Rationale for combining salivary MS tests and PFRI for prediction of caries risk Like the inflammation induced in gingival soft tissues adjacent to dental plaque, carious lesions that develop on the individual enamel surface beneath bacterial plaque should be regarded as the net result of an extraordinarily complex interplay between harmless and harmful bacteria, antagonistic and synergistic bacterial species, their metabolic products, and their interaction with the many other external (fermentable carbohydrates etc) and internal (saliva and other host factors) modifying factors,
which are discussed in more detail in chapters 2 and 3. In other words, enamel carious lesions develop only on the specific tooth surfaces where thick plaque with a high percentage of acidogenic and aciduric bacteria remains too long¾its “acid slag products” demineralize the underlying tooth surface.
As discussed earlier in the chapter, the quantity of plaque that forms on clean tooth surfaces during a given time represents the net result of interactions among etiologic factors, many internal and external risk indicators and risk factors, and protective factors. This observation was the rationale for the development of the Plaque Formation Rate Index by Axelsson (1984, 1989, 1991). The index, based on the amount of plaque freely accumulated (de novo) 24 hours after PMTC, is described in more detail earlier in the chapter.
An earlier 30-month longitudinal study showed a very strong correlation between development of approximal carious lesions and the level of MS colonization (Axelsson et al, 1987). Other studies have shown that salivary MS counts are correlated to the number of tooth surfaces that are colonized by MS (Lindquist et al, 1989; for review, see Bowden, 1997; Bowden and Edwardsson, 1994; Bratthall and Ericsson, 1994). Therefore, it seems reasonable that MS-positive individuals with high and very high PFRI scores (4 and 5, respectively) should be more caries
susceptible than MS-negative individuals or MS-positive individuals with very low or low PFRI scores (1 and 2, respectively). That is because the total number of the most cariogenic bacteria (MS) should be significantly higher on tooth surfaces in subjects with a PFRI score of 4 or 5 than in subjects with a PFRI score of 1 or 2, if the percentage of MS in their plaque is the same.
Prediction study. In 1984, a large-scale, combined cross-sectional and longitudinal study was initiated with the following objectives:
1. To determine the distribution of the PFRI in a large number of schoolchildren and the distribution characteristics of plaque formation on the individual tooth surfaces in the dentition.
2. To determine whether there is a correlation among the salivary S mutans level, a Cariostat test, and the PFRI score, separately or in combination, and the prevalence of smooth-surface caries.
3. To determine whether a combination of the SM level and the PFRI score is more closely related to caries prevalence than are the variables individually.
4. To determine whether there is any association between the SM level and the PFRI score.
5. To determine the influence of individual factors on the PFRI score (these data were not available at the time of writing and are not included in this chapter).
6. To determine whether caries development can be predicted by a combination of salivary S mutans levels and the PFRI.
All 716 14 year olds in Karlstad, Sweden, were recruited to participate in the study.
Each was given two dental appointments, precisely 24 hours apart. The first appointment comprised the following procedures (for details on methods and materials, see Axelsson, 1989, 1991):
1. Measurement of salivary secretion rate.
2. Salivary S mutans test using the spatula method described by Kohler and Bratthall
(1979).
3. Cariostat test (based on a sample of approximal plaque). The acidogenic capacity of
the sample was estimated from the colorimetric indicator in the test tube showing
different pH values.
4. Gingival index.
5. Plaque index based on disclosed plaque (O’Leary et al, 1972).
6. Professional mechanical tooth cleaning. The subject was instructed to refrain from
all oral hygiene until the appointment scheduled for the following day.
7. Examination of all smooth surfaces for caries, with the following diagnosis: sound
surface, enamel caries, dentin caries, or filled surface. This was recorded as the
subject’s mean decayed or filled surface (DFS) score.
On day 2, 24 hours later, the appointment began with plaque disclosure. The presence
of adherent plaque mesiobuccally, buccally, distobuccally, mesiolingually, lingually,
and distolingually was noted for each tooth. The percentage of tooth surfaces with
plaque was calculated according to the following formula:
Total number of surfaces with plaque x 100
Number of teeth x 6
Each subject’s PFRI was then scored according to the scale described earlier in the
chapter.
At the same visit, a thorough 24-hour dietary history was recorded for subsequent
evaluation of the influence of dietary factors. Many other indicators and factors
possibly related to the PFRI were also evaluated, including gingival inflammation,
salivary levels of glucosyl transferase, agglutinin levels in resting saliva, and oral
hygiene, dietary, and fluoride habits.
Of the 716 children aged 14 years, 667 participated in the PFRI study. Figure 12
presents the frequency distribution of PFRI scores in the population. Figure 17
presents the plaque distribution on various surfaces.
