In Kyoto, Japan, there is a famous rock garden where 15 rocks are laid out in a pattern, and all are not visible at once from any one position on the viewing deck. The lesson to be learned from this Buddhist-inspired garden is that no matter how hard we try, we cannot see the entire situation (all the rocks) at any one time, and if we finally can indeed see all the rocks at once, we are ready for enlightenment. (In endodontic terms, that might be considered mastery of the specialty and seeing everything that is actually present.) The analogy of this garden to the interpretation of endodontic radiographs is striking. Based solely on one radiograph, we very rarely can see the entire picture (all the rocks) that reflects the true and complete clinical situation.
Interpreting the quality of previous endodontic therapy from radiographs (especially a single image with the limitation of having been taken from only one angle) is at best a challenge and at worst misleading. Every endodontic radiograph shows some part of the whole picture, but at least 3 radiographs taken from 3 different angles (mesial, distal, and straight on) are required to show a complete image of the previous root canal that is as close to 3-dimensional as possible. A single radiograph may show an apparently well accomplished treatment, which when retaken from a second or third viewpoint may demonstrate an important discrepancy relative to the first view. Untreated canal space, hidden curvatures, and obscured separated files often can be seen in these multiple angles that were not visualized on the initial picture.
It is also imperative that interpretation of the clinical situation not be made solely by the radiographs. The patients objective and subjective clinical examination is the primary evidence upon which treatment decisions are based. Simply put, radiographs often lie, and single radiographs lie even more. The purpose of this paper is to discuss the key areas in which diagnostic endodontic radiographs must be evaluated and how to interpret findings in order to obtain the clearest true picture.
INTERPRETING RADIOGRAPHS
Several questions are invaluable when interpreting a previous root canal treatment (RCT) radiographically. These questions are the following:
(1) Radiographically, does the coronal seal appear to be compromised? Compromise of coronal seal can be indicated radiographically (amongst many possible such indications) by open crown margins, recurrent decay, voided spaces in the pulp chamber,* diffuse or indistinct buildup material that does not appear in intimate contact with the gutta-percha, voided spaces next to canal filling material of all types (silver cones, gutta-percha, paste, Thermafil, etc), and restorations that have been placed on crown margins to control recurrent decay. If coronal microleakage is present and clinical examination bears out such microleakage be aware that gutta-percha exposed to saliva even for short periods of time (weeks to months) is contaminated with bacteria and should be re-treated, even in the absence of symptoms.
 |
| Figure 1. Significant microleakage occurred under this crown. The existing void is highly suggestive of microleakage. |
*Note: Some bases, liners, and buildup materials are not visible radiographically, and confusion is possible in trying to judge if a void stems from an absence of buildup material under a crown or if the space is actually taken up by such a material. Caution is advised in these uncommon cases. In this authors experience, virtually every void radiographically visible under a crown that has been observed and subsequently re-treated has proven to be a true void that helped facilitate significant coronal microleakage and many failed RCTs (Figure 1).
(2) Has a post and buildup been used to restore the tooth? If so, one must be suspicious of perforations, root fractures, and fulcrum stresses that may lead to fractures, especially cervical fractures in the anterior region and vertical root fractures in the posterior region. In addition, the length of the post, the type of post (active, passive, and combination), the method of bonding, the posts circumference, etc, all have a direct impact on the potential for an iatrogenic event. For example, a screw post that is actively engaged in dentin but only to the level of the gingival crest may cause a cervical fracture. Oversized, actively threaded posts (excessively long and/or wide posts) also have significant iatrogenic potential. If a post is off center of the true canal, one must be strongly suspicious of a perforation until proven otherwise.
 |
| Figure 2. This defect is probed to the apex and, given the teardrop radiolucency and drainage from the sulcus, a diagnosis of a vertical root fracture was made. |
In addition, the canal that was chosen for post placement as well as the number of posts per tooth are important considerations. All things being equal, posts placed into the mesial roots of lower molars risk perforation far more than those placed into the distal roots. Placement of more than one post per tooth is also a strong visual clue that little remaining coronal tooth structure was present and that the tooth is susceptible to crown and/or root fracture (Figure 2).
