One of the main reasons for tooth re-treatment is cusp fracture. A differential diagnosis is required to decide whether to amputate the cusp, retain it, or shoe it.

When teeth are painful, typically to hot, cold, and biting pressures, the fractured cusp must be amputated. If, however, the tooth is asymptomatic, amputation may be elective and shoeing can also be considered see Cusp Shoes and Bevels The location and extent of the crack also guides the decision.


If there is a visible complete crack in the buccal or lingual enamel, there is no choice: the cusp must be amputated in its entirety.

As well, if the external cavosurface has visible cracks at the edges of the cusp, but not joined, then there is still little choice. This external cracking almost always follows internal cracking, and only a small amount of uncracked dentin remains.The whole cusp must be amputated.


But if the tooth has no visible external cracks, has not developed symptoms, and a dentinal crack is encountered after an old restoration is removed, for example, in the photo above, at the pulpo-axial line angle of a Class II,what are the treatment options?

If the crack is visible only in dentin, and not externally, we have three clinical choices; crack-bonding, intentional fracture, or shoeing.

  1. Crack-bonding: the crack is followed with a small bur to its terminus within dentin, and bonded
  2. Intentionally amputate, making a premptive cut on the outside of the tooth to lead the fracture into a desirable final cavosurface location. After the cusp is removed, the crack “fault line” can be followed with a bur to eliminate it from the preparation.
  3. Do neither of the above: shoe the cusp


A first option is to follow the crack into the heart of the tooth, a half mm. or so, using a 330 bur, wet, at low rpm, or a small round bur, like a #2 round, or a fissurotomy bur, at slow speed, in a high speed handpiece. Preserving a cusp this way reduces the complexity of restoring the tooth and conserves the maximum viable tooth structure.

Often the end of the crack is quickly reached. Transillumination of the tooth from the outer surface is necessary to determine if the crack terminus has been reached. An Addent Microlux with 2mm or 3mmcontrangle  tip is helpful for this task.

However, vision is an imperfect sense. If the crack terminus has not been actually reached, symptoms might develop after restoring. To minimize this risk,follow this protocol: note that this protocol is based upon a total etch bonding system see Standard Bonding Protocol . In synopsis, this system uses low viscosity components; liquid 37% phosphoric acid etch, Microprime B as a primer/desensitizer, and Kuraray Clearfil Photobond , a dual-cured Generation V low viscosity bonding agent, achieving a film thickness of 5 to 7 microns.

Crack bonding will not work with thick SE adhesives, for example, with film  as thick as 40 microns; the viscosity is too high to effectively penetrate a crack.

Viscous etchant gel, thick bonding agents, and various compilations of S/E and One-step systems all conspire to  fail in crack bonding for the same reasons.

  1. Etch with liquid 37% phosphoric acid for ten seconds, rinse, blow dry but do not dessicate
  2. Apply Microprime B, remove excess with air.
  3. Apply Microprime B again, blow out completely. We are trying to encourage capillary ingress of the primer into the crack. As the water evaporates it leaves a concentration of HEMA in the crack, which will invite the bonding agent to penetrate.
  4. Apply bonding agent, air thin, do not cure. The alcohol in the bonding agent is chasing out the residual moisture of the primer.
  5. Reapply bonding agent, air thin, and cure. We are attempting to ensure that the entire volume of crack is now filled with adhesive. This presents the best likelihood of penetration of a crack extension incompletely removed with the 330 bur.

I repeat, a bonding protocol involving aqueous etch, separate primer, and thin bonding film is essential to success. Any protocol using gel etch, or thick S/E bonding agents, 30 to 50 microns thick, have lessened probability of crack-penetrating capacity.

  1. Following double-bonding, fill the undermined cut following the crack with a small amount of flowable, adapting it with an explorer tip and popping bubbles.
  2. Fully cure this tiny amount of flowable. This minimizes its contraction.
  3. Build triangulation to protect the weakened cusp from contraction force deriving from large composite volumes against a small volume of weakened tooth structure.
  4. Advise the patient. A suitable dialogue states the following: ” We have an unknown factor in this situation, and we have begun with the most conservative approach. However, biological systems are not entirely predictable, and there is a chance that symptoms may develop, which is very unlikely according to our experience, and would be limited to temperature sensitivity that does not dissipate shortly after this treatment. If necessary we will rectify that problem by removing the cusp and adding to the restoration. This could shorten the life of the restoration or require  escalation to a more expensive onlay or crown. However,experience has shown that when the tooth is asymptomatic in the first place, this is almost never necessary.”