Six hundred fifty-four children, 333 boys and 321 girls, formed the population for
further analysis; for each of these, complete examinations were available. The other
62 children among the 716 originally selected for the study were excluded from the
statistical analysis mainly because of incomplete examinations, antibiotic treatment,
orthodontic bands, refusal to participate, or illness.
Results of prediction study. The examination for caries showed that 70% of the
children had no dentin caries or restorations on smooth surfaces. Of the total number
of lesions on these surfaces, enamel caries constituted more than 80%. Figure 42
shows the mean number of approximal carious lesions per individual in the extreme
groups in relation to the PFRI score, salivary S mutans level, and Cariostat test. The
group with a PFRI score of 5 had, on average, twice as many carious or restored
surfaces as did the group with score 1. The difference between S mutans counts of
fewer than 100,000 CFUs/mL of saliva and more than 1 million CFUs/mL of saliva
was much less marked and of the same order as the differences between Cariostat blue
(low acidity) and greenish yellow (high acidity).
An analysis of caries prevalence (mean DFS) related to different PFRI scores
indicated a thres- hold for caries risk between PFRI scores 2 and 3 (see fig 13), and
this was subsequently confirmed in the longitudinal part of the study, over 5 years
(Axelsson 1989, 1991). For S mutans, this critical level was between 0 and 100,000
CFUs/mL. The Cariostat test offered no additional diagnostic advantages over S
mutans counts.
Table 1 presents the mean values for caries prevalence per individual in relation to
different scores of salivary SM and PFRI. The mean values were low in the
approximately 20% of subjects with 0 SM, irrespective of the PFRI score. On the
other hand, individuals among the 80% of the SM-positive subjects with a PFRI score
³ 3 showed markedly higher caries prevalence than did others. The level of salivary S
mutans appeared to lack significance in this context.
These results were are also confirmed by the lack of correlations between salivary S
mutans levels and PFRI scores (Table 2).
In the cross-sectional part of the study, a score of 0.0 DFS was regarded as a truenegative
value for low caries risk and > 10.0 DFS was chosen as a true-positive cutoff
value for high caries risk because the mean value for the whole group was only 3.5
DFS. The results in Table 2 indicate that the SM level and PFRI score were
independent, ie, there was no association between the variables. The contingency
coefficient was low: 0.16 (upper limit = 0.87).
From the data in Table 3, it can be calculated that a combination of the PFRI and S
mutans gave values of 92.1, 60.9, and 67.3%, for sensitivity, specificity, and
diagnostic power, respectively; For these extremely low- or high-risk groups, the
values obtained were better than were those for sensitivity, specificity, and diagnostic
power for any of the variables alone (PFRI, Cariostat, SM > 0, and SM > 106
CFUs/mL).
Five years later, the children were reexamined. Table 4 shows the mean caries
incidence on the approximal surfaces in the predicted nonrisk (SM-negative) group
and the predicted risk group (SM-positive subjects with PFRI scores 3 to 5). The risk
group developed five times more new approximal carious lesions in dentin per
individual per 5 years than did the no-risk group, in spite of ongoing preventive
programs. The question remains, without answer, how big the difference would have
been without any preventive program.
The 14-year-old age group was selected as particularly appropriate for this
investigation. At this age, smooth-surface caries¾particularly approximal
lesions¾would have developed within the previous 2 years; that is, the caries
prevalence on these surfaces would largely correspond to the caries incidence over
this period, and the subjects would still have a suboptimal number of intact
approximal surfaces at risk. The occlusal surfaces of the molar teeth were deliberately
excluded from the investigation because indications for restoration of these surfaces
vary widely.
Recommendations derived from prediction study. For caries prevention, S mutanspositive
subjects with a PFRI score > 3 should be encouraged to clean their teeth more
frequently than other 14 year olds. These subjects should probably clean their teeth
twice as often as others, that is, morning and evening with an efficient use of fluoride
toothpaste. For patients at extreme risk, cleaning immediately before meals and the
use of fluoride chewing gum as a “dessert” after every meal may be recommended. In
lingual plaque up to 12 hours old, critically low pH values do not occur after rinsing
with a 10% sucrose solution. However, in approximal 3-day-old plaque, there is a
potentially dangerous drop in pH. Rinsing with a sucrose solution does not cause a
critical drop in pH approximally if these surfaces have been cleaned just prior to
rinsing (Imfeld, 1978). Wright et al (1979) demonstrated, in a split-mouth study, that
approximal cleaning once a day gave a reduction of slightly more than 50% in
approximal caries.
Artificial cleaning of the palatal surfaces is, for caries control purposes, unnecessary.