 |
 |
| Figures 3a and 3b. The gutta-percha is not centered in the distal root, highly suggestive of a second untreated canal (3a). The completed re-treatment is pictured (3b). |
(3) How many canals have been obturated? Canal system anatomy is anything but predictable and the unexpected is the norm. That said, methodical skepticism is advised when evaluating previous root canal treatment when less than the number of potential canals are found. For example, the vast majority of upper first molars have 4 canals; only seeing 3 filled canals is an obvious clue that an untreated fourth canal (usually an MB2 canal) may be present. In a similar fashion, seeing only one treated canal in a lower anterior tooth should alert the observer to the presence of an untreated second lingual canal. Every tooth in the mouth (except perhaps the upper incisors) should be evaluated with the recognition that there exists the potential for additional canals. Also, if the root canal filling is not centered in the canal, it is a virtual certainty that a second canal exists within the root that is untreated (Figures 3a and 3b).
 |
 |
| Figures 4a and 4b. Significant untreated canal space is present in both root apices (4a). The completed retreatment is pictured (4b). |
(4) Where do the root canal fillings terminate in relation to the radiographic apex? Controversy exists regarding the correct location of the most desirable termination point for root canal fillings. Some believe that the best location is an arbitrary point at some distance back from the radiographic apex, based on anatomical averages. Others believe it is the radiographic terminus (RT). Anatomic studies have shown that the apical foramen can be as far as 3 to 4 mm away from the RT. An important aside must be made at this point: filling to the RT will by definition often cause sealer and gutta- percha to be extruded beyond the apical foramen of the tooth. The danger of filling teeth arbitrarily short of the RT, however, is that many millimeters of canal space may be left untreated. For each linear millimeter of distance back from the RT, quite a bit of untreated space in the apical delta canals may remain if the entire canal space is not obturated to a reproducible reference point. The farther from the RT that the previous root canal filling terminates, and the less ideal its shape, taper, and density, the greater the chance that the tooth contains significant uncleaned and unfilled space that may require re-treatment (Figures 4a and 4b).
 |
 |
| Figures 5a and 5b. This canal shape fulfills all the requirements of root canal preparation and facilitates obturation. |
(5) What is the shape of canal preparation? Excellent canal preparation should (a) resemble a tapered funnel from the orifice to the radiographic apex, (b) leave the canal in its initial position, (c) leave the foramen in its original position and size, (d) facilitate obturation, and (e) be as small as is practical. Any other shape given to canal preparations is a deviation from ideal and represents potential for debris to have been left within the root canal system, thus risking subsequent failure (Figures 5a and 5b).
 |
| Figure 6. The difference in the 3-dimensional obturation obtained in the central incisor versus the cold lateral condensation done in the lateral incisor is striking. |
(6) Has the whole canal system been treated? Are there sealer puffs present that indicate treatment of all lateral and delta canal systems within the tooth? Not every well-obturated tooth will exhibit sealer puffs, but sealer puffs most often demonstrate that all portals of exit (POEs) of the canal system have been treated and all pulp tissue in these POEs (vital and nonvital) was extruded during the compaction of gutta-percha. Necrotic pulp tissue is a potent toxin, which if left in the tooth greatly increases the risk of root canal failure. Once pushed into the periodontal ligament by the compaction of warm gutta-percha, such toxins can be addressed by the immune system. A lack of sealer puffs, while not diagnostic, does imply the possibility of untreated canal space (Figure 6).
 |
| Figure 7. The faint and nondistinct appearance of the root canal treatment in this upper second molar is highly suggestive of a paste root canal fill that was clinically verified. |
(7) What is the density of the root canal filling? Clear voids, faint or diffuse areas of filling, spirals of gutta-percha with empty space next to the master cone, etc, all are strong indicators that uncleaned and unfilled space exists within the canal system. In addition, filling material that appears lighter in shade or appears “grainy” is a strong visual clue of a paste root canal filling. Considerations for re-treatment of paste root canals are beyond the scope of this paper, but suffice it to say, if the operator is reasonably certain a paste root canal has been done, and especially if a new coronal restoration is needed, the tooth should be re-treated ASAP (Figure 7).
 |
| Figure 8. Lack of a rubber dam and lack of recognition that the roots and crown were not aligned led to a cervical perforation at the distal of this upper bicuspid. |
(8) Is there evidence of iatrogenic misadventure; ie, perforation, fractured instruments, ledges, torn apical foramen, etc? The presence of iatrogenic misadventure often requires referral to determine whether such issues need further observation, re-treatment, extraction, or possible surgery. Such iatrogenic events could be manifest by (amongst many possible radiographic appearances) the presence of little if any furcal dentin (near perforation), furcal bone loss (either root fracture, furcal floor fracture, or true perforation), distinct white lines in the canal more opaque than gutta-percha (diagnostic for separated instruments), gutta-percha filling that doesn’t follow root curvature (severe ledging), etc (Figure 8).
