. J Clin Periodontol 1989;16:563-568. 21. Persson GR, Alves MEAF, Chambers DA, et al. A multicenter trial of PerioGard in distinguishing between diseased and healthy periodontal sites. I. Study design, methodology and therapeutic outcome. J Clin Periodontol 1995;22:794-803. 22. Gibbs CH, Hirschfeld JW, Lee JG, et al. Description and clinical evaluation of a new computerized periodontal probethe Florida probe. J Clin Periodontol 1988;15:137-144. 23. Fleiss JL. Statistical Methods for Rates and Proportions, 2nd. ed. New York: John Wiley and Sons; 1981;165-168,173-175. 24. DeRouen TA. Statistical models for assessing risks of periodontal disease. In: Bader JD, ed. Risk Assessment in Dentistry. Chapel Hill: University of North Carolina Dental Ecology; 1990;239-244. 25. Goodson JM, Cugini MA, Kent RL, et al. Multicenter evaluation of tetracycline fiber therapy II. Clinical response. J Periodont Res

1991;26:371-379.
26. Chambers DA, Imrey PB, Cohen RL, Crawford JM, Alves MEAF, McSwiggin TA. A longitudinal study of aspartate aminotransferase in human gingival crevicular fluid. J Periodont Res 1991;26:65-74. 27. Persson GR, DeRouen TA, Page RC. Relationship between gingival crevicular fluid levels of aspartate aminotransferase and active tissue destruction in treated chronic Periodontitis patients, i Periodont Res

1990;25:81-87.

28. Persson GR, DeRouen TA, Page RC. Relationship between levels of aspartate aminotransferase in gingival crevicular fluid and gingival inflammation. J Periodont Res 1990;25:17-24. 29. Page RC. Host response tests for diagnosing periodontal disease. / Periodontol 1992;63:356-366. 30. Schmidt E, Schmidt W. Aminotransferases in human pathology and clinical chemistry. In: Christian R Metzler DE, eds. Transaminases. New York: John Wiley and Sons; 1985;586-590.
Send

Biology,
7480.

reprint requests to: Dr. Roy C. Page, Research Center in Oral Box 357480, University of Washington, Seattle, WA 98195for

Accepted

publication

December

4, 1995.