The uncinate process (UP) has high anatomical diversity, including variations such as bullous variants, medial/lateral deviations, and the superior attachment of the uncinate process (SAUP). It is essential for surgeons to know the anatomical variations in this region before an operation, especially in functional endoscopic sinus surgery (FESS), to avoid complications. The primary method used in the radiological evaluation of the paranasal region is computed tomography (CT) [1,2].

The nasal uncinate process (NUP), together with the middle turbinate, is among the key parts that form the ostiomeatal complex (OMC). It is the first anatomical barrier surgeons encounter when entering the nasal cavity [3,4]. The NUP is usually removed during endoscopic sinus surgery (ESS). This technique is known as uncinectomy. In the presence of a post-operative remnant, the frontal FESS procedure terminates with failure. Therefore, SAUP has great importance in the management of ESS. It is necessary to know where the NUP attaches in order not to damage the ethmoid roof [5]. The easiest way in radiological reporting is to specify the SAUP as lamina papyracea, middle turbinate, or skull base. However, there are two different classifications formally for this subject, known as the Landsberg and Friedman (LF) and Stammberger and Hawke (SH) classifications [6,7]. SH types are denoted by roman and LF types by arabic numerals. The LF classification includes four groups, and SH comprises six groups [6,7]. Different classifications have been used in studies. For example, Barosso [8], Srivastava [7], and Doğan [9] used the LF classification, while Güngör [6], Kansu [5], and Ercan [10] used the SH classification in their studies. This situation creates confusion, makes comparison difficult, and prevents a healthy meta-analysis study. The last meta-analysis study in the English literature on this subject belongs to Papadopoulos in 2021 [1]. In this study, variations in almost all points of the paranasal region were compared. However, the NUP part was probably overlooked due to this situation [1].

To the best of our knowledge, in our study, SAUP classifications (LF and SH) are compared for the first time. In addition, having a confusing anatomical structure with ethmoid cells and many variations makes it difficult to follow the NUP, especially on the roof, thus causing non-standard subjective evaluations. In the Material and Methods section, the common pitfalls of SAUP evaluations are also emphasised. Our study aims to identify the common and diverging points in SAUP classifications and to draw attention to the pitfalls in these evaluation processes.

The uncinate process is a hook-shaped structure and is one of the four anatomical regions called by the same name in the body. Apart from the NUP, a part connected with the head of the pancreas is called the uncinate. In addition, there is a neural tract in the brain called the uncinate fasciculus. Furthermore, there is an uncinate process of the cervical vertebra. Therefore, the abbreviation used is the nasal uncinate process (NUP) instead of the uncinate process (UP), unlike other studies [11,12,13]. Our study aims to identify the common and diverging points in the superior attachment of the nasal uncinate process (SAUP) classifications and to draw attention to the pitfalls in these evaluation processes.

This study was approved by the Muğla Sıtkı Koçman University Human Research Ethics Committee with the document number 200022/2020. Between January 2018 and December 2020, 239 healthy patients who underwent paranasal CT scans for various indications were retrospectively evaluated. All the images were obtained from picture-archiving communication systems (PACSs). Thirty-two patients were excluded from the study for the following medical reasons: six patients had a history of uncinectomy operation, two had polyposis, four had rhinitis, eleven had sinusitis, four had major deviation, one had NUP medialisation, two had NUP lateralisation, and two had ethmoid hypoplasia. In addition, seven CTs with artefacts were not appropriate for evaluation. They were also excluded from the study. In total, 200 patients were accepted for general evaluation, and 400 sides were bilaterally evaluated. Power analysis was performed with the G-power test. The sample group was calculated at 44 for 95% power, 0.05 alfa, and 0.2 beta parameters. In conclusion, our study population size was fairly sufficient for analysis.

CT scans were performed with a 256-slice multi-detector CT scanner (Somatom, Siemens Healthcare, Erlangen, Germany). The patients’ position was prone, with the head in extension; the slice thickness for coronal imaging was 1.5 mm, and for axial imaging, it was 1 mm. The slices were perpendicular to the intraorbital meatal line, with dose parameters of 100 KVp and 40 MAs. The evaluation was performed at a workstation with a high-resolution medical monitor in a paranasal window (window width (WW): 2000–2400 Hounsfield unit (HU), window level (WL): 400–450 HU) and bone window (WW: 1800 HU, WL: 400 HU). The palatine bone for coronal imaging and the orbitomeatal line for axial imaging were taken as the reference lines. The CT sections started from the anterior wall of the frontal sinuses and ended with the posterior wall of the sphenoidal sinus.

