Nonsurgical Versus Surgical Endodontic Retreatment: “How Do I Choose?”

Dentistry Today


The purpose of endodontics is to prevent or cure lesions of endodontic origin. Endodontic retreatment is focused on curing lesions of endodontic origin (LEOs) where incomplete endodontic healing exists.
Retreatment endodontics is a discipline within a discipline. Success requires a different set of skills, leading-edge technologies, and interdisciplinary thinking. The intention of this article is to examine the treatment planning considerations for nonsurgical versus surgical endodontic retreatment, and offer the clinician decision-making guidelines for optimal treatment outcomes.
Five critical considerations influence endodontic retreatment success. Each of the five retreatment considerations will be discussed and then summarized in an easy-to-understand flowchart, followed by my personal decision-making “gray area,” and finally my corresponding guideline for each consideration.


Figure 1. Quality of Obturation. (a) This maxillary left central incisor not only has a poor obturation, it essentially has no obturation at all. The retrorestoration also apperars to be “floating away.” (b) Clinical photograph of gutta-percha cone tracing sinus tract. (c) Radiograph showing gutta-percha cone tracing to underfilled apical POE. (d) Nonsurgical retreatment ensures that the body of the root canal system is sealed and does not risk further scarring in the aesthetic zone of this patient with a high smile line. In 3 months the lamina dura appears to be returning. (e) Three month post-treatment clinical of healed sinus tract.

Healing of LEOs after nonsurgical retreatment plateaus within the first year.1 By one year, approximately 90% of the teeth that heal eventually demonstrate signs of healing, and almost 50% are already completely healed. At 2 years, the majority of LEOs are healed, while others demonstrate further reduction or elimination of the LEO. This healing process may continue for several years or longer. It is important, therefore, to diagnose whether a radiographic LEO is actually healing or not healing. In order to measure increasing, decreasing, or static radiographic LEO size, it is valuable to compare current and previous radiographs. Signs and/or symptoms, however, clearly identify failure to heal, such as a sinus tract, palpation, percussion, or cellulitis.

Quality of Obturation

Figure 2. LEO in apical third. (a) The obturation is solid radiographically, but the POEs are clearly transported internally. Good candidate for surgical retreatment since there are no independent lateral LEOs. (b) Apical portion of tooth surgically removed before retrorestoration, revealing multiple patent POEs. (c) 10-year recall demonstrates radiographic healing of the LEO and healthy lamina dura.

Is the existing endodontic obturation fairly solid or inadequate? What role does obturation quality play in deciding nonsurgical (NSRCT) versus surgical (SRCT) endodontic retreatment? The quality of endodontic obturation is a 2-dimensional measurement of a 3-dimensional structure, which means quality of obturation is, at best, a judgment call. Nevertheless, most dentists can get an instant and empirical sense that the obturation is good or poor. Obturation quality is, of course, an exact expression of the quality of the cleaning and shaping. Have the five mechanical objectives of classic quality cleaning and shaping been achieved?2,3 Usually, if the dentist has failed in one objective, he or she has failed in other mechanical objectives as well.


Figure 3. LEO is lateral, and coronal microleakage is likely. (a) Nonsurgical retreatment gives maximum chance for healing since the surgical discovery of a lateral POE would be challenging, access could be invasive to the attachment apparatus, and microleakage, if present, would continue. (b) Bridge removal with CORONAflex (KaVo). (c) Eleven-year recall radiograph revealing excellent lateral LEO healing with new, individually cemented full crown.

• Continuously tapering cone.
• Each cross-sectional diameter of the preparation becomes narrower apically.
• The root canal preparation “flows” in the same 3-dimensional planes as the original root canal system.
• The apical portal of exit (POE) is neither transported internally (blocked or shelved) or externally (perforated or torn).
• The apical minimal diameter is preserved to be as small as is practical.
With the exception of external transportation, endodontic retreatment due to poor obturation is usually easiest and most predictable through nonsurgical retreatment. Often, the ferrule width and height have not been compromised due to inadequate access and shaping. Often, the initial “glide path” was not developed due to a block or ledge.4 When a poor obturation is present, the dentist often simply gave up any “deblocking” or “deledging” attempts early in the endodontic treatment and obturated whatever minimal space had been created. Since larger areas of vertical and therefore emanating lateral POEs are undiscovered and not filled in these cases, nonsurgical retreatment would be the treatment of choice for maximizing the obturation of potentially significant POEs along the root surface. Depending on the size of the original canal, usually the better the obturation the better the cleaning of viable and nonviable irritants and bacteria, and the better the obturation due to larger obturation hydraulic spaces. In addition, when the main body of the root canal system has been treated, any subsequent failure to heal the LEO’s underfilled POE would probably occur apically and therefore could be easily corrected with surgical endodontic retreatment. In addition, when an obvious missed canal (for example the fourth canal in maxillary and mandibular molars) is the cause for lack of healing, nonsurgical retreatment is advised because surgical apical correction leaves the main body of the canal full of irritants capable of spilling out of lateral POEs or from a poor apical seal, thus potentiating a LEO.

