Outcome of endodontic treatment – the elephant in the room

Outcome data from well designed and conducted clinical trials are essential to evaluate the safety and effectiveness of endodontic procedures from vital pulp therapy to microsurgical re-treatment. Such information allows clinicians to discuss risks, success rates, prognostic factors and alternative management options with patients, and only with such information can patients make properly informed decisions about their care. It is not uncommon for endodontic disease to be clinically asymptomatic (Kirkevang & Hørsted-Bindlev 2002, Huumonen & Ørstavik 2002), and radiological assessment is therefore essential to objectively determine its presence for both untreated and previously treated teeth (Patel et al. 2012). Dentists have traditionally relied on a combination of clinical evaluation and periapical radiographs as their reference standard to evaluate the presence or absence of endodontic disease. Apical periodontitis usually manifests radiographically as a periapical radiolucency, due to reduced mineral density of the affected bone in response to localized inflammation. In most circumstances, this is promoted by microbial infection within the root canal system (Bender 1982, Patel et al. 2012). The absence of a radiolucency indicates the absence of apical periodontitis, and by inference, the absence of root canal infection; a key determinant of treatment success (European Society of Endodontology 2006). The advent of 3D imaging systems such as cone-beam computed tomography (CBCT) has complicated matters by identifying more periapical radiolucencies than traditional methods, though uncertainty remains about the healing dynamics of lesions identified and monitored by CBCT, and their histological status at intervals after treatment. Increasingly, tooth survival has become established as an alternative outcome measure. Survival data identify teeth that are lost due to symptoms or risks that are unacceptable to patients, but accepts the presence of teeth that may have signs, symptoms and radiographic features of apical periodontitis, but which are judged as acceptable by patients (Friedman et al. 2003, Doyle et al. 2006). This gradual shift in focus begs the question whether asymptomatic apical periodontitis is an important disease, and whether persistent radiolucencies identified on CBCT images are associated with significant risks of local flare-up or systemic consequences, and if so, whether particular patient groups are at risk. These uncertainties become increasingly relevant as populations age with increasing frailty, vulnerability and co-morbidities caused by medical conditions, medical treatment and the effects of ageing. The aim of this editorial is to appraise the use of periapical radiographs and CBCT in determining endodontic treatment outcomes and highlight a number of important questions (elephants in the room) that must be addressed if we are to provide more complete data to inform clinical decision-making. For more than 100 years, periapical radiographs have been the standard method to detect the presence or confirm the absence of apical periodontitis, with numerous studies showing high levels of endodontic treatment success (de Chevigny et al. 2008, Ng et al. 2011, Patel et al. 2012). Despite the simplicity and objectivity of traditional radiographic methods, and the acquisition of readily quantified data, the limitations of 2D images are well recognized (Bender & Seltzer 1961, Brynolf 1967, Kanagasingam et al. 2017a). Periapical radiolucencies may be missed and/or under-estimated due to the superimposition of overlying cortical bone and/or geometric distortion. Despite this, not one outcome study using periapical radiographs has acknowledged their limited sensitivity or questioned the accuracy of the outcome assessment (Wu et al. 2009). Over the last two decades, the use of small field of view (FOV) CBCT has steadily increased in endodontics (Setzer et al. 2017). Clinical studies have shown 11–39% and 20–34% increased detection of periapical radiolucencies with CBCT compared to periapical radiographs in teeth treatment planned for primary and secondary endodontic treatment respectively (Table 1). Live animal and human cadaver studies have confirmed the enhanced accuracy of CBCT in detecting periapical radiolucencies that may not be visible on periapical radiographs (Paula-Silva et al. 2009, Kanagasingam et al. 2017b, Kruse et al. 2019). The recent observation from a human cadaver study that approximately 20% of roots with canal fillings and histologically ‘healthy’ periapical tissues may have been judged as diseased on CBCT (Kruse et al. 2019) provides an incomplete picture as the original periapical status of the teeth examined was unknown. This study does, however, point to the potential dangers of over-treating previously root filled teeth that have radiolucencies identified on CBCT images. Lesions may simply take longer to resolve completely when monitored by 3D imaging techniques. In an outcome study conducted 12 months after primary root canal treatment, Patel et al. (2012) reported that 87% of periapical radiolucencies had healed (i.e. complete resolution) on PR, compared with 62.5% on CBCT images. Ninety-five percent of periapical radiolucencies were judged to be healing (i.e. periapical radiolucencies reducing in size) on periapical radiographs, compared with 84.7% on CBCT images. For teeth with no preoperative apical periodontitis, CBCT images revealed periapical radiolucencies in 17.6% of cases, compared with just 1.3% on periapical radiographs. The difference was highly significant. However, for teeth with preoperative apical periodontitis, the difference was not statistically