Furcation involvement

Furcation involvement is loss of the bone between the roots of a molar, starting where the roots divide. On a radiograph it shows as a radiolucency in the interradicular bone. The radiograph is specific but misses many involvements, particularly in maxillary molars, and doesn't grade extent, so furcation probing remains the reference [1, 2, 3, 4, 5].

Synonyms: furcation defect; furcation radiolucency; furcation bone loss; interradicular bone loss

What is furcation involvement?

The furcation is where a molar's roots divide. On a radiograph of a healthy molar, bone fills the furcation right up to the fornix, the roof of the angle between the roots [1]. Furcation involvement is the bone loss that empties that space.

Clinically, it's graded by how far a probe passes horizontally into the furcation. In the Hamp classes, Class I is horizontal loss of no more than one-third of the tooth's width (less than 3 mm), Class II is more than one-third without passing through, and Class III is through-and-through [2]. A radiograph doesn't grade these. It mostly tells you whether involvement is there [2, 5].

What does furcation involvement look like on a radiograph?

On a periapical radiograph, start at the alveolar crest and follow the bone into the angle between the roots. In a healthy furcation that bone reaches the fornix [1]. Involvement shows as loss of that interradicular bone [1].

The call is usually present or absent, not how much. In one study, intraoral radiographs typically picked up involvement only once it had reached Glickman Class 2 or 3 [5].

Maxillary molars add a sign of their own. A proximal involvement, at the mesial or distal furcation, can show as a small triangular shadow over the mesial or distal root in the furcation area. This is the furcation arrow [6].

It's a good sign when it's there and a poor one when it isn't. In the original dry-skull study, the arrow seldom appeared over uninvolved furcations, yet it was absent from a large number of Degree 2 and 3 involvements [6]. A clinical study of 164 maxillary molars, checked against the finding at surgery, put its sensitivity at 38.7% and its specificity at 92.2% [7]. Examiners agreed on whether an arrow was there with a kappa of only 0.489 [7].

Why is it easier to see in mandibular molars than maxillary molars?

Because of overlap, though not every study finds the gap. In maxillary molars the roots overlap on the radiograph, so furcation involvement may not show at all [4].

Site-level data fit that picture. In a retrospective study of first molars, CBCT matched the clinical finding more closely than intraoral radiographs did, most of all at the distopalatal furcation of the maxillary first molar (16, 26) [5]. In a panoramic study reported by a 2022 systematic review, readings agreed with the surgical finding better for mandibular lesions, for experienced and less experienced examiners alike [2]. Deep learning models reading radiographs were also more accurate on mandibular molars, with sensitivity 0.96 and specificity 0.97, against 0.93 and 0.94 across all molars [8]. Those figures describe algorithms, not clinicians.

The pattern isn't universal. In a 1980 study that compared radiographs with probing, radiographs found furcation involvement in maxillary molars more often than the clinical examination did, while in mandibular molars the clinical examination found more [9]. That study used a periodontal probe and an explorer rather than a furcation probe [2].

Probing has an arch problem of its own. Against CT, probing agreed best at mandibular buccal furcations (κ 0.52), less well at maxillary buccal furcations (κ 0.38), and lower still at lingual (0.35), mesial (0.29) and distal (0.27) furcations [10].

Keep visibility and frequency apart. Furcation involvement is also more common in maxillary molars: one study found it three times as often as in mandibular molars [9], and a Swedish radiographic survey found the maxillary first molar most often affected and the mandibular second molar (37, 47) least [1]. A periodontitis clinic sample went the other way for first molars, with mandibular first molars (36, 46) affected in 64.5% of cases and maxillary first molars in 58.5% [11]. Either way, that's how often it happens, not why it's harder to see.

How well does the radiograph agree with clinical furcation probing?

Moderately, and lopsidedly. Radiographs are highly specific, and furcation probing is more sensitive [3].

Across the studies in a 2022 systematic review, intraoral radiographs agreed with the reference finding (open-flap surgery, CBCT, and in one study clinical probing) in 38.7% to 83.1% of furcations, overestimated in 12.7% to 16% and underestimated in 0.9% to 38.9% [2]. The low end of that range is the furcation arrow's 38.7% sensitivity [7]. The spread has causes. The studies used different reference standards, and the location of the furcation entrance had the largest influence on validity [2].

Probing isn't a perfect reference either. Against open-flap surgery, the probed class was confirmed in 56% of furcations, overestimated in 15% and underestimated in 29% [12]. For Class III involvement, radiographs got 57% right and probing 32% [12].

The two also measure different things. Probing reads horizontal and vertical clinical attachment, while a radiograph shows the bone level, which makes a straight comparison between them questionable [2].

Whether experience improves radiographic reading is unsettled. One study found reader experience mattered (weighted κ 0.618 experienced against 0.426 less experienced) [12]; another found no effect of experience [3].

A Nabers probe remains the reference for diagnosing furcation involvement, according to the systematic review [2], and two of the primary studies conclude that radiographs are best used alongside furcation probing [3, 12].

What else produces a radiolucency in the furcation?

Not every radiolucency between the roots is periodontal. The pulp can reach the furcation directly. Accessory canals open into the furcation region of permanent molars: they were found in 28.4% of extracted molars in one dye study [13], and patent canals on the pulp chamber floor in 16% to 24% of extracted molars in another [14]. An endodontic-periodontal lesion is a pathological communication between the pulpal and periodontal tissues of one tooth [15].

A periapical lesion reaching the furcation doesn't need a row of its own. It's an endodontic lesion arriving by another route. In root-filled molars, periapical lesions larger than 5 mm were associated with furcation involvement [18], and a case report describes a lesion of endodontic origin extending towards the furcation [17].

The distinction matters for the tooth. Endodontic-periodontal lesions that involve root damage, such as a perforation, have the worst prognosis [16].

What it means clinically

Furcation involvement is one of the complexity factors used to stage periodontitis, alongside pocket depth, infrabony defects, tooth hypermobility and masticatory dysfunction [21]. Its extent is measured with the probe, since the radiograph mostly records presence [2].

How to be sure

• Probe it. A systematic review found a Nabers probe remains the reference [2], and one study concluded radiographs are best used in combination with furcation probing [3].
• Mind the horizontal angle. In simulated defects in mandibular first molars (36, 46), detection depended on the horizontal angulation of the X-ray beam relative to the roots. Accuracy didn't differ between -10 and 20 degrees, and the angulations used for proximal caries detection suited furcation defects too [22].
• Keep CBCT for the hard cases. Against intrasurgical measurement, CBCT's pooled sensitivity and specificity were both 0.98 [23]. One review concluded it should only be used where clinical evaluation and conventional radiographs don't give the information needed [24].

Common errors

• Reading no arrow as no involvement. The furcation arrow's sensitivity was 38.7% in the clinical study [7], so its absence settles nothing.
• Grading from the radiograph. An intraoral radiograph shows presence, not the Hamp class [2].
• Treating the radiograph's call as certain. Across studies, intraoral radiographs overestimated involvement in 12.7% to 16% of furcations [2].
• Assuming every furcation radiolucency is periodontal. Accessory canals can carry the pulp to the furcation, and the study that counted them stressed the need for a differential diagnosis when pulpal-periodontal disease exists [13].
• Assuming the radiograph always loses to the probe. For Class III involvement, radiographs got 57% right against 32% for probing in one surgical comparison [12].