|Increase in Cariogenic Bacteria after Initial Periodontal Therapy
For figures, tables and references we refer the reader to the original paper.
Epidemiological surveys have shown that root-surface caries as well as periodontitis are common in adults (Hellyer et al., 1990; Papapanou, 1996). Several longitudinal studies reported a significant correlation between the number of Streptococcus mutans species (and, to a lesser extent, of lactobacilli) in the saliva and the prevalence/incidence of root caries (Ellen et al., 1985; Emilson et al., 1988; Ravald et al., 1993). In spite of the complexity of the pathogenic flora, mutans streptococci and lactobacilli seem to be useful target micro-organisms for monitoring root caries in clinical practice (Klock et al., 1990; Hunt et al., 1992; Ravald, 1994). Some studies linked root caries with periodontal disease, even in patients with good to excellent periodontal conditions after therapy (Ravald and Hamp, 1981; Keltjens et al., 1987; Reiker et al., 1999). Two recent papers (Quirynen et al., 1999a; Van der Reijden et al., 2001) reported relatively high detection frequencies for S. mutans, both supra- and subgingivally, after periodontal therapy.
In the present study, we undertook a longitudinal examination of whether initial periodontal therapy causes an intra-oral microbial shift, both supra- and subgingivally, from a periopathogenic to a more cariogenic flora, and if so, whether such a shift can be prevented. [The clinical observations will be discussed in a separate paper (Quirynen et al., 2005).]
MATERIALS & METHODS
Seventy-one Caucasian volunteers (from 30 to 75 yrs of age; mean, 48 yrs; 31 females; 18 smokers) volunteered for this single-blind study. They suffered from severe periodontitis (at least 2 multi-rooted and 3 single-rooted teeth in the first quadrant, with at least 6 sites showing a probing depth of 7 mm or more). Radiographic evidence of severe bone loss (≥ ½ the root length) was present. All subjects were in good general health, and none of them had used antimicrobial agents 4 mos prior to the study. After an explanation of the therapy, all participants signed an informed consent. The protocol was approved by the Clinical Trials Committee of the University Hospital.
A clinician who was informed about the baseline clinical data (but not about the content of the treatment strategies) randomly allocated the participants consecutively to one of the following groups:
a negative control group (NC, n = 15)—scaling and root planing (quadrant by quadrant) at two-week intervals (thus, 6 wks between first and last quadrants), without adjunctive products;
a control group (FRP, n = 14)—one-stage full-mouth root planing without the use of adjunctive products (Quirynen et al., 1999b);
or one of the 3 positive control groups (n = 14 per group)—one-stage full-mouth disinfection (Quirynen et al., 1995), followed by the use of different antiseptic mouthrinses: either chlorhexidine 0.2% (Corsodyl®, GlaxoSmithKline, Genval, Belgium) for 2 mos (CHX), or amine fluoride/stannous fluoride (AmF/SnF2, Meridol mouthrinse®, GABA International AG, Switzerland) for 2 mos (F), or chlorhexidine 0.2% for 2 mos and AmF/SnF2 for another 6 mos (CHX+F) (Table 1⇔).
Treatment Strategy in Different Intra-oral Niches for Each Group
For all except the NC group, scaling and root planing were completed in 2 sessions within 24 hrs (starting with the lower jaw). For the positive control groups, mechanical debridement was combined with a chair-side chlorhexidine application (Quirynen et al., 1995). All groups were relatively comparable with respect to age, smoking, and degree of periodontal destruction (Table 2⇔). The 4 deepest pockets around single- and multi-rooted teeth in the right maxillary quadrant of each patient were selected as experimental sites (thus, 8 sites per subject).
Descriptive Statistics of Patient Population (n = 71) Sorted by Treatment Strategy
Immediately after the scaling and root planing procedures, and at the end of months 2, 4, and 8, one clinician (who was unaware of the treatment strategy or the previous registrations) recorded the probing depth for the selected reference pockets.
Just prior to the first session of scaling and root planing (baseline), and after 2, 4, and 8 mos, microbial samples were taken from:
approximal supragingival plaque between lateral incisor, and between central incisor and canine (single-rooted)—and, consequently, between both premolars, and first and second molars (multi-rooted), respectively—by means of sterile wooden wedges (Tenovuo et al., 1990);
subgingival plaque (pooled samples from reference pockets around single- and multi-rooted teeth, respectively), by means of 4 paper points (Roeko®, Roeko, Langenau, Germany) per site;
the dorsum of the tongue, by means of a sterile cotton swab (Biomerieux S.A., Montalieu-Vercieu, France) (Danser et al., 1994); and
from unstimulated saliva (0.5 mL), by means of a sterile syringe (B. Braun®, Melsungen, Germany).