significant, with 13.9% failures observed on CBCT images and 10.4% on periapical radiographs. Using the same methodology to assess the 12-month outcome of root canal re-treatment, 93% of periapical radiolucencies were judged to be healing on periapical radiographs, compared with 77% on CBCT. Healed rates were 78% with periapical radiographs and 61% with CBCT (Davies et al. 2016). Though the recall intervals were short in both studies, there was a higher detection of radiographic signs of post-endodontic disease in molar teeth with CBCT. It is again uncertain how outcomes would change following longer-term recall. Recently, the 12-month data from three primary and secondary root canal treatment outcome studies involving 354 teeth, identified signs of healing in 90% of anterior, premolar and molar teeth on periapical radiographs (Al Nuaimi et al. 2018). These figures are comparable to previous studies (de Chevigny et al. 2008, Ng et al. 2011). For anterior and premolar teeth, there was almost perfect agreement on lesion detection with CBCT and periapical radiographs (90%), but only 75% of molars were judged to have healed or healing lesions on CBCT (Al Nuaimi et al. 2018). This difference in outcome was primarily associated with the larger number of pre-treatment (baseline) radiolucencies detected with CBCT (188 with periapical radiographs vs 264 with CBCT in a cohort of 346 teeth). The increased number of radiolucencies associated with molar teeth observed by CBCT may indicate delayed healing compared to observations on periapical radiographs. However, when data from two outcome studies on primary and secondary root canal treatment in posterior teeth are pooled (Patel et al. 2012, Al Nuaimi et al. 2017, 2018), there is a highly significant difference (Fisher’s exact test two-tailed P = 0.0063) between the number of new, enlarging and unchanged radiolucencies detected with CBCT images compared to periapical radiographs, indicating increased radiographic signs of post-endodontic disease detected by CBCT (Table 2), at least at 12-month follow-up. These differences in detection rates between periapical radiographs and CBCT in the posterior region are largely attributed to anatomical noise resulting in periapical radiolucencies being hidden from detection on periapical radiographs. What remains uncertain once again is the progress of such lesions over longer-term recall intervals. New radiolucencies Enlarged radiolucencies Unchanged radiolucencies Resolved radiolucencies Reduced radiolucencies No radiolucencies before treatment and at recall Based on current evidence, periapical radiographs combined with clinical evaluation of signs and symptoms is appropriate for routine outcome assessment in anterior and premolar regions, both in clinical practice and in clinical trials. However, given the obvious lack of sensitivity of periapical radiographs in detecting radiolucencies in the molar region, it may be questioned whether researchers should still advocate their use as the gold standard method of assessing treatment outcomes in clinical trials. At the present time, small FOV CBCT cannot be recommended for the routine assessment of endodontic treatment outcomes. However, in specific situations it may be relevant, particularly in the management of posterior teeth, for example where apical periodontitis is suspected but not revealed on periapical radiographs. In vital pulp therapy, Hashem et al. (2015) examined molar teeth with reversible pulpitis by periapical radiographs and CBCT pre-treatment and 12 months post-treatment. Diagnosis and treatment planning were determined solely on the basis of clinical examination and assessment using periapical radiographs, and clinicians gained access to the pre-treatment CBCT only when they assessed the outcome of treatment. At 12-month follow-up, use of periapical radiographs identified 90% of cases was successful, compared to 65% with CBCT, judged by the presence or absence of a lucency. For teeth with preoperative apical periodontitis identified by CBCT, 37% were successful at 12 months, compared with 84% where no lesion had been identified by CBCT. The exclusion of teeth with such preoperative CBCT-detected radiolucencies from a successive clinical trial on deep caries management was associated with improved success rates (Ali et al. 2018). These findings suggest that CBCT may be helpful for the assessment of teeth being considered for vital pulp therapy to detect potential ‘hidden’ radiolucencies which may not be detected on periapical radiographs and which may reflect a less advantageous pulpal status. There is growing evidence that CBCT imaging improves diagnostic accuracy, decision-making and/or change/enhances treatment planning (Ee et al. 2014, Davies et al. 2016, Rodríguez et al. 2017a,b). A recent study has also revealed that the additional information gained from CBCT images reduced the stress levels of a cohort of clinicians treating molar teeth and also improved the radiographic quality of root fillings; a factor that has long been associated with treatment outcome (Patel et al. 2019a). However, further research is needed to assess whether such observations are true for all clinicians, especially those with no previous experience of working with CBCT. Uncertainties also remain on the impact of preoperative CBCT scans on improving clinical safety and efficacy, and how these factors impact on treatment outcomes in terms of periapical health, tooth survival, reduced disease risk and patient satisfaction. The benefits of CBCT must always be balanced with concerns for radiation protection and this may be enhanced by guidelines to minimize dose (European Society of Endodontology 2019). The effective dose of small FOV (<5cm) CBCT scans