(9) Is there a surplus of GP beyond the radiographic apex? Small amounts of gutta-percha and sealer beyond the RT will cause no harm and are well tolerated by the body if the entire canal system has been obturated to the minor constriction of the apical foramen. There is a clear distinction, however, between an “overfill” and a “surplus” of gutta- percha and sealer. A surplus implies that the canal is indeed obturated ideally and that there is a small amount of excess filling material. The extrusion of a small amount of sealer usually causes no difficulties other than possibly some mild and self-limiting inflammation.
 |
 |
| Figures 9a and 9b. A lack of an apical capture zone for gutta-percha led to this overextension (9a). The completed re-treatment is pictured (9b). |
Overfilling implies that there is untreated canal space within the tooth while there is an overextension of gutta-percha and sealer. The difference is not academic. Bacteria, untreated canal space, and apical dead pulp are a source of possible future infection. While the “overfill” is often blamed as the source of failure, the true problem is the lack of ideal cleansing, shaping, and filling that has left significant canal debris in the apical third, which the body usually cannot tolerate, leading to later infection (Figures 9a and 9b).
 |
| Figure 10. The teardrop radiolucency present at the apex of this lower bicuspid is highly suggestive of a vertically fractured root that was clinically confirmed. |
(10) What radiolucencies are present? Obviously, radiolucencies are anticipated at root apices, but it bears mentioning that lesions of endodontic origin (LEOs) can arise anywhere along the periodontal ligament, and LEOs are present laterally as often as they are present apically. Lesions form opposite POEs, and after obturation sealer puffs are frequently found leading into LEOs. In addition, it is very important to distinguish between LEOs and normal anatomic spaces. The mental nerve bundle as well as the incisive canal (in addition to other normal anatomic structures) can be misinterpreted as pathology. A lesion in the presence of apparently well-accomplished root canal therapy should cause the operator to take a second look for a normal anatomic structure or possible nonodontogenic pathology mimicking a LEO. In addition, it is noteworthy that a pattern of bone loss in the furca, where the entire furcal bone complex is missing, is a strong indication that either one of the roots is fractured or that there is a mesial-to-distal fracture running through the furcal floor, especially in an upper molar. The prognosis for a tooth with this condition (even in the absence of probing pre-operatively) is very poor, all things being equal. Similarly, a teardrop radiolucency that extends up a root, especially in the presence of a post and buildup, is most often associated with a vertical root fracture, and a probing to the apex of the affected root is virtually diagnostic (Figure 10).
ONE FINAL QUESTION
While not directly germane to the topic of this paper, a final question that should be asked is, what should be done if a given root canal can obviously be improved based on the radiograph (obviously missed canals and/or other significant defects), and yet there is no apical pathology and the patient is asymptomatic? While there is not a completely correct answer to the question, the following may illuminate this often-grey area. If the tooth is to be restored, especially if it is a strategic tooth (lone distal abutment, etc) and the root canal treatment is less than ideal, all things being equal, re-treatment is preferable to later having to make access through a new bridge and risk possible damage. If there is no immediate restorative treatment planned, the patient should be informed of the clinical findings, told of all risks and benefits, and be allowed to make his or her own choice.
CONCLUSION
Multiple angles (mesial, distal, and straight on) of radiographs are essential for adequate evaluation of previous root canal therapy. In combination with a comprehensive objective and subjective clinical examination, such radiographic interpretation from multiple angles is made much more decisively when consideration is given to the questions presented above. This approach will go a long way toward allowing the dentist to see the often-elusive complete clinical picture …and as at the garden in Kyoto, see all the rocks at the same time.
Dr. Mounce is in private endodontic practice in Portland, Ore. He lectures worldwide and has written numerous articles for journals, including Dentistry Today and the Journal of Endodontics. Please send e-mail if you are interested in attending one of Dr. Mounce’s lectures worldwide. He can be reached at lineker@aol.com.