The extremely low plaque formation is probably attributable to the constant friction of
the rough surface of the dorsum of the tongue on these surfaces. Needs-related tests of
patients’ oral hygiene have clearly shown that special efforts should be concentrated
on the linguoapproximal surfaces of the mandibular molars and premolars and the
buccoapproximal surfaces of the corresponding maxillary teeth. Experience has also
shown that the recall visit on day 2 is the ideal occasion for successful introduction of
such individual needs-related oral hygiene practices.
Traditionally, a salivary S mutans count of > 1 million CFUs/mL has been regarded as
a critical value in assessing caries risk. The results of the previously described
investigation do not support this concept. Rather, the critical limit for salivary S
mutans is 0 CFUs/mL. Similar findings were reported in another study of Karlstad
schoolchildren of the same age (Kristoffersson et al, 1986). However, when the same
material was analyzed using approximal tooth surfaces as the unit, a very clear
association emerged between the different levels of S mutans colonization and caries
risk (Axelsson et al, 1987b).
Table 1 could serve as a guideline for selecting nonrisk and risk individuals. A
salivary S mutans test screens out SM-negative subjects (about 25%) as not being at
risk. Of the remaining 75% or so (SM-positive subjects), those with a PFRI > score 3
are selected as risk patients (approximately 20%). From these subjects, an extremely
high-risk group may be further selected: those with a PFRI score of 4 or 5 and an SM
score of 2 or 3 (around 5%). Such a guideline is illustrated in Fig 43.
In general, if the aim of screening is to direct intensive preventive treatment toward
high-risk subjects, a screening procedure offering high sensitivity and predictive value
is preferable both for individual patients and for community dental health planning. A
false-negative diagnosis would deny a subject at risk the benefit of additional
preventive measures. The occurrence of many false-positive diagnoses would make
unnecessary demands on community dental health resources. In this study, Cariostat
tests and high salivary S mutans counts were less reliable as predictors than were
salivary S mutans-positive status and high PFRI (scores 3 to 5).
Evaluation of the influence of several individual risk factors on the PFRI is in
progress. It is anticipated that the three or four main factors will be identified.
Identification of which of these factors dominates in patients with a PFRI score > 3,
should make it possible to design an individual preventive program in which, where
feasible, preventive measures are specifically directed toward minimizing the
influence of these factors.
Individuals with a PFRI score of 1 or 2 were stable over 5 years, while scores in
individuals with a PFRI score of 3 to 5 sometimes changed over time. This
observation indicates that plaque formation rates in individuals with a PFRI score of 4
or 5 can be reduced: Such individuals should be thoroughly evaluated to identify the
factors contributing to their rapid plaque formation. Needs-related preventive
measures could then be introduced.
High-caries rate prediction study. This study was followed up by a 3-year
longitudinal study in Polish children (Axelsson et al, 2000a). Selected 12-year-old
schoolchildren in Warsaw were randomly assigned to a test or a control group. At the
baseline examination, caries prevalence (DFSs), salivary SM counts (Strip-SM), the
PFRI, the Plaque Index (O’Leary, 1967), etc, were recorded. Figures 44, 45, and 46
show the frequency distribution of DFSs, PFRI scores, and Strip-SM scores,
respectively, among all the children at the baseline examination.
Based on the baseline examination, the subjects were assigned to low-risk, risk, and
high-risk groups, according to the following criteria:
1. Low-risk groups: Streptococcus mutans-negative individuals with a PFRI score of 1
or 2 (test: n = 47; control: n = 43).
2. Risk groups: Streptococcus mutans-positive individuals with a PFRI score of 3
(test: n = 30; control: n = 32).
3. High-risk groups: Streptococcus mutans-positive individuals with a PFRI score of 4
or 5 (test: n = 14; control: n = 13).
During the following 3 years, the children in the test group participated in a
preventive program based on professional mechanical toothcleaning at needs-related
intervals but with very simplified methods. For ethical reasons, the children in the
control group were maintained in the regular, simple, school-based preventive
program, based on oral hygiene instructions and topical fluoride administration.
After 3 years, the children in the low-risk test group had developed significantly fewer
new DFSs per individual per 3 years than had the children in the risk test group, who
also had developed significantly fewer new DFSs than both high-risk groups.
However, all the test groups had developed fewer new DFSs than had the control
groups (Fig 47). This study showed that future caries development can be predicted,
even in populations with very high caries incidence, by a combination of salivary S
mutans counts and the PFRI (Axelsson et al, 2000a).
The clinical carious lesion that develops on the tooth surface beneath undisturbed
bacterial plaque represents the net result of an extraordinarily complex interplay
among harmless and harmful bacteria, antagonistic and synergistic bacterial species,
their metabolic products, and their interaction with the many salivary and other host
factors. In other words, dental caries not only is a multifactorial disease but also has a
complicated etiology. It is more difficult to demonstrate a correlation between one
single species of cariogenic bacteria and future caries development in populations
with low caries prevalence than it is in populations with high caries prevalence.
