Systemic Doxycycline as an Adjunct to Scaling and Root Planing in Diabetic Patients With Periodontitis

A Systematic Review and Meta-analysis

Kenneth Chou Hung Yap; Shaju Jacob Pulikkotil


BMC Oral Health. 2019;19(209) 

In This Article


Results of 3-month Bleeding On Probing (BOP), Clinical Attachment Levels (CAL), Pocket depth (PD) and HbA1c levels are taken due to the limited number of studies done for 6 months and above for each of these clinical parameters. Systematic reviews like Grellmann AP et al. showed that the use of doxycycline did not significantly improve probing depths whereas Ronaldo Lira Junior et al. and Wang-Tze Fang et al. showed that adding doxycycline to periodontal therapy with SRP does not significantly improve metabolic control in patients with type 2 diabetes mellitus and chronic periodontitis.[13–15] Although systemic doxycycline inhibits metalloproteinase activity and also has antimicrobial effects which helps in reducing inflammation of periodontal tissues,[15] HbA1c levels did not statistically improve after 3 months after administration of systemic doxycycline as an adjunct to SRP in our study. In terms of periodontal status, CAL, PD and BOP generally showed no significant improvement when systemic doxycycline is used as an adjunct to SRP, although 1 study showed an a favorable increase in CAL when systemic doxycycline is used.[16] On the other hand, another study revealed a significant improvement in probing depths when systemic doxycycline is used as an adjunct to SRP.[7] This study that we have done also proves the current systematic reviews that systemic doxycycline as an adjunct to scaling and root planing (SRP) in diabetic patients with periodontitis does not show any benefit versus SRP alone in boosting metabolic control (HbA1c) as well as periodontal status in terms of clinical attachment levels (CAL) after 3 months of treatment.

The incorporation of doxycycline in SRP can have a higher chance to alter the pathogenic bacterial group, thus making it a more favorable environment for stable recolonization in the long run in gingival pockets which are recently scaled. This consequently creates a stable biofilm community which can also be found in individuals with no periodontal disease.[17–19] Nevertheless, bacterial resistance is an issue when there is broad usage of systemic antibiotics such as doxycycline used in periodontics.[20] Therefore, every clinician should weigh the pros and cons of prescribing these antibiotics to prevent bacterial resistance and other undesirable side effects of antibiotics. At this moment of time, about 75% of patients with chronic periodontitis showed that periodontal bacteria such as Aggregatibacter Actinomycetemcomitans are now resistant to at least one of the common antibiotics which includes doxycycline.[21] Thus, their prescription is only considered in clinical situations where the benefits of prescribing outweighs the undesired effects of antibiotics.

As shown in our study, there is no significant difference in CAL when systemic doxycycline is used as an adjunct to SRP and borderline statistical difference for PD. Although CAL is the gold standard for diagnosis of periodontitis, probing depths are also as important as CAL because the presence of deep pockets increases the risk of development of periodontal disease which will signals the need for more treatment.[22,23] Studies from Sokransky and Haffajee et al. have demonstrated that more bacteria are found in deep pockets.[17] Increase in PD shows inflammatory changes, and clinical attachment levels will increase when inflammation of periodontium decreases and long junctional epithelium starts to develop.[24]

Wang et al. have evaluated the efficacy of SRP with systemic doxycycline on reduction of HbA1c in diabetic patients in which 3 trials were included in the meta-analysis and there was no significant change in HbA1c levels by using systemic doxycycline as an adjunct to SRP.[25] As for Ronaldo Lira Junior et al's study, they had 3 extra trials and found results complementing Wang et al's study, proving that systemic doxycycline is not beneficial when used as an adjunct to SRP over SRP alone on in terms of reducing HbA1c levels.[14] These systematic reviews and meta-analyses are similar to our study which also shows that systemic doxycycline showed no significant benefits when used as an adjunct to SRP as compared to just SRP alone [− 0.13 (− 0.41, 0.15)].

In terms of heterogeneity of the studies involved, the hetereogeneity percentage for BOP, CAL, HbA1c and PD are 8, 36, 3 and 48% respectively. According to Cochrane, 0–40% indicates low heterogeneity and 40–60% indicates moderate heterogeneity.[29] Results BOP and HbA1c are categorized as low heterogeneity, whereas CAL has a low to moderate heterogeneity. On the other hand, PD is categorized as moderate heterogeneity. The moderate heterogeneity of PD might be due to difference in dose and duration of systemic doxycycline as stated in the characteristics table in Table 1. However, since the primary outcomes are CAL and HbA1c which are calculated as low heterogeneity, it would not affect the results of the meta-analysis significantly.

From the GRADE assessment, certainty of the results of this meta-analysis is moderate which indicates that further research is will most probably have a major effect on our confidence intervals in the estimate of effect and may change the estimate. Certainty was moderate due to the moderate overall risk of bias in the 6 studies involved. The importance of this meta-analysis results is stated as "critical" because this systematic review aims to avoid the excessive use of systemic doxycycline without any tangible benefit, reducing the risk of antibiotic resistance against systemic doxycycline in the long run.

According to the funnel plots in Figs. 8, 9, 10 and 11, the studies involved are evenly distributed among the four funnel plots, suggesting that there was no publication bias. However, according to British Medical Journal (BMJ), tests for uneven distribution of studies in funnel plots are not used when there are 9 or less studies involved in any meta-analysis due to the low test power to differentiate substantial asymmetry from pure coincidence.[26]

Figure 8.

3 Month bleeding on probing (BOP)

Figure 9.

3 Month clinical attachment levels (CAL)

Figure 10.

3 Month glycated hemoglobin (HbA1c)

Figure 11.

3 Month pocket depth (PD)

The main reason a meta-analysis is done is due to the fact that it has the superiority of having a better statistical power and its value as an evidence-based resource able to extrapolate confirmatory data analysis across the affected population. Nonetheless, the inadequate number of evidence in the present meta-analysis may not add to the current clinical guidelines with confidence. The limitations of this systematic review includes a lack of timeframe which only includes a 3-month follow-up after systemic doxycycline is used.[7,10,11,16,27,28]

Besides that, another issue is the allocation concealment in which the risk of bias is fully unclear. No information was given on the allocation concealment, which affects the selection bias of all the six studies. Another concern is that the definition used in the present study lacked radiographic evidence of bone loss and were single-point in time measurements, although it would be a difficult task in epidemiological studies. Also, randomization of participants were not stated for three of the studies, namely Gaikwad et al., O'Connell et al. and Al-Nowaiser et al.,[10,16,27] which might lead to selection bias. Strength of the reviews include prior protocol registration, subgroup and sensitivity done, grade shows the quality of evidence.