When a Crown Would Have Made It Worse: The Sandwich Technique for Worn Upper Anteriors
- Jun 25
- 5 min read
My patient Beth came in full of life. Talkative, vivacious, the kind of patient who lights up the room. And her teeth were melting away.
Acid erosion. Dry mouth from a Sjögren’s diagnosis. Old crowns on most of her posterior teeth. Composite patches on the anteriors where her previous dentist had been trying to stay ahead of the damage. She knew what she wanted — she showed me pictures from her late twenties and said, I want those teeth back. Same shape. Same feeling. I just don’t want to look like I have veneers.
When I looked at her intraoral photos, I could see exactly what she meant. The lingual cervical enamel on her upper anteriors was largely intact. But from the middle third to the incisal edge, everything was thinning out in her functional envelope. The interproximals were demineralized and dark. Her laterals were so worn I couldn’t add length without solving the occlusion first. And she’d had premolar extractions, which meant everything was tight.
She needed coverage on the facial and the lingual. And that’s where the decision gets interesting.
Why Not a Crown
The obvious answer is a full-coverage crown. You need the front covered, you need the lingual covered, just do a 360. But look at what that means for a tooth that’s already been thinned by years of acid exposure. You’re removing what’s left. You’re grinding down structure that isn’t there to give.
There’s also a bilaminar option — two separate porcelain veneers, one facial and one lingual. It works in theory, but the cost doubles because the lab charges for two restorations per tooth. And the seating has to be perfect, because if the lingual doesn’t seat exactly right, it affects the facial. The margin for error is thin.
Here’s the pearl: when you need facial coverage and lingual coverage and you don’t want a crown, do composite on the lingual. Inject it through a matrix, get your coverage, and if it ever needs to be replaced, you remove it, micro-abrade the surface, and re-inject. It’s a failure of the material, not the tooth. The porcelain stays untouched. The template exists. The fix takes a single appointment.
Setting the Case Up
Before I touched a tooth, I did Invisalign. Six months or less. I intruded her lower anteriors to give me more room, pushed the upper anteriors out just enough to stay conservative — leaving about a millimeter for the veneer material — and widened the buccal corridors. Because I was treating the full upper arch, I could push everything out together, which gave me more to work with and let me keep every anterior tooth no-prep or near no-prep.
This is the thing I want everyone to hear: when you have width-length discrepancy, the more teeth you include, the less you have to remove from each one. That’s not a reason to do 14 veneers on every case. It’s a clinical reality that sometimes the conservative option is the bigger treatment plan.
The Lingual Injection Mold
I had the lab design the facials of her existing teeth and add the linguals of the upper anteriors only. The key was not designing both surfaces on every tooth — if the full arch had been designed, composite would have flowed all over the facial surfaces during the injection. One surface per tooth, then the other.
And always, always use retraction cord for lingual injection molding. I did the first pass without it on Beth’s case. It was fine. By the time I cleaned up and was ready for the second pass, there was bleeding. Cord first, every time, soaked in hemostat. Don’t learn that one the way I did.
The Slice Preps and the Smile Design
For the interproximals, I used slice preps so the porcelain could wrap into the dark, demineralized areas between the teeth without me having to prep the full facial surface. It’s a targeted approach — you’re covering what needs to be covered without removing what doesn’t.
On smile design, Beth’s canines were not at the zero repose position. Textbook says they should be. But I wasn’t going to shorten canines on a patient where I’m trying to give her back what she had. So I accepted the canine position as my benchmark and built the smile curve from there. The goal wasn’t to hit every metric — the goal was to not make anything worse while giving her teeth that looked like hers.
I raised the tissue on #6 and #7 to harmonize the gingival line on the right side, which had been dropping slightly. Electrosurgery for the tissue. I left the bone alone — her bone was thick but not so far out that it would rebound badly. For a case where the bone is jutting, you have to flap it. This one didn’t require that.
The Final Result
The ceramics were feldspathic, made by Juan Rigo. They were beautiful. You could see the lingual composite sandwiched behind the facial porcelain, protecting the little tooth structure remaining underneath. Tissues healed well. She looked exactly like her late-twenties photos.
Beth didn’t want to look like she had veneers. I think we got there.
What We Covered in This Discussion
Why the sandwich technique — porcelain on the facial, injection-molded composite on the lingual — is the preferred alternative to a full-coverage crown when both surfaces need coverage on a structurally compromised tooth
The clinical and cost reasons why bilaminar porcelain (two separate veneers per tooth) is a less practical option than composite on the lingual
How to sequence the lab design for lingual injection molding: facial design only on anterior teeth, lingual design added separately, to prevent composite flowing onto unprepared facial surfaces
Why retraction cord is non-negotiable for lingual injection molding, and what happens when you skip it
The width-length discrepancy principle: when discrepancy exists across the arch, including more teeth in the restorative plan often allows each individual tooth to be prepared more conservatively
Slice preps for interproximal coverage: how to wrap porcelain into demineralized, discolored interproximal surfaces without a full facial prep
Smile design when the patient’s anatomy doesn’t match textbook metrics: using the existing canine position as the benchmark and building from there rather than reducing teeth to hit an ideal number
Managing subgingival crown margins under isolation: sequencing posterior crown cementation before rubber dam placement when the margin depth makes full-arch dam placement impractical
Are you a member? Log in to your account to watch this discussion: The Sandwich Technique
