Perio Surgery Went Exactly as Planned — So Why Did the Site Not Heal?
When a periodontal surgical site fails to respond as expected despite a technically correct procedure, the reason is almost always that the surgical plan itself was built on an incomplete picture of the defect. Three imaging-related gaps account for most of these cases: a wrong wall-count assumption, buccal or lingual bone destruction that was never visualized, and a furcation involvement graded lower than it actually was. All three trace back to the same source — 2D imaging that structurally cannot show what a defect looks like in three dimensions.
The Post-Surgical Scenario That Doesn't Add Up
The flap goes up cleanly. Debridement is thorough. If a graft was planned, it's placed and contained the way it's supposed to be. Sutures go in, the site is closed; healing instructions are given, and everything about the procedure itself was, by any reasonable clinical standard, done correctly.
Then the re-evaluation happens — three months, six months, sometimes a year later — and the numbers don't match the effort. Probing depths haven't reduced the way the case should have responded. Radiographic bone fill is minimal or absent. Occasionally there's continued attachment loss at the site, or a graft that seems to have simply not integrated into the way it should have.
The instinct here is to question technique — was the debridement thorough enough, was the graft material handled correctly, was the flap closure tight enough. Sometimes that's the answer. But in a large share of these non-responder cases, the surgery matched the plan perfectly. The plan itself was built on an assumption about the defect that turned out to be wrong. And the reason it was wrong almost always traces back to what a periapical radiograph or bitewing simply cannot show a three-dimensional bony defect.
Gap One — The Defect Wasn't the Wall Count You Planned For
Infrabony defects get classified by wall number — one-wall, two-wall, three-wall, or combinations — because the number of remaining bony walls is what determines whether a defect is a good regenerative candidate and what material or technique gives it the best chance of filling in. A well-contained three-wall defect behaves completely differently to a shallow, wall-deficient defect, and the surgical approach should differ accordingly.
Here's the problem: wall count is fundamentally a buccolingual question, and a periapical radiograph does not show the buccolingual dimension at all. It shows mesiodistal width and vertical depth reasonably well — which is exactly what makes it deceptive. The film looks informative. It gives you a number for how deep the defect appears to go. What it cannot give you, structurally, is any information about whether the buccal or lingual wall is present, thinned, or gone entirely, because that dimension is completely collapsed in a 2D projection.
Studies comparing CBCT to periapical radiographs against actual intra-operative measurements have found exactly this pattern: CBCT and periapical radiographs perform similarly for mesiodistal width and are not dramatically different for vertical depth in some analyses — but only CBCT can assess the buccolingual dimension at all, since a periapical radiograph provides no data on that axis whatsoever. Wall count assigned from a periapical film isn't really a measurement. It's an inference — often a reasonable one, but an inference nonetheless, made from a view that was never capable of confirming it.
The consequence shows the moment the flap is raised. A defect assumed to be a well-contained three-wall lesion, ideal for a straightforward regenerative approach, turns out to be missing its buccal wall entirely — effectively a one-wall or combination defect with nothing to contain a graft material the way the plan assumed it would be contained. The graft has nowhere to sit against, the membrane (if one was planned) doesn't have the architecture it needs to be stable, and the site doesn't fill in the way a true three-wall defect would have, because it was never that kind of defect to begin with.
This is one of the clearest illustrations of why bone-level imaging needs to be genuinely three-dimensional before a regenerative decision gets made — a point covered in more depth in this look at how 3D imaging changes periodontal treatment planning, which walks through how bone loss patterns that look straightforward on a 2D film often aren't.
It's worth being fair to the periapical radiograph here: for mesiodistal width and, in several comparative studies, for vertical depth, it isn't dramatically inferior to CBCT. The problem isn't that periapical imaging is universally unreliable — it's that it's selectively blind in exactly the dimension a wall-count decision depends on, while looking complete enough in the dimensions it does show that the gap rarely announces itself. A clinician reading a film that gives a confident-looking depth measurement has little reason to suspect that the same image is offering no information at all about the plane that actually determines whether the defect is regenerable the way it's been planned.
Gap Two — Buccal or Lingual Plate Destruction That Was Never Visible
Dehiscence — a vertical loss of the buccal (or occasionally lingual) cortical plate extending up from the alveolar crest — and fenestration — an isolated window-shaped defect in the plate that doesn't connect to the crest — are both, in a very literal sense, invisible to a periapical radiograph. The X-ray beam passes through the buccal cortex, the root, and the lingual cortex all at once, and everything superimposes into a single flat image. A missing section of buccal plate doesn't leave an obvious gap on the film the way a missing tooth would; it just quietly isn't represented, because the projection was never built to show that surface separately from everything behind it.