Figure 4. Structural integrity determinants. (a) Radiograph of maxillary left canine showing gutta-percha cone tracing to lateral LEO. Disassembly and nonsurgical retreatment would be a significant structural risk. The ferrule is already minimal because of the large post. The POE source responsible for the lateral LEO appears to be accessible surgically. (b) Sinus tract facial to maxillary left canine. (c) Ultrasonic preparation. (d) MTA retrorestoration. (e) Five-year recall with lateral LEO healing. (f) Sinus tract closed immediately after microsurgery.

Decision-Making Gray Area: Though the apparent quality of the obturation does reflect the cleaning and shaping, it is possible that the root canal network was disinfected and cleaned beyond the borders of the obturation material. The radiograph, therefore, can be deceptive and partly misleading, or not measure what has been removed and what has been left behind in the root canal system.
Guideline 1: When the initial endodontic obturation is inadequate, when coronal microleakage is suspected, and when the tooth structure can sustain disassembly, then nonsurgical retreatment over surgical treatment is the optimal choice.


Is the LEO at the apex, is it located laterally along the root, or is it both? What important influence does this play in advising the patient that the best option to save the tooth is nonsurgical or surgical?
The LEO location consideration is one of the key critical distinctions in deciding a nonsurgical versus surgical endodontic correction.5 If the LEO is coronal to the apical third of the apex, or if the LEO extends coronally past the apical third, the likelihood of an exclusive surgical endodontic seal lessens. Even though ideally 3 mm of root and therefore root canal system is removed during the surgical retropreparation, and even though an additional 3-mm apical preparation and restoration are placed, much root surface remains that may contain significant POEs. The surgical seal alone will not capture the POEs that are coronal to the retro-obturation. It should be noted that when a radiographic LEO is present anywhere along the root surface, an offending POE is often lurking within the LEO’s radiographic architecture. Lateral LEOs are quite distinctive in both initial and retreatment endodontics. In addition, even if a 3-mm retrorestoration is placed, and even though the body of the root canal system is incarcerated, the lateral branches still exist and are capable of being a LEO source of infection and inflammation.

Figure 5. Maxillary left central incisor has mid root palpation soreness and is percussion sensitive. (a) LEO incisal to lateral surgical amalgam has been increasing in size during the last 3 months. Sufficient tooth structure warrants nonsurgical retreatment. (b) Sliding small manual endodontic file into lateral POE in order to clean and confirm patency for obturation. (c) Twelve-year recall; tooth is asymptomatic, and lateral LEO has healed.

Generally speaking, if the LEO extends coronally a third of the root or more, or if a separate lateral LEO is present, nonsurgical retreatment should be chosen over surgical retreatment. A useful way to discover the location of the offending POE is to trace the sinus tract (when present) with a gutta-percha cone. Magnification and illumination (operating microscope) are helpful in discovering the sinus tract lumen, which is not always obvious and will sometimes need to be gently milked with a Q-tip to expose its opening. Cut the gutta-percha cone in half, and using the apical, curved portion, follow the sinus tract (similar to a small file following a canal) to the sinus tract foraminal source.
Decision-Making Gray Area: One or more POEs may be the problem today, but another may be a LEO source in the months and years to come. SRCT is more corrective than preventive, and NSRCT is as much preventive as corrective. Also, when considering NSRCT versus SRCT, the quality of the coronal seal and the structural integrity of the ferrule are not known. As discussed later, these 2 considerations are key in choosing NSRCT versus SRCT retreatment.
Guideline 2: If the LEO is confined apically, and the previous endodontic treatment is reasonably sufficient, and if nonsurgical retreatment were to compromise the structural integrity of the tooth, then surgical correction is the treatment of choice.


Structural considerations include the height, width, and location of the ferrule, and whether any of these essential structures would be compromised during disassembly (ie, post or foundation removal) or new access preparation. Remember, the ferrule must be measured after the access preparation. Before the access, dimensions do not count. In summary, structural or restorability considerations are as follows:6
• 1.5 to 2.0 mm facial and lingual ferrule is minimum. The lingual ferrule is more critical in maxillary anterior teeth, and the labial ferrule is more critical in mandibular anterior teeth. In posterior teeth, the ferrule is essential on both the buccal and lingual due to the magnitude and multiple directions of force.
• Minimum 1.0-mm ferrule wall thickness.
• Crown-to-root ratio is at least 1:1.
• Third, third, third rule: The canal or canal preparation should not be wider than one third of the mesio-distal root diameter after endodontic access.
If there is not enough ferrule present in an endodontically involved tooth, then either forced orthodontic eruption or osseous recontouring, or both, are required to avoid biologic width invasion. Four millimeters (4 mm) must be present from the height of the bone to the height of the ferrule in order to be restorable. Two-and-a-half millimeters are required for the periodontal ligament and the epithelial attachment, ie, the biologic width, and a minimum 1.5 mm is needed for the ferrule height. Hence, 2.5 mm + 1.5 mm = 4.0 mm.
The following are considerations for forced orthodontic eruption:7

• root length
• root form
• level of fracture
• aesthetics
• importance of tooth
• endoperio prognosis.

