Our main finding is the extreme variability of the number and topographic distribution in midline and lateral LF. The first consideration to be pointed out is that the absence of LF is quite rare but possible, despite several studies have reported at least one LF in their series. It could be postulated that we wrongly missed the LF in those two patients of our series, but also Sekerci et al. [11] and Demiralp et al. [12] found few patients without LF in their series, in 1.8% and 3.4% respectively, thus supporting our results. Concerning the maximum number of LF per patient, we found up to seven LF in one out of 300 patients and eight LF in another patient. Indeed, a large number of LF can be rarely encountered as highlighted by previous studies. Similarly, He et al. found up to seven LF in 2% of their patients [13]. Notably, Patil et al reported up to 11 LF in 0.5% of their CT study on 300 patients [14]. However, the total number of LF per patient is generally quite lower. In fact, in our series, two LF were observed in 32.3% of patients, three in 27%, and four in 17.7%. These results are in line with those previously published, given that two LF were reported in 28.2% of patients by Sekerci et al, in 34% by Patil et al., and in 32.8% by Demiralp [11, 12, 14]. Conversely, higher variability has been reported concerning the frequency of patients with only one LF that was 10% in our series, similarly to what reported by He et al. (12%) and Demiralp et al. (10.3%) [12, 13], but quite lower than in other series in which one LF has been observed as the most common rate of LF (in up to 40% patients)[14, 15].
Concerning the midline LF, two previous studies confirmed the chance of having no LF in the middle portion of the symphysis with frequencies of 3-3.8% [16, 17]. Nevertheless, no other studies reported more than four LF in the midline, although we found five LF in one patient. We also found four LF in 4/300 patients (1.3%), which is a rare but possible condition also reported by Sheikhi et al. and Wang et al. in 2.9% and 2% of cases, respectively [16, 18]. Further, we reported the highest frequency of patients presenting two midline LF ever published (in 57.3% of cases). Of note, two LF has been the highest frequency of midline LF also in several previous studies, having been observed in 53.9% by Von Arx et al, in 52.9% by Sheikhi et al., in 43.6% by Wang et al, and in 43% by Rosano et al. [5, 16–18]. As a matter of fact, other studies have reported one LF as the highest frequency of LF in the midline, with frequencies reaching 72% [19], while one midline LF was observed in 22.7% of our patients. On the other hand, our frequency of three midline LF (17.7%) is in line with previous studies, in which the frequency of three LF per person has never been reported as the most common rate of midline LF. Regarding the relationship with the genial tubercle, we found that a midline LF above the tubercle is almost invariably detected, similarly to what reported by Tagaya et al (95%) and Sheikhi et al (99%)[18, 20]. Further, when a second midline LF is identified, it is located below the genial tubercle in most of the cases. Out of 300 patients, 178 (59.3%) presented at least one midline LF below the genial tubercle, with a slightly lower frequency to that reported by previous studies with frequency ranging from 74.5–85% [18, 21].
The frequency of lateral LF reported in literature is quite variable. We observed lateral LF in 62.3% of cases, similarly to Liang et al. (62%)[19], although this frequency was much lower in the series by He et al. (30.1%) and higher in that by Tagaya et al. (80%)[13, 20]. Generally, lateral LF are bilateral and symmetrical, as already reported by previous studies [22]. The total number of LF has been also investigated by Xie et al. who reported one lateral LF in 37.3%, two in 19.7%, three or more in 5,4% [23], in much the same way of our study (one lateral LF in 36.2%, two in 17.7%, three or more in 0,4%). Concerning the relationship between the position of lateral LF and the teeth, we observed 13% of LF in the zone of incisors, 9.8% canines, 46.6% premolars, and 5.4% molars, confirming what already highlighted by other authors on the highest frequency of lateral LF in the zone of premolars [11, 15, 17, 23]. We have resumed in Table 3 the number of LF reported in previous CT and cadaveric studies conducted on different populations.
Table 3
Number of LF reported in different CT and cadaveric studies.
| Study | Country | Sample | LF |
| Our Study | Italy | 300 patients, CBCT | 592 midline, 283 lateral |
| Von Arx et al [17] | Switzerland | 179 patients, CBCT | 86 midline, 131 lateral |
| Trost et al [24] | Germany | 460 patients, CT | 613 midline, 231 lateral |
| Sheikhi et al [18] | Turkey | 102 patients, CBCT | 205 midline |
| Sekerci et al [11] | Turkey | 500 patients, CBCT | 476 midline |
| Liang et al [19] | Belgium | 555 patients, CT | 132 midline |
| He et al [13] | China | 200 patients, CBCT | 683 |
| Rosano et al [5] | Italy | 60 cadaveric mandibles | 118 |
| Vandewalle et al [3] | Belgium | 354 dry mandibles | 347 |
No previous studies have investigated the association of mandible width with number and distribution of LF, thus a comparison with the literature is not possible. We did not find any statistically significant association with mandible width, as well with gender, although a progressive increase of the number of lateral LF has been observed from the first group (with the smaller mandible width) presenting 78 LF, to the second group with 99 LF and the third group (with the bigger mandible width) with 106 LF, thus some considerations should be pointed out. In the first group, only 1% of subjects showed three or more lateral LF, while in the second and third groups the frequency was 6% and 9%, respectively. Further, considering only the premolars zone, 61% of patients of the first group did not show any lateral LF, as well as the 53% of the second group and the 43% of third.
Some limitations of our study should be considered. First, the relatively small sample size of our series, indeed, we cannot exclude that a larger study population would have allowed to obtain a more powerful statistical analysis, even reaching interesting association of the number and anatomic distribution of LF with mandible width. Second, we did not evaluate the distance of LF from the alveolar ridge and tooth apex that, in turn, could be an important pre-operative finding to be evaluated before proceeding with implant procedures. Third, the methodology that we used to estimate the mandibular width has not been previously validated, but allowed us to investigate the correlation between the distribution of LF and mandible width. However, it is not felt that this fact was a real limitation, but rather something that deserves further investigation. Last, our CBCT analysis is based on the detection of bone canals, rather than the direct visualization of the vessels. The differences in the number and distribution of LF with some previous studies might be partly related to the different approaches used to assess the LF foramina, for instance through cadaveric skulls which might enable to demonstrate a higher number of LF than CBCT studies. The limited, albeit high, spatial resolution of CT probably might be responsible for a lower detection rate of the LF, although according to several authors CBCT provides highly accurate data concerning mandible anatomy and state that the different frequencies reported in literature is mostly related to the anatomical variability related to different geographical regions. For instance, Rosano et al [5] found LF in 100% of cases in their cadaveric study, while Tagaya et al [20] published a double study on five cadavers and 200 patients using CBCT reporting the occurrence of LF in all cadavers and in 95% of patients.
