A perforation is an unintended opening between the canal space and the periodontal tissues, cut by a bur, a file or a post drill, or produced by caries or resorption. It is one of the few endodontic complications that can cost the tooth on its own, and also one of the most repairable when it is recognised at once and sealed well. This guide covers where perforations happen, how to recognise one, what to do in the first minutes, how to judge the defect, when to repair, refer or extract, the principles of repair, and what the outcome data actually show.
Last updated: September 2026
Iatrogenic perforations arise at every stage of treatment: a bur misdirected during access, a file straightening a curve until it exits the root, a rotary thinning the furcal wall of a mesial root, a post drill leaving the canal. Pathological perforations arise from caries and from internal or external resorption. Reported perforation rates run from 0.6% to 17.6% depending on the series, and the perforation-repair meta-analysis notes that iatrogenic perforations account for a measurable share of the endodontically treated teeth that end up extracted. The defect itself is a portal for bacteria into the periodontium; what decides the tooth's fate is how quickly it is sealed, whether the site is already infected, and where it sits relative to the crestal bone.
Recognise, isolate, seal. A fresh, small perforation sealed promptly with a calcium-silicate cement under magnification has a reasonable outlook: the pooled nonsurgical repair success in the 2015 meta-analysis was about 73% across all materials and about 81% with MTA. The defects that fare worse are the ones left open, the ones already surrounded by a radiolucency, and the ones at or near the crestal bone where the gingival epithelium can reach them. If the site cannot be seen, dried and sealed with confidence, that is the moment to refer rather than to pack material blind.
Naming the site tells you both how it happened and how it is likely to behave. The rows below follow the procedure in which each perforation typically arises.
How it typically happens
Bur driven through the floor while searching for canals in a calcified or receded chamber, or the roof mistaken for the floor in a tooth with a shallow chamber.
Why it matters
Sits at the furcation near the crestal bone; contamination from the sulcus and rapid attachment loss are the concern. Large defects need a matrix to keep the repair material out of the furcation.
How it typically happens
Misaligned bur during access in a tilted or rotated tooth, orifice enlargement that strays laterally, or a post drill that leaves the canal.
Why it matters
Often at or above the crestal bone. Supracrestal defects can be restored like a cavity; those close to the attachment risk epithelial downgrowth and a pocket.
How it typically happens
Over-preparation of the furcation-facing wall of a curved root, classically the distal wall of a mandibular mesial root, by large tapers, Gates-Glidden drills or ultrasonic troughing in a calcified canal.
Why it matters
The wall is thinnest here. The defect is long rather than round, hard to see and hard to seal; the larger the defect area, the harder a tight seal becomes.
How it typically happens
Canal transportation in a curve: a stiff file straightens the canal, ledges, then exits the root; or the apical constriction is driven through during shaping.
Why it matters
Usually below the attachment and less exposed to the mouth, so it behaves more like an over-prepared apex. A well-fitted cold obturation, or a calcium-silicate plug, often manages it.
How it typically happens
Post drill following the root's external contour incorrectly, or a fibre-post removal done without magnification.
Why it matters
Frequently discovered late, after cementation; the site may already be infected. Retreatment planning starts with the perforation, not the filling.
How it typically happens
Caries extending to the root or furcation; internal or external inflammatory resorption breaching the canal or the root surface.
Why it matters
Restorability is the first question; the resorption guide covers the diagnostic side. Extensive pathological defects often make the tooth unrestorable.
Over-instrumentation through the natural foramen is over-preparation, not a perforation. It is still a procedural error, but the outcome cohort discussed below found it did not by itself worsen healing, unlike a canal that was deviated and left uninstrumented.
Most perforations announce themselves at the moment they happen. The signs are easy to dismiss in a busy canal, so the habit is to stop and look whenever one appears.
Bleeding also comes from vital pulp remnants and from a furcation already inflamed by disease. Bright, persistent bleeding that does not settle with the usual irrigation and a paper point that keeps staining at a fixed level point toward a perforation; a film with a marker settles it.
Time is a prognostic factor in every source. What happens in the first minutes decides whether the defect is sealed clean or seeded.
Withdraw the instrument, dry the chamber and look. Place a paper point or a small file in the suspected site and take a film. Do not keep instrumenting through a defect: every pass enlarges it and carries debris into the periodontium.
A well-sealed rubber dam is the precondition for any repair. Gentle pressure with a paper point, irrigation, and a short application of a haemostatic agent or calcium hydroxide are the usual first measures; sodium hypochlorite through an open defect is avoided.
Explain plainly what happened, that it is a recognised complication, what repair involves and what the outlook is. Document the site, size, the bleeding, the radiograph and the conversation.
A fresh, accessible, dry defect is generally sealed at the same visit. If bleeding cannot be controlled, if the site cannot be seen, or if the chairside time is not there, calcium hydroxide and a well-sealed temporary hold the situation until the next visit or the referral appointment.
Once the defect is sealed or temporised, disinfect the canals as usual and keep the perforation in mind at obturation: warm or high-pressure techniques can push material through a weak or repaired wall.
Six variables shape the plan. Most can be judged from the film, the microscope view and the history.