If short root length or conical root form make forced orthodontic eruption undesirable, and if an inadequate ferrule exists, then the post needed to retain the foundation must be designed to be load bearing. As soon as the post becomes loaded, the risk of root fracture increases.
Decision-Making Gray Area: Essentially none. Whether a tooth is structurally restorable or not restorable has specific rules. These rules apply whether a tooth needs no endodontic treatment, needs endodontic treatment, or needs endodontic retreatment. Ferrule is the main restorability determinant. However, being able to judge the amount of ferrule under a crown is impossible from a radiograph.
Guideline 3: An endodontically diseased tooth must have sufficient structure and ferrule integrity in order to predictably save it and to predictably restore it.
Approximately two thirds of full-coverage endodontically treated teeth that become symptomatic or that develop a LEO(s) demonstrate coronal bacterial leakage or frank caries.3 Therefore, coronal leakage plays an essential role in a NSRCT versus SRCT endodontic retreatment.8,9 When a patient presents with a crown and a LEO, careful evaluation for possible marginal leakage or coronal micromovement (best observed under the operating microscope) must be made. In addition, if considerable disassembly is required (post and/or foundation removal), then it is possible to actually create microleakage and potentiate a future endodontic failure. Of course, if coronal leakage is suspected, crown removal and NSRCT would be the treatment of choice. In addition, if NSRCT were performed, as with initial nonsurgical treatment, it is imperative to protect the gutta-percha from possible future coronal leakage by discontinuing the gutta-percha back pack 1 mm short of the access chamber and placing a bonded barrier over the gutta-percha and chamber floor at the end of the obturation visit.
Analysis of extracted teeth due to endodontic failure reveals that 85% have no full coverage.10 The number of extracted teeth with no crown was 4.8-fold higher in anterior teeth, 5.8-fold higher in premolars, and 6.2-fold higher in molars as compared to teeth with a crown. When restorations and post and core foundations experience micromovement during masticatory cyclic loading, and where inadequate endodontic coronal seals exist, teeth may have a much greater potential for failure because of bacterial leakage down the obturated canal, which can occur almost immediately after restoration.
Finally, if a crown or bridge abutment is cemented with zinc phosphate cement, removal is often successful. Stronger bonded cements may make removal impossible. Of course, better long-term marginal integrity should significantly reduce coronal leakage and reduce endodontic LEOs in the future. The key is for the  clinician to determine, if possible from dental records, which cement was used in cementing the crown.
Decision-Making Gray Area: Unfortunately, it is impossible to know for certain the degree of coronal microleakage, if any, when evaluating NSRCT versus SRCT. Sometimes identification of leakage is only possible after crown removal or access cavity.
Guideline 4: If coronal leakage is suspected, then NSRCT and a new crown should be chosen over SRCT retreatment.
Full mucoperiosteal surgical endodontic retreatment in the aesthetic zone risks the loss of the proximal papilla, or creates what is known as a gingival “black triangle,” when the distance from the base of the interproximal contact point to the crest of bone is greater than 5 mm.11
SRCT incisions in the attached gingiva can result in an unaesthetic scar, especially in patients with a high smile line. This risk can be minimized using microsurgical techniques and sutures. However, these two potentially devastating aesthetic outcomes must be considered when determining whether an NSRCT versus SRCT endodontic retreatment approach offers the best treatment options for the patient.
Decision-Making Gray Area: Preventing gingival “black triangles” and scarless incisions depends on proper surgical design and operator skill, care, and judgment.
Guideline 5: If a gingival “black triangle” or unaesthetic incision is a high risk, choose NSRCT retreatment.
(1) If NSRCT is chosen and an existing crown is present, or if the tooth is an abutment to a bridge, first remove the restoration, if possible. If not, perform the endodontic retreatment through the existing crown or abutment and advise the patient that the crown or abutment could become loose and may require replacement. This is especially true if insufficient tooth structure exists under the crown, which, at the beginning of treatment, is often unknown. Also, advise the patient that caries observation using caries indicator paste after the access may reveal caries or leakage under the crown, requiring a new restoration or a new crown after the endodontic retreatment. Pave your way with words!
(2) If the tooth is not healing endodontically, and if it does not have full coverage, then retreat the endodontics nonsurgically and restore. If full coverage already exists, or if the tooth is a bridge abutment, first attempt to remove the crown or bridge. If removal cannot be accomplished, then retreat nonsurgically through the crown or consider a surgical approach depending on the LEO location, structure, and aesthetic considerations.
(3) Identify LEO location, whether it is buccal, lingual, lateral, or in the furca. If the sinus tract traces to a position on the root somewhere other than the apical third, consider nonsurgical correction. Choose a nonsurgical retreatment approach when lateral portals of exit are suspected of preventing healing.
(4) Choose a surgical retreatment plan if the LEO is apical or at least surgically accessible, and if the nonsurgical retreatment would be structurally invasive, leaving the tooth unrestorable or structurally compromised.
(5) In general, evaluate the existing endodontic quality. If poor cleaning, shaping, and obturation have been previously performed, or if there is a likelihood of a missed canal or other significant anatomy, then the nonsurgical retreatment approach would be the first choice. Where surgical guidelines have been met, however, the surgical retreatment is ultimately more conservative. Literature and clinical experience suggest that failure to seal in endodontics does not mean failure to heal, but failure to heal means failure to seal.
This article has discussed five guidelines for managing endodontic patients who have previously endodontically treated teeth with persistent or developing LEOs. The role of endodontic obturation, LEO location, structure, coronal seal, and aesthetics have been reviewed in detail. The intent of this article has been to emphasize the importance of interdisciplinary thinking, decision making, and sequencing when planning endodontic retreatment for patients who have recurring endodontic disease.