The samples were dispersed in Reduced Transport Fluid (Syed and Loesche, 1973), homogenized by being vortexed for 30 sec, transferred to the Laboratory of Microbiology, and processed in under 24 hrs. The samples were cultured under aerobic and anaerobic conditions in specific and α-specific media, including a TYCSB medium for the isolation of S. mutans (Van Palenstein-Helderman et al., 1983) and a Rogosa medium for lactobacilli (Rogosa et al., 1951; Schüpbach et al., 1995). Details concerning the growth conditions, colony selection, pure culturing, and final identification of specific species (cariogenic species as well as periopathogens) have been summarized previously (Quirynen et al., 1999a). The microbiological evaluation was performed blindly.
A linear mixed model was fitted. The repeated character of the data was modeled in the error matrix, of which the Akaike's Information Criterion showed that a compound error structure was the best fit. Residual analysis required the data to be log-transformed. P-values for the multiple comparisons were corrected for simultaneous hypotheses according to the Tukey-Kramer method. For all data, corrections for differences between and among groups at baseline were carried out, even though these differences were never statistically significant. For statistical significance, P-value was set at p ≤ 0.05.
Detection Frequencies and Numbers of Cariogenic Species and Periopathogens
The changes in the detection frequencies of S. mutans within each treatment group were nearly identical for all sample locations (i.e., subgingivally, supragingivally, tongue, and saliva) (Table 3A⇔). However, clear differences could be observed between therapies. In the NC and FRP groups, the detection frequency clearly increased over time, with largest increases for the NC group. The F group showed a slight reduction for month 2 only and a recurrence afterward. In the CHX group, S. mutans was not detected at month 2, but recovered partially afterward. In the CHX+F group, S. mutans was not detected for the entire observation period. When the number of colony-forming units (CFU) for positive sites was considered over time (comparison between months 0 and 8), only the CHX+F group showed obvious changes (1 to 2 log reductions) for all sample locations (Table 3B⇔). In the other groups, the changes remained small (< 1 log value), except for some clear reductions in the CHX group.
Mean Detection Frequency (number of positive sites on 14 [15 in the NC group] examined sites) for Cariogenic Species and Periopathogens in Samples from Supragingival and Subgingival Plaque, Tongue, and Saliva, at Baseline and after 2, 4, and 8 Mos
Mean Number of Colony-forming Units (CFU) in Samples from Supragingival and Subgingival Plaque, Tongue, and Saliva, at Baseline and after 8 Mos, and per Treatment Group
For the lactobacilli, the changes within a treatment group were similar for all sample locations. The changes over time were less impressive for all treatment strategies (Table 3A⇔). Only for the CHX+F group could a long-term reduction of up to 50% be maintained over the eight-month period. For the other groups, only temporary reductions were seen, except for the NC group, for which only a few changes were recorded. The number of CFU for this species remained unchanged for the subgingival, tongue, and saliva samples for all treatment strategies (Table 3B⇔). Only supragingivally were some reductions above 1 log value detected, but clear tendencies were not seen.
The detection frequency for periopathogens in subgingival plaque samples decreased over time in all groups, with the largest reductions for the CHX+F group (Table 3A⇔). The reductions were most obvious and longstanding for P. gingivalis, but were negligible for F. nucleatum. The number of CFU for black-pigmented bacteria (Table 3B⇔) in pockets around single-rooted teeth showed a significant reduction (p ≤ 0.05, borderline for NC) over time for all treatment groups. Also, for multi-rooted teeth, reductions were observed up to month 8, but these reductions reached a level of statistical significance only for the CHX and the CHX+F groups, respectively. For the FRP and F groups, the reduction was statistically significant up to month 2 only.
Numbers of CFU for Anaerobic Bacteria
Compared with baseline, the FRP, CHX, and CHX+F groups showed significant reductions in the numbers of CFU in supragingival plaque (p always < 0.01; up to month 2 for CHX, and up to month 8 for FRP and CHX+F), for both single- and multi-rooted teeth (Table 3B⇔). For the subgingival flora around single- and multi-rooted teeth, significant reductions (p < 0.01) were observed in all groups up to month 8, except in the NC group. The changes in the numbers of anaerobic species in tongue samples were generally negligible. Compared with baseline, only CHX+F (p < 0.001, entire period) and CHX (p = 0.002; month 2) were successful in reducing the number of anaerobic bacteria. No significant changes were detected in the saliva.