This isn't a marginal gap in detection. Research comparing CBCT with intraoral radiography for exactly this purpose has found CBCT to be significantly superior for detecting dehiscence and fenestration — not marginally better, but the kind of difference where one modality can see the finding and the other essentially cannot.
Clinically, this gap tends to surface at the worst possible moment: mid-surgery, after the flap is already reflected. A treatment plan built around an assumed intact buccal plate suddenly has to be improvised around a dehiscence or fenestration nobody knew was there — a different flap design might have been more appropriate, a different regenerative material or barrier technique might have been indicated, and the whole risk conversation with the patient beforehand was built on a picture of the defect that wasn't accurate. Even when the surgeon adapts well in the moment, an unanticipated dehiscence changes the healing prognosis in ways the patient was never actually consented for, because nobody knew to expect it.
This is the same category of problem covered from the surgical-planning side in DMD Imaging's imaging checklist for bone grafting and sinus lift cases — knowing the true state of the surrounding bone before you commit to a surgical approach, rather than finding out once you're already inside the flap.
There's a compounding risk here beyond the immediate procedure: an unanticipated dehiscence changes the biomechanics of primary closure. Flap design and suturing technique planned around an intact buccal plate may leave inadequate soft tissue coverage over a thinner-than-expected bony architecture, raising the risk of membrane exposure or graft material becoming compromised during early healing — which is frequently the actual mechanism behind a site that looked well-managed at surgery but failed to hold its healing trajectory over the following weeks.
Gap Three — The Furcation That Was a Grade Higher Than Recorded
Furcation involvement is graded clinically with a Nabers probe, checking horizontal attachment loss into the furcation entrance of a multi-rooted tooth. In principle, this is a direct physical measurement. In practice, it's limited by exactly the same anatomy it's trying to assess root trunk length, the angle of access at mesiopalatal and distopalatal furcation entrances on maxillary molars, adjacent tooth contacts, and simply how much room there is to maneuver a probe tip into a space that may only be partially open.
The result is a grading system that's frequently less reliable than it feels in the moment. One study comparing clinical furcation grading against actual intra-operative findings found agreement between clinical assessment and surgical reality was weak — with roughly a third of sites underestimated compared to what surgery actually revealed. CBCT-based assessment in the same study matched surgical findings far more closely, with underestimation occurring in less than one in ten sites. Other studies looking specifically at Grade III (through-and-through) involvement have found radiographic and CBCT-based detection identifying significantly more true Grade III sites than clinical probing alone managed to catch.
What this means in practice: a furcation recorded as Class I or borderline Class II based on probing and a periapical shadow can turn out, once the flap is raised, to be a full Class III through-and-through defect. That's not a small reclassification — it's the difference between a furcationplasty or regenerative approach and a root resection, tunnel preparation, or, in some cases, a decision that the tooth's prognosis doesn't support keeping it at all. When that distinction is made correctly before surgery, the procedure and the patient conversation both match reality. When discovered mid-flap, the surgery that was planned and the surgery that's actually needed are two different procedures, and outcomes for the site tend to reflect that mismatch.
Furcation mis grading is especially relevant in molars where periodontal and endodontic pathology overlap — a scenario examined in more detail in how 3D imaging changes the treatment decision in endo-perio lesions, where the true extent of bone loss at the furcation often changes which specialty and which treatment sequence should be leading the case.
There's also a directional pattern worth knowing: clinical probing tends to underestimate furcation involvement far more often than it overestimates it. In other words, the error almost always runs one way — toward a falsely reassuring grade, not a falsely alarming one. A clinician relying on probing and a periapical shadow is statistically far more likely to walk into surgery underprepared for what they'll find than to be pleasantly surprised that a case is simpler than expected. That asymmetry is exactly why a pre-surgical CBCT matters more for furcation cases than intuition might suggest — it isn't there to occasionally catch a rare surprise, it's correcting a systematic blind spot that shows up in the majority of borderline cases.
When These Three Gaps Combine
Individually, each of these three gaps is manageable to reason about. The genuinely difficult cases are the ones where more than one gap is present at the same site — which happens more often than clinicians tend to expect, particularly in maxillary and mandibular molars, where furcation anatomy, buccal plate thinness, and complex multi-wall defects naturally co-exist in the same crowded anatomical space.
A maxillary first molar with advanced periodontitis, for instance, commonly has all three problems layered on top of each other: a mesial or distal infrabony defect whose true wall count depends on buccal plate integrity, a furcation entrance that's difficult to probe accurately because of root trunk length and palatal root proximity, and a buccal plate that's frequently already thin or dehisced in that region even before disease progression thins it further. A periapical radiograph of a tooth like this can look moderately concerning — enough to justify surgery, not alarming enough to suggest anything unusual. The reality once the flap goes up can be considerably more complex than that single film ever suggested, because each of the three gaps was quietly compounding the other two.