All the images were independently assessed by two experienced radiologists on the coronal, axial, and sagittal planes. In case of contradictory results, the images were evaluated by both radiologists together.

The data were stored in a Microsoft Office Excel file (Excel 2010, Microsoft). We used the SPSS software (version 22.0, IBM) for statistical analysis. Continuous variables are expressed as mean ± standard deviation (SD) values. Qualitative variables were counted and calculated as percentages. All data were statistically compared according to gender. The Pearson chi-square (χ²) analysis and the Spearman and Kendall tau b test were used to evaluate the relationship between the variables. P values < 0.05 were considered statistically significant, and values of <0.01 were highly significant.

In total, 200 patients (males/females, 100/100; mean age/ range: 37.13 ± 16.14/16–84) were included in this study. In all the patients, 400 NUPs were evaluated according to SAUP classifications: 200 on the right side and 200 on the left side.

On the right side, according to the SH classification, 72/200 (36%) of the patients had Type I SAUP, 90 (45%) had Type II, 25 had Type III (12.5%), and 13 had Type IV (6.5%). In the evaluation of the same NUPs according to the LF classification, 13 (6.5%) of the patients had Type 0, 14 (7%) had Type 1, 63 (31.5%) had Type 2, 8 (4%) had Type 3, 34 (17%) had Type 4, 43 (21.5%) had Type 5, and 25 (12.5%) had Type 6 SAUP (Table 1).

The equivalents of all the LF Type 0 SAUPs were ST Type IV. All patients with LH Type 1 were in the SH Type I group. However, 58/63 of the patients with LF Type 2 were SH Type I, and 5/63 of them were SH Type II. LF T3, T4, and T5 were completely included in the SH Type II group. However, the majority of SH Type II group were in the LF T4 and T5 groups. Type II was the most inhomogeneous group. The patients were distributed among four separate groups, namely LF T2, 3, 4, and 5. All the LF T6s were Type III.

On the left side, 98 (49%) of the patients had Type I SAUP, 67 (33.5%) had Type II, 24 had Type III (12%), and 11 had Type IV (5.5%) according to the SH classification. In the LF classification, 11 (5.5%) of SAUPs were Type 0, 8 (6.5%) were Type 1, 93 (46.5%) were Type 2, 11 (5.5%) were Type 3, 20 (10%) were Type 4, 33 (16.5%) were Type 5, and 24 (12%) were Type 6 (Table 2).

In total, according to SH, 170/400 (42.5%) were Type I, 157/400 (39.25%) were Type II, 48/400 (12%) were Type III, and 24/400 (6%) were Type IV. According to LF, 24/400 (6%) were Type 0, 22/400 (5.5%) were Type 1, 156/400 (39%) were Type 2, 19/400 (4.75%) were Type 3, 54/400 (13.5%) were Type 4, 76/400 (19%) were Type 5, and 48/400 (12%) were Type 6.

In comparison, the following observations were made:

SH Type I: 22/170 (12.9%) were Type 1, while 148/170 (87.1%) were Type 2 according to LF;

SH and LF classifications were separately compared using the Pearson chi-square (χ²) and the Spearman and Kendall tau B (τ) tests on the right and left sides. In the chi-square (χ²) test, the Pearson correlation value corresponding to 1 for LF was 0.292 for the right side and 0.370 for the left side in the calculation of ST; the correlation coefficient in the Kendall tau b test was 1 for LF and 0.543 for ST on the right and 0.587 on the left. In the Spearman test, the correlation coefficient was 1 for LF, 0.528 for ST for the right side, and 0.577 for the left side. The approximate b was 4.292 for Pearson chi-square (χ²), while it was 8.753 for the Spearman tests. According to these parameters, there was a highly significant relationship between both classifications. In addition, there was no statistically significant difference relative to gender or side.

SH and LF classifications are used in assessing the SAUP of NUPs. Stammberger and Hawke first classified SAUP in 1991 [17]. In the following years, Landsberg and Friedman (1996) developed a different classification system and classified SAUPs into six groups [6]. However, since ST Type IV does not have an equivalent in LF, it was modified by adding Type 0 to the classification [16]. We observed a statistically high degree of correlation between the two groups (p < 0.01). This relationship is due to the fact that the SH groups are compatible with only some groups in LF. One-to-one compatibility was determined, respectively, between SH IV/III and LF 0/6. For other groups, such harmony cannot be mentioned, but SH I and LF 1/2, and SH II and LF 2/3/4/5 are matched. However, without seeing the CT images of the patients who were evaluated in terms of NUP, it is not possible to find the equivalent of the groups. Therefore, a comparative evaluation is also not possible between studies. Given that different classifications are used in the studies, a meta-analysis study on SAUP would be incomplete (Table 4) (Figure 3).