Three different theories for the etiology of dental caries have been proposed: the
nonspecific plaque hypothesis, the ecological plaque hypothesis, and the specific
plaque hypothesis. However, the true etiology is none of these, but a complex
combination of all three processes. The criteria for cariogenic bacteria is that they
must be acidogenic; that is, organic acids are formed as waste products of the
fermentation of carbohydrates. In addition, the bacteria must be aciduric to survive in
the resultant acidic environment (low pH) in the plaque and the carious lesion. Even
bacteria-producing enzymes, which destroy the organic components of root cementum
and dentin, may be involved in the development of root caries and dentin caries.
The basic principle of the nonspecific plaque hypothesis is that thick plaque on the
tooth surface, if left undisturbed for long periods, allows the total amount of acid
produced within this plaque to initiate the development of a carious lesion.
Accordingly, very high plaque formers (PFRI scores 4 and 5) would be expected to
develop more carious lesions than low or very low plaque formers (PFRI scores 1 and
2), if the standards of oral hygiene and the composition of the microflora were the
same in the two groups.
In addition, carious lesions tend to develop on the particular tooth surfaces on which
most plaque reaccumulates between toothcleaning procedures (mesiolingual and
distolingual surfaces of the mandibular molars and mesiobuccal and distobuccal
surfaces of the maxillary molars), and, in a toothbrushing population, where the
toothbrush has limited access (the approximal surfaces of the molars and premolars).
This is confirmed in studies on the pattern of caries prevalence in different
populations.
Not only the frequency but also the main target of needs-related oral hygiene
procedures should be based on the score and the pattern of the PFRI. Because the
quantity of plaque that forms on clean tooth surfaces during a given time represents
the net result of interactions among etiologic factors, many internal and external risk
factors, and protective factors, future research should be directed toward methods of
identifying the major factors causing rapid plaque formation in the individual patient.
If possible, these factors should be reduced or eliminated.
The ecological plaque hypothesis is based on the principle that a change in a key
environmental factor (or factors) will trigger a shift in the balance of the resident
plaque microflora and this might predispose a site to disease. For example, the thicker
the plaque, the less accessibility there is for the saliva to dilute and buffer the organic
acids formed by fermentation of carbohydrates by acidogenic plaque bacteria. As a
consequence, the pH will continuously decrease the more fermentable carbohydrates
(for example, sucrose) are supplemented and the longer the plaque remains
undisturbed. The lowered pH (< 5) in the plaque will promote a shift of the
composition of the plaque bacteria toward an increased number and assortment of
acidogenic and aciduric species such as the cariogenic mutans streptococci and
lactobacilli.
According to this hypothesis, the strategy for caries prevention should be to maintain
a high pH on all tooth surfaces (microenvironments) by frequent removal of plaque,
thereby limiting the thickness of undisturbed plaque; and by reduction of the “sugar
clearance time,” through a diet that stimulates saliva, and use of supplementary
fluoride chewing gum as a dessert directly after every meal, particularly in patients
with reduced salivary flow. Future research should focus on efficient methods for
achieving homeostasis of dental plaque, maintaining high pH (> 6).
There is abundant support for the so-called specific plaque hypothesis, which
proposes that some specific species of the plaque flora should be regarded as major
pathogens in the etiology of dental caries. The most significant of these bacteria are
some of the mutans streptococci. This group includes seven species, although two, S
mutans and S sobrinus, are most closely associated with dental caries in humans. In
longitudinal studies, particularly at surface level, a very strong correlation has been
shown between S mutans and development of caries lesions on smooth surfaces.
However, in populations with low caries prevalence, a correlation between various
levels of salivary S mutans counts and caries incidence seems to be less significant:
The threshold would appear to be S mutans-negative or S mutans-positive status.
The second genus closely associated with caries etiology is Lactobacillus, commonly
isolated from the dentin in both coronal and root caries lesions, its main habitat in the
mouth. Compared to Streptococcus, Lactobacillus has been less extensively studied.
Actinomyces odontologica, Actinomyces naeslundii, and other species of the MS
group are also associated with the etiology of dental caries but are considered to be
less cariogenic than S mutans, S sobrinius, and Lactobacillus.
Future research involving DNA probes and so-called genetic-fingerprinting
techniques will result in tools for evaluation of (1) how S mutans is transmitted among
individuals; (2) how stable the oral population is; (3) how many types of S mutans an
individual carries; (4) if particular clonal lines of S mutans are more virulent
(cariogenic) than others; and (5) if individuals with higher levels of carious activity
carry particular types of S mutans.