1. Friedman S. Expected outcomes in the prevention and treatment of apical periodontitis. In: Orstavik D, Pitt Ford TR, eds. Essential Endodontology: Prevention and Treatment of Apical Periodontitis. 2nd ed. Oxford: Blackwell Science (in press).
2. Schilder H. Cleaning and shaping the root canal. Dent Clin North Am. 1974;18:269-296.
3. West J. Endodontic update 2006. J Esthet Restor Dent. 2006;18:280-300.
4. West JD. Perforations, blocks, ledges, and transportations: overcoming barriers to endodontic finishing. Dent Today. Jan 2005;24:68-73.
5. West JD. Restore or remove: the endodontic perspective. Seattle Study Club Journal. 1998;2:29-37.
6. Kinzer G. Interdisciplinary management of traumatic fracture. Advanced Esthetics and Interdisciplinary Dentistry. April 2005;1(1):12-18.
7. Kokich VG. Adjunctive role of orthodontic therapy. In: Newman MG, Takei HH, Klokkevold PR, et al, eds. Clinical Periodontology. 10th ed. St Louis, Mo: Saunders-Elsevier; 2006:856-870.
8. Torabinejad M, Ung B, Kettering JD. In vitro bacterial penetration of coronally unsealed endodontically treated teeth. J Endod. 1990;16:566-569.
9. Ray H, Trope M. Periapical status of endodontically treated teeth in relation to the technical quality of the root filling and the coronal restoration. Int Endod J. 1995;28:12-18.
10. Farzaneh M, Abitbol S, Friedman S. Treatment outcome in endodontics: the Toronto study. Phases I and II: orthograde retreatment. J Endod. 2004;30:627-633.
11. Tarnow DP, Magner AW, Fletcher P. The effect of the distance from the contact point to the crest of bone on the presence or absence of the interproximal dental papilla. J Periodontol. 1992;63:995-996.

As the founder and director of the Center for Endodontics, Dr. West continues to be recognized as one of the premier educators in clinical and interdisciplinary endodontics. He received  his DDS in 1971 from the University of Washington, where he is affiliate associate professor. He then received his MSD in endodontics in 1975 at Boston University Henry M. Goldman School of Dental Medicine, where he is a clinical instructor and was awarded the 1995 Distinguished Alumni Award. Dr. West has presented more than 400 days of continuing education in North America, South America, and Europe while maintaining a private practice in Tacoma, Wash. He co-authored “Obturation of the Radicular Space” with Dr. John Ingle in Ingle’s 1994 and 2002 editions of Endodontics and was senior author of “Cleaning and Shaping the Root Canal System” in Cohen and Burns’ 1994 and 1998 Pathways of the Pulp. Dr. West’s memberships include The American Academy of Esthetic Dentistry, The Northwest Network for Dental Excellence, and The International College of Dentists. He is the scientific endodontic editor for Boston University’s Communiqué, where he is also a member of the school’s Board of Visitors. Dr. West is a Thought Leader for Kodak Digital Dental Systems and serves on the editorial advisory boards of The Journal of Advanced Esthetics and Interdisciplinary Dentistry, The Journal of Esthetic and Restorative Dentistry, Practical Procedures and Aesthetic Dentistry, and The Journal of Microscope Enhanced Dentistry. He can be reached at (800) 900-7668 or