Retained Root Tips After Extraction. When to Retrieve, When to Stop, and When Coronectomy Is Safer
Few moments in oral surgery create more anxiety than watching a root fracture during an extraction. The instinct is immediate. Get it out. A complete extraction feels definitive, while leaving anything behind can feel like failure. Yet modern oral surgery increasingly shows that this instinct can become dangerous when pursuit of a small fragment threatens a much more important structure.
The better question is not whether a good dentist leaves root tips. It is whether removing this particular fragment creates more benefit than harm.
That distinction matters because “retained root” describes very different clinical situations. A small, vital, immobile fragment buried in bone is not the same as an infected root with apical pathology. A mandibular third molar root touching the inferior alveolar nerve is not the same as a maxillary root displaced into the sinus. A primary molar fragment beside a developing permanent tooth is not the same as a root in the path of a future implant. Size matters, but anatomy, pathology, mobility, future treatment, and retrieval risk matter more.
Coronectomy has changed this discussion. Also called intentional partial odontectomy, coronectomy removes the crown of a mandibular third molar while deliberately leaving healthy roots near the inferior alveolar nerve. It is not an accidental broken root given a better name. It is a planned nerve sparing operation for carefully selected teeth.
The evidence is now difficult to dismiss. A 2024 systematic review and meta analysis found an 86 percent reduction in the relative odds of inferior alveolar nerve injury with coronectomy compared with complete extraction. A 2025 meta analysis involving more than 7,000 third molars found the risk of IAN injury was roughly one tenth that of complete extraction. A 2026 meta analysis reached the same general conclusion, finding about a 77 percent reduction in the relative odds of nerve injury in high risk cases.
The tradeoff is that some retained roots later migrate, become exposed, or require removal. Across modern reviews, secondary surgery generally occurs in only a small minority of cases, roughly in the low single digits. Root migration itself is common and is not automatically a complication. Roots often move coronally during the first year or two, frequently away from the mandibular canal. If later removal becomes necessary, the anatomy may actually be safer than it was during the original operation.
Professional guidance reflects the same principle. The American Association of Oral and Maxillofacial Surgeons recognizes coronectomy when a third molar requires treatment but complete root removal threatens the IAN. The Royal College of Surgeons also accepts coronectomy as a nerve preserving option in appropriately selected high risk cases. Both emphasize diagnosis, imaging, patient selection, informed consent, and follow up.
The lesson is not that broken roots are harmless. It is that completeness is not always the best measure of surgical quality.
Pathology changes the equation. Infection, persistent pain, mobility, caries, vertical fracture, cystic disease, exposure, or interference with implants, prosthetics, or orthodontic treatment all strengthen the case for removal. Contemporary oral surgery references also recognize situations where a root may appropriately remain because retrieval threatens the IAN, adjacent teeth, supporting bone, or maxillary sinus.
That is why a simple millimeter rule is inadequate. A tiny infected root tip may need removal. A much larger healthy third molar root complex intimately associated with the IAN may be exactly what should remain during a coronectomy.
Anatomy beats millimeters.
The maxillary sinus illustrates the point. A 1 or 2 millimeter apical fragment still embedded in alveolar bone near the sinus floor is fundamentally different from a root displaced freely into the sinus. Chasing the embedded fragment may create an oroantral communication or push the root into the sinus, converting a manageable problem into a larger one.
Once a root is actually inside the sinus, the risk calculation changes. Published cases document chronic odontogenic sinusitis associated with displaced dental roots, including one extraordinary report of a root retained in the maxillary sinus for more than 25 years that was eventually associated with chronic sinusitis and an Aspergillus fungus ball. Case reports cannot tell us how often this occurs, but they show why a foreign body in the sinus should not be treated like a tiny fragment safely buried in bone.
Imaging helps determine which situation you are facing. Panoramic radiographs are useful first line studies, but they compress three dimensional anatomy into two dimensions. A maxillary molar root that appears to project into the sinus may actually lie beside it. CBCT is not required for every extraction, but when knowing the true relationship to the sinus or inferior alveolar canal would change the treatment plan, three dimensional imaging can be decisive.
The same thinking should occur before high risk mandibular third molar surgery. If panoramic findings suggest an intimate relationship with the IAN, the dentist should consider before surgery whether the case is best managed by complete extraction, coronectomy, surveillance, or referral. Waiting until a root fractures to discover that the case was really a coronectomy candidate is less desirable than recognizing the risk in advance.
Pediatric dentistry provides another clear example. The American Academy of Pediatric Dentistry states that a fractured primary root tip is not automatically an indication for removal. If the fragment is easily accessible, it should generally be retrieved. If it is very small, deep, close to the permanent successor, or increasingly difficult to remove, leaving it to resorb may create less harm than damaging the developing permanent tooth. The anatomy changes, but the principle does not.
Operator limits matter too. There is a difference between a fragment that is anatomically dangerous to retrieve and one that is simply beyond the clinician’s current surgical ability. In the second situation, referral may change the equation. Fatigue, embarrassment, and sunk cost can become dangerous after 45 or 60 minutes of chasing a root. The rational stopping point can arrive before the technical stopping point.
Instrument debates often distract from this larger issue. Elevators, luxators, root picks, endodontic files, periotomes, and burs all have legitimate roles. The important question is not which instrument has the strongest fan club. It is whether the operator can gain adequate visualization and controlled access without sacrificing unnecessary bone or damaging vital anatomy.
Antibiotics should not become a psychological safety net. ADA guidance emphasizes that antibiotics are not routinely indicated for most localized pulpal and periapical problems in immunocompetent adults when definitive dental treatment is available. A retained root fragment by itself is not an indication for amoxicillin. Whether the root should remain and whether the patient needs systemic antibiotics are separate decisions. Spreading infection, fever, malaise, cellulitis, or systemic involvement may justify antibiotics, but medication does not compensate for inadequate source control when source control is required.
Patient communication requires the same precision. Telling a patient before surgery that a root will probably break and be left behind is too casual unless intentional root retention is actually the treatment plan. A better discussion explains that difficult anatomy or proximity to important structures can increase the possibility of root fracture, and that if a fragment cannot be retrieved safely, leaving it may create less risk than removing additional bone or injuring the nerve, sinus, adjacent tooth, or permanent successor. If coronectomy is being considered, it should be planned and consented as such.
When a fragment is retained, disclosure is only part of the responsibility. Document what remains, why it was left, its relationship to surrounding anatomy, what imaging was obtained, and what follow up is planned. Reassess if pain, infection, exposure, migration, sinus symptoms, or interference with future treatment develops. If a fragment has entered an anatomical space and cannot be directly visualized and easily grasped, blind retrieval is a poor substitute for imaging and referral.
The practical framework is simple. Pathology pushes toward removal. Dangerous anatomy can push toward retention. Future treatment changes the equation. Operator limitations may favor referral. Imaging should be obtained when it is likely to change what you do.
The old debate asks whether a good extraction leaves anything behind. Modern evidence asks a better question, which choice leaves the patient with the lowest probability of meaningful harm?
Join the Conversation!