A fresh defect made in a clean field is a wound; an old one discovered at retreatment is an infected communication. Every source treats prompt sealing as favourable, even though the pooled data could not quantify the effect.
Small round defects seal predictably. A large or strip-shaped defect is harder to fill without extruding material, and defect area tracks the risk of an inadequate seal. In the 2015 pooled data the size bands did not reach significance, but the numbers were small.
Defects at or near the crestal bone are exposed to the sulcus and to epithelial downgrowth; supracrestal ones can be restored from outside; apical ones are protected by the attachment. The pooled analysis found no significant difference by level, while clinical reviews still treat the crestal zone as the one that decides the prognosis.
A radiolucency adjacent to the perforation roughly halved the odds of success in the meta-analysis (absence of radiolucency OR 2.57). Apical periodontitis at the apex, by contrast, did not change the repair's outcome.
Remaining tooth structure, periodontal status and strategic value decide whether repair is worth attempting. Maxillary teeth healed more often than mandibular ones in the pooled data (OR about 2.2), which the authors relate partly to blood supply and partly to radiographic overlap.
A defect that can be visualised, dried and reached with a carrier under the microscope is a chairside repair. One that cannot is a referral or a surgical question, not a blind attempt.
The two overall options are repair or extraction, and repair is nonsurgical whenever the defect can be reached from inside the tooth. The table matches the common situations to the usual first approach.
Usual first approach
Immediate orthograde repair with a calcium-silicate cement under magnification, then complete the root canal treatment.
Notes
The best-case scenario in every source. Keep the material inside the root; a small defect needs no matrix.
Usual first approach
Haemostasis, calcium hydroxide, sealed temporary; repair at the next visit or refer.
Notes
A short delay in a clean field is generally acceptable; leaving the site open to saliva is not.
Usual first approach
Internal matrix (a resorbable barrier placed through the defect) then calcium-silicate cement; consider referral.
Notes
The matrix keeps the cement out of the furcation. Strip defects are the hardest to seal and the ones most often referred.
Usual first approach
Repair the perforation first, then remove the root filling and continue.
Notes
Otherwise irrigant and filling debris travel through the open defect into the periodontium.
Usual first approach
External restoration of a supracrestal defect; orthodontic extrusion may bring a crestal defect into reach.
Notes
An alternative to surgery for anterior and premolar teeth.
Usual first approach
Surgical repair with a calcium-silicate cement, intentional replantation, or extraction.
Notes
Bone loss, defect age, periodontal status and the surgeon's experience decide between them. Extensive pathological defects usually end in extraction.
Whatever the branch, the goal is a sealed defect with a healthy periodontium next to it. If a pocket has already formed at the site, the aim shifts to re-establishing attachment and the prognosis is more guarded.
The principles below recur across the sources; they are what a competent repair looks like, not a substitute for training under magnification.
Material pushed through the defect is a foreign body in the periodontium, not a repair. When the defect is large, when it cannot be dried, or when its edges cannot be seen, a matrix or a referral is the safer path than more pressure.
The 2015 systematic review pooled 188 repaired perforations from twelve cohort studies: overall nonsurgical success about 73% (CI 62–81%), about 81% with MTA (CI 67–90%). Dedicated repair studies performed under the microscope by trained operators reported around 90%; retreatment cohorts in which perforation was one problem among several reported 50–60%.
The best management of a perforation is not to make one. The habits below are grouped by treatment stage.
How urgent is repair?
Short answer
Every source treats prompt sealing as favourable: a fresh defect in a clean field is a wound, an old one is an infected communication.
What is the success rate of perforation repair?
Short answer
The 2015 meta-analysis pooled 188 repaired perforations: about 73% success overall and about 81% with MTA.
Which material should be used?
Short answer
Calcium-silicate cements — MTA and its successors, tricalcium-silicate cements and premixed bioceramic putties — are the materials of choice in every source read for this guide.
Does the location matter?
Short answer
Clinically, yes: defects at or near the crestal bone are exposed to the sulcus and to epithelial downgrowth, apical ones are protected by the attachment, and strip defects are the hardest to seal.
Is a CBCT needed?
Short answer
Not routinely.
Should the perforation or the old root filling be dealt with first at retreatment?
Short answer
The perforation.
Is over-instrumentation through the apex a perforation?
Short answer
No.
When is extraction the right answer?
Short answer
When the tooth is not restorable, when the defect is extensive or pathological, when reaching it would destroy too much bone or root, or when a repair has failed with persistent symptoms and bone loss.
Chairside references that pair with this guide — the complications hub, the separated-instrument decision guide, the retreatment and restoration guides, and imaging for the CBCT call.
This guide is an educational reference for dental professionals. It is not a substitute for clinical judgment, individual patient assessment, or referral to an endodontic colleague where indicated. Repair techniques are described at the level of principles, not as a step-by-step tutorial.
Reviewed by
Uzm. Dt. Levent Yüksel
DDS · Endodontist
Endodontist — DDS, Hacettepe University; endodontics specialty training, Ankara University. 10+ years of clinical experience. Sole author and reviewer of EndoGuide content.