This is precisely why relying on 2D imaging scales badly with case complexity. A straightforward single-rooted tooth with a simple vertical defect is far more forgiving of an incomplete picture than a multi-rooted molar with several interacting variables. The more anatomically complex the tooth, the more a 2D film's blind spots compound rather than simply add up — and molars are disproportionately represented in exactly the cases where periodontal surgery is being considered in the first place.
What These Three Gaps Actually Have in Common
None of these three scenarios involve a surgeon making a technical error. In every case, the procedure performed matched the plan made beforehand. The failure point sits earlier — in the imaging step that was supposed to tell the surgeon what they were actually dealing with.
A periapical radiograph can show you a shadow of bone loss. It cannot show you the true shape of a defect, the integrity of the buccal and lingual plates, or the real depth of a furcation. Every one of these gaps is a dimension the film was never built to represent.
This is precisely the gap that structured CBCT training for periodontal cases targets specifically — learning to read defect morphology, plate integrity, and furcation anatomy in three dimensions before a surgical plan is finalized, rather than discovering the true picture once the flap is already up. It's also worth noting that a defect site being evaluated today often becomes a future implant site if the tooth is eventually lost — which is where the anatomy questions extend naturally into implant-focused CBCT training as well.
Why This Matters Beyond One Surgical Site
A site that doesn't respond to periodontal surgery isn't just a clinical disappointment — it has real downstream costs. Regenerative material placed into a defect it was never suited for is material and cost that didn't help the patient. Bone graft, membrane, and biologics for a single site are not inexpensive, and when they're placed against an assumption about wall count or plate integrity that turns out to be wrong, that expense buys very little clinical benefit. A second surgical procedure to correct or manage what the first one didn't resolve means additional healing time, additional risk, and a patient who now has meaningfully less confidence in the treatment plan than they did the first time around.
There's a referral relationship cost too, one that's easy to underweight. A General Dentist who refers a periodontal case expects the Periodontist to come back with a plan that reflects what's actually happening at the site — not a plan that later needs revising mid-surgery because the defect turned out to be something else. Repeated experiences like that, even when nobody did anything technically wrong, quietly erode the confidence that makes a referral relationship work smoothly over time.
For General Dentists referring periodontal surgical cases, and for Periodontists managing their own case-planning, the practical shift is straightforward: treat 2D imaging as a screening step, not a planning step, for anything more surgically complex than routine scaling and root planing. The same discipline applies to reading a scan thoroughly once it's in front of you, a point covered more generally in this look at what most Dentists still miss after opening a CBCT — the procedure isn't the variable that needs scrutiny first. The completeness of the picture it was planned from is.
Key Takeaways
Wall count from a periapical film is an inference, not a measurement — the film has no way to show the buccolingual dimension that actually determines wall number.
Buccal and lingual plate defects are frequently invisible on 2D imaging because the cortical plates superimpose directly over the defect in a flat projection.
Clinical furcation grading underestimates the true grade in roughly a third of cases compared to surgical findings, largely due to probe access limitations.
All three gaps share the same root cause — 2D imaging cannot represent the three-dimensional shape of a periodontal defect, regardless of how carefully it's read.
A non-responding surgical site is rarely a technique failure — it's usually a sign the plan was built on an incomplete picture of the defect from the start.
Recognizing these three gaps after a disappointing re-evaluation is one thing. Learning to read defect morphology, plate integrity, and furcation anatomy correctly before the surgical plan is finalized is a different skill — and it's the specific focus of the CBCT Periodontics course at DMD Imaging Academy. For cases where the same site may later be considered for an implant, the CBCT Implantology course covers the related planning questions that come up once a tooth in a compromised periodontal site is eventually lost.
Frequently Asked Questions
-
In most cases, the surgery itself wasn't the problem — the surgical plan was based on an incomplete picture of the defect. The three most common imaging-related gaps are an incorrect wall-count assumption, undetected buccal or lingual plate destruction, and a furcation involvement that was graded lower than its true extent, all of which 2D imaging is structurally limited in showing.
-
Wall count depends on the buccolingual dimension of the defect, and a periapical radiograph is a flat, two-dimensional projection that doesn't capture that axis at all. It can show mesiodistal width and vertical depth reasonably well, which is what makes it easy to mistake for a complete picture of the defect.
-
Less accurate than it's often assumed to be. Studies comparing clinical probing against actual surgical findings have found meaningful rates of underestimation, particularly for distinguishing Class II from a true Class III through-and-through involvement, largely due to limited probe access at the furcation entrance.
-
Not every case — routine scaling, root planing, and shallow single-wall pockets rarely need it. But for cases involving suspected multi-wall infrabony defects, regenerative treatment planning, or furcation involvement in molars, a CBCT is one of the few ways to confirm defect morphology before committing to a surgical approach rather than discovering it mid-procedure.