When a dentin crack is incomplete, i.e., has not progressed all the way to the outer enamel, but you wish to amputate, how best to amputate? This following method works well.

Choose a desirable break-out point on the outer surface of the tooth. With a 330 or 1157 high speed bur,wet, cut through the enamel into dentin, aiming towards the cracked pulpoaxial line angle. When the the entire cusp has been suitably undermined, grasp the cusp with Howe pliers and flex towards the middle of the tooth, not to the outside of the tooth. If the cusp does not fracture readily, deepen the cut, avoiding encroachment into the pulp zone.

The goal is to intentionally create a path that will initiate controlled breakage to a restoratively benign margin. See the adjacent photo sequence

Accidents with the upper molars

Without the above precautionary measure, the lingual cusp of upper molars may fracture into a pulp exposure. This is because the lingual pulp horn is large, and the typical pulpoaxial line angle of a Class II is located mid-tooth. The path from the pulpoaxial line angle to the cavosurface may lead diagonally to the CEJ, intersecting the pulp in that pathway.

A second disaster that may occur if one does not create an intentional exit point, is fracture below the CEJ subgingivally to the osseus crest, possibly dooming the tooth to extraction. This is a hard pill for a patient to swallow if they presented asymptomatically, needing only routine amalgam replacement. Then their dentist found the crack, broke the cusp off hopelessly, and the tooth had to be extracted. Bad story. This is great fodder for anti-dentist conversations at social gatherings, and it is hard chairside to represent such a situation positively as something other than operator error. Prudence suggests: inform the patient pre-operatively and make a precautionary fracture path!

If a thin and weak cusp presents with almost no dentin support,the path of fracture is self-evident – a straight line to the nearest cavosurface.


See Cusp Shoes and Bevels


After amputating, ensure that the gingival margin is strong and free of crumbly or shattered remnants of enamel. A scaler works well to remove friable enamel.

As well, ensure that there are no sharp dentin internal forms which can initiate resin fracture, see Dentin Axioms

Then a bevel will be required on the cavosurface, from 6 to 12 degrees on enamel if mid-crown or higher, a 45 degree bevel if close to the CEJ. See Axiom #5 always Proportionate Bevels

See also articles by the author,“Preparations In Composite Resin Part I: Principles And Instrumentation For Class V, Cusp Tips, And Incisal Attrition” Oral Health Journal, December 2011 Pgs.48-61

and “Design Principles for Class II Preparations” Oral Health Journal December 2012,pgs 60-68

A second follow-up article, detailing instrumentation, was published in July 2013, at “Modifications to Class II GV Black Preparations for Composite Resin”, Oral Health Journal, July 2013.

“WET-PACK MARGIN PLACEMENT TECHNIQUE See Wet-pack margin placement method

Restoration completed with the Bandbender

Lower Molar 3 – surface preparation

When restoring a beveled margin on the buccal or lingual where finishing is accessible, supragingival and  non-iatrogenic, a wet-pack margin placement method ensures a hermetically sealed and cosmetically perfect margin.

Place a small amount of flowable with a periodontal probe on the margin, then use a regular Bendabrush to wick up excess. The amount of flowable placed should constitute, conceptually, not more than needed to take up irregularities and potential voids between the margin and the final resin.Think of it as a thickened layer of bonding agent. Then a full-bodied resin is condensed over this, while still uncured..

This method eliminates both bubbles and unsightly lines of flowable at the cavosurface that arise when one places flowable and cures before placing final resin. The downside of the “wet margin” method is that there may be considerable flash expressed as the main resin is condensed over top of the wet flowable. The flowable may form part of the final resin where it is not wanted. Hence the quantity must be judiciously limited.

Never use wet-pack interproximally. Finishing  interproximally with rotary burs is iatrogenically dangerous. There are schools of thought that do advocate “Injection Molding” interproximally but this writer considers it a flawed concept.