can be 3–20 times higher when compared to periapical radiographs depending on the scanner used, exposure parameters and the resolution setting. The quality of reconstructed CBCT images varies considerably between small FOV scanners, which in turn has an impact on the diagnostic accuracy of detecting apical periodontitis (Patel et al. 2019b). Radiation protection agencies and/or maxillofacial radiology societies should be encouraged to study the impact of image quality and artefacts on diagnostic accuracy and ultimately provide clear and evidence-based guidelines on dose reduction protocols and image quality for small FOV CBCT scans. The European Society of Endodontology Quality Guidelines (European Society of Endodontology 2006) emphasize the role of periapical radiographic healing in determining treatment success. When a periapical radiolucency has diminished but not completely resolved within 4 years, the outcome is considered ‘unfavourable’, requiring discussion of further treatment. Such judgements are based on the interpretation of periapical radiographs. The healing dynamics of periapical radiolucencies monitored by CBCT are not addressed, and the likely presence of more persistent periapical radiolucencies at 4 years compared with those identified on periapical radiographs begs the question whether 4 years is an appropriate cut-off for monitoring with CBCT or if longer periods of monitoring are indicated before considering further intervention. Although the accuracy of CBCT in detecting apical periodontitis associated with endodontically treated and non-endodontically treated teeth with moderate to severe inflammation is around 90%, the accuracy for detecting mildly inflamed or healthy periapical tissues is lower in root filled teeth (70-80%) (Kanagasingam et al. 2017b, Kruse et al. 2019). The pathological nature of such radiolucencies and the risks they may present to healthy and vulnerable individuals remains uncertain, with only a limited number of cadaveric studies correlating radiolucencies identified by CBCT with histopathological findings. Periapical radiolucencies that persist but are healing slowly may reasonably be defined as ‘healing’ rather ‘unfavourable’ as defined by current guidelines when judged 4 years after treatment (European Society of Endodontology 2006). Definitions of outcome are also relevant to patient decision-making, notably between endodontic treatment and implant replacement, where survival data reveal that restored endodontically treated teeth have comparable survival rates to restored implants, and where both modalities of treatment may be associated with varying degrees of periapical or peri-implant inflammation (Doyle et al. 2006, Ng, et al. 2010). The true local and systemic risks of apical periodontitis and peri-implantitis are yet to be fully elucidated. As populations age with increasingly complex co-morbidities and as the effectiveness of antibiotic treatment becomes less certain, the systemic risks of acute and chronic dental infections must not be dismissed. Case reports identify risks of sepsis associated with infected teeth in susceptible individuals (Newman 1996, Olsen & van Winkelhoff 2014, Oriol et al. 2015), and work continues on other systemic associations (Aminoshariae et al. 2017), including the elevation of inflammatory mediators associated with endodontic infections and endothelial dysfunction (Cotti et al. 2011). Potential patient vulnerability is illustrated by the greatly increasing numbers of patients undergoing joint replacement (National Joint registry 15th Annual Report 2018), organ transplantation (Organ Donation & transplantation activity report 2018/2019) and immunosuppressive treatments for neoplastic disease. Post-transplantation sepsis from a suspected dental source was, for example, acknowledged in 27% of surveys of dental care protocols among US organ transplant centres and 80% of the respondents routinely requested a pre-transplant dental evaluation (Guggenheimer et al. 2005). Patients and their medical teams may justifiably seek advice on asymptomatic periapical lesions, whether identified on periapical radiographs or 3D images (Rustemeyer & Bremerich 2007, Olsen & van Winkelhoff 2014). Further research is necessary to understand the nature of periapical radiolucencies identified on periapical radiographs and CBCT images and if there are any patient groups that are at special risk (Shendi et al. 2018). It may be that minimalist definitions of treatment success, such as tooth survival are insufficient to identify and manage risks. Equally, greater certainty is required on periapical radiolucencies that are resolving slowly, and whether they are likely to become a risk in susceptible patients. Other parameters related to quality of life and patient perceptions should also be considered in the decision-making process before planning further treatment (Liu et al. 2014, Leong & Yap 2019). The practice of judging endodontic outcomes on the basis of periapical radiographs may be insufficient to provide the answers that dentists and their patients need. Equally, if apical periodontitis is a significant disease in terms of local and systemic risks, then minimalist, patient-based outcomes such as tooth survival may be insufficient. The current reality is that endodontic procedures are considered safe, effective and evidence-based to prevent and heal apical periodontitis and manage its associated risks. Endodontics must not, however, be complacent and a renewed era of scientific rigour is necessary to understand more fully the nature of the periapical radiolucencies that are detected with CBCT imaging, their healing dynamics, histological nature and risks they may present to vulnerable patient groups.