Numbers of CFU for Aerobic Bacteria
Compared with baseline, only the FRP, CHX, and CHX+F groups showed significant reductions in the numbers of CFU in supragingival plaque (p always < 0.01; up to month 2 for CHX and month 8 for FRP and CHX+F), from both single- and multi-rooted teeth (Table 3B⇔). For the subgingival flora around single- and multi-rooted teeth, significant reductions (p < 0.01) were observed for all groups, but with the smallest changes for the NC group. These changes could be maintained up to month 8 for all treatment strategies. The changes in the numbers of aerobic species in tongue samples were negligible. Compared with baseline, only CHX+F (p < 0.01, entire period) and CHX (p = 0.002, month 2) were successful in reducing the numbers of aerobic bacteria. No significant changes were detected in the saliva.
Our hypothesis—that an initial periodontal therapy could result in a shift from periopathogens toward cariogenic species—seems confirmed by the significant increase in the detection frequency for S. mutans in the NC and the FRP groups in which no antiseptics had been used. This observation is in accordance with our previous pilot observations (Quirynen et al., 1999a) and with the data from a cross-sectional study (Van der Reijden et al., 2001), both suggesting a similar shift. Surprisingly, this shift occurred not only supragingivally, but also in all other sample locations (tongue, saliva) and even subgingivally. A clear explanation for the latter cannot be determined from this study. One might speculate that the changes in microbial composition after periodontal therapy and/or the healing of the periodontium resulted in more favorable growth conditions for S. mutans. Because several longitudinal studies have reported a positive relationship between salivary counts of S. mutans and the incidence and/or prevalence of root caries (Ellen et al., 1985; Emilson et al., 1988; Ravald et al., 1993), this microbial shift should be taken into consideration when the prevention of root-surface caries is considered.
The changes in detection frequency of lactobacilli were less impressive. In contrast to S. mutans, this species showed a decrease in detection frequency for all groups.
Analysis of the data from the antiseptic groups gave some indications of how this overgrowth of S. mutans could be prevented. It was shown that chlorhexidine is a very potent antimicrobial against S. mutans. In both groups where chlorhexidine was used, S. mutans was decreased below detection levels in all sample locations within the oral cavity, at least as long as the antiseptic was used. This finding is in agreement with those from several other studies that evaluated only saliva and/or supragingival plaque samples (for review, see Emilson, 1994). The beneficial impact of amine fluoride/stannous fluoride on the suppression of S. mutans is in accordance with findings reported in several papers on the suppression of S. mutans by stannous fluoride alone (Zickert et al., 1987; Wallman et al., 1998). Moreover, other clinical studies clearly showed significant reductions of S. mutans after subjects rinsed with a stannous-amine fluoride rinse (Meurman et al., 1989).
The differences between the CHX and CHX+F groups seem to indicate that the subgingival colonization by S. mutans is influenced by the supragingival area. The only difference between these groups, after month 2, was the use of amine fluoride/stannous fluoride in the CHX+F group for the remaining 6 mos. The mouthrinses could have an impact only on the supragingival plaque, since they cannot penetrate subgingivally (Eakle et al., 1986). Thus, in the CHX group, S. mutans probably could re-establish itself supragingivally after cessation of the chlorhexidine, followed by subgingival colonization. This hypothesis must be confirmed by new studies designed for this purpose.
The improved supragingival plaque control could also have played a role, although the impact of the supragingival environment on subgingival recolonization after periodontal therapy is still controversial (Petersilka et al., 2002). It is obvious, however, from our data, that the groups with significantly better plaque control (CHX+F and CHX, the latter up to month 2 only; Quirynen et al., 2005) harbored the lowest numbers of S. mutans.
The relative proportions of S. mutans in the supra- and subgingival samples were quite similar, indicating that this species can grow in habitats providing different conditions. Surprisingly, the microbial load on the tongue did not show major changes after periodontal therapy, including toothbrushing, except when the patients rinsed with chlorhexidine. Probably the extreme roughness of the tongue, with its deep fissures, prevented significant removal of bacteria, thus enhancing regrowth. This observation is in accordance with our previous observations (Quirynen et al., 2004).
In comparison with the 4 other treatment modalities with a more global approach, the inferiority of the NC group in reducing the numbers of anaerobic species, as well as specific periopathogens subgingivally, confirms the importance of one-stage, full-mouth disinfection as we previously reported (Quirynen et al., 1999b). The fact that the NC group, together with the CHX+F group, had slightly more severe (statistically insignificant) periodontal destruction initially, when compared with the other groups, does not explain these differences.
The observation period of this study was too short for significant clinical differences in root caries prevalence to be determined. The microbiological observations from this study, together with the data from previous papers, suggest the need for a caries-preventive program after periodontal therapy.
This study was supported by a grant from GABA SA International, Switzerland, and by a grant from the Catholic University, Leuven (OT/03/52).