NUP is an important bone structure associated with the lateral nasal wall. It is the first anatomical structure encountered in the anterior section of the nasal cavity. It borders Hiatus semilunaris (HS) from the anterior together with the adjacent anterior ethmoid cells. The maxillary sinuses open to the posterior side of the infundibulum via the ostium [1,18,19].

There are many variations in the NUP. Less importance is attributed to pneumatisation and hypoplasia among these variations. However, pneumatisation is a relatively common one (2–14%), and when extensive, it can obstruct the anterior entrance of the nasal passage. Deviations of the NUP and SAUP are one of the factors that primarily affect the HS and PNS ostium outlets. Physiopathologically, they directly affect sinus ventilation by compressing the infundibulum [1,20,21].

The relationship of the SAUP type with the HS determines the anatomical outlet point of the frontal recess. NUP may cause pathologies by closing the ostium externally. In such cases, it is important to specify the SAUP type in the radiological reporting [22]. Frontal sinuses consist of bony anterior and posterior walls. Depending on the type of SAUP, they drain into the middle meatus or ethmoid infundibulum medially and posteriorly via the frontal recess. If NUP attaches to the lamina papyracea, the frontal sinus drains to the middle meatus via HS. If NUP attaches to the skull base or the middle turbinate, the frontal sinus drains into the ethmoid infundibulum before draining into the middle meatus [23]. In other words, the SAUP type directly affects the frontal sinus drainage path.

In addition, NUP shapes and SAUP types are associated with facial pain, rhinorrhoea, and nasal obstruction. The lateral contact of the NUP with the nasal septum or walls while reaching the roof is the main cause of facial pain. It is also involved in the aetiology of sinusitis by obstructing sinus recess and ostium. This triggers rhinorrhoea and pain [24]. Apart from the NUPs, it has been associated with many pathologies in the literature. It even has a place among the anatomic variations associated with the antrochoanal polyp [25].

NUP is closely related to the maxillary sinus ostium, as it is to the frontal sinus recess. Mostly, uncinectomy is the first step in ESS [24]. Uncinectomy is necessary for the maxillary sinus outlet during FESS [1]. There are two common types of uncinectomy techniques, known as antegrade and retrograde techniques. The basic surgical principles of both techniques are based on removing the UP to avoid a blind operation on the maxillary sinus and ostium [26]. Sickle knives or freer elevators are used in the anterograde method, while small backbiting forceps are used in retrograde uncinectomy. Aside from these techniques, other methods can be used, such as using microdebriders or other types of equipment. Knowing the SAUP type is essential, especially during the use of forceps. In the case in which the NUP adheres to the base of the skull (SH II or LF 3, 4, or 5), the base of the cranial fossa may become damaged while the uncinate process is being pulled with forceps [27]. However, it would be unfair to limit the importance of the SAUP type only to uncinectomy. Knowing the SAUP type is essential not only in uncinectomy but also in techniques for preserving the NUP for determining the maxillary ostium [28].

The orbital-related SAUP variants are important in orbital decompression operations. NUP excision is also essential in this operation technique. In LF Type 1 and 2, or SH Type I variants, it should be carefully considered against the risk of complications in the orbital interventions from inferior [29,30].

In summary, concha-related variants are the cause of facial pain. The types that adhere to the skull base are important in uncinectomy operations, and the types related to the orbit are important in orbital decompression operations.

The SAUP types are important in clinical and surgical management. Concha-related variants are associated with chronic facial pain. It is essential to know the types related to the skull base before uncinectomy and the types adhered to the orbit before decompression operations to avoid possible complications. NUP can fuse with some anatomic structures such as the ethmoid cell and Agger nasi cell until it reaches the insertion point. This situation can cause misconceptions by making follow-up difficult. Following the thick band and paying attention to concavity and convexity helps to accurately evaluate the anatomical structure.

SH and LF are the guides used in SAUP classifications. These two methods showed a statistically high significant correlation with each other. However, the use of different classifications in research prevents meta-analysis and causes confusion. SH Type IV/LF Type 0 and SH Type III/LF Type 6 completely overlapped. SH Type I and Type II, respectively, corresponded to LF Type 2/3, and SH Type II to LF Type 2/3/4/5 groups. There was no significant difference between males and females or between right and left, according to the LF or SH types.