Netter's classic drawing of the human nasal septum. We have mapped onto this illustration the locations of the nasopalatine duct (NPD) and vomeronasal organ as described in the literature. Area X comprises the vomeronasal organ area as described by 4 different groups38,41-43 and the location where we located the NPD. Copyright 1989. Novartis. Reprinted with permission from the Atlas of Human Anatomy,2 illustrated by Frank H. Netter, MD. All rights reserved.
Potiquet's 1891 anatomical illustration3 shows the openings of both the nasopalatine duct and vomeronasal organ. The original captions (translated from the French) are as follows: 1, pad made of Jacobson's cartilage; 2, opening of Jacobson's canal, into which a probe has been inserted; 3, septal tubercle; 4, nasopalatine infundibulum (funnel) leading to the nasopalatine canal through the bone; 5, opening of the sphenoidal sinus; and 6, frontal sinus.
Nasopalatine duct fossa shapes. A and B, The distribution of the nasopalatine duct's 4 fossa shapes is based on their eccentricity (N=207 fossae). C and D, The photographs provide exemplars of the 4 fossa shapes. Note the vessel entering the round fossa; this photograph also shows how the endoscope's angle may prevent visualization of the aperture. The photograph of the oval fossa shows an aperture and an air bubble on the floor of the fossa.
Gross anatomy of the nasopalatine duct (NPD) and vomeronasal organ (VNO), depicted by an excised human nasal septum with a closeup of the region of interest. Gutta-percha points inserted into the fossae of the NPD and VNO mark the location of the structures on the nasal septum.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Jacob S, Zelano B, Gungor A, Abbott D, Naclerio R, McClintock MK. Location and Gross Morphology of the Nasopalatine Duct in Human Adults. Arch Otolaryngol Head Neck Surg. 2000;126(6):741–748. doi:10.1001/archotol.126.6.741
To characterize the nasal opening of the human nasopalatine duct (NPD; a major duct of the vomeronasal system used in animals for chemical communication) by identifying its location, categorizing variations in physical characteristics, and assessing frequency of detection.
Two studies: (1) general study incorporating endoscopic examinations documented by videotapes and photographs, and an investigation of detection bias in terms of method of visualization and defining criteria for NPD identification; and (2) cadaver dissections examining NPD gross anatomy and proximity to the putative vomeronasal organ (VNO), the second major duct of the vomeronasal system.
Department of Otolaryngology, University of Chicago Hospitals, Chicago, Ill.
A total of 125 university community volunteers, with a mean age of 23 years.
(1) General study: NPD was located 1.9±0.02 cm (mean±SEM) dorsal to the columella nasi, and 0.2±0.01 cm above the nasal floor/septum junction, in both nostrils (90% bilateral), and was symmetrical in shape. An NPD was detected in 94% of 221 nostrils unobstructed in the region of interest. Fossa shapes were oval (57%), round (18%), spindle-shaped (18%), and tubular (7%). A small, round aperture was visualized in 30% of fossae. Among 3 observers, NPD detection frequency ranged from 94% to 40%, with the disparity due to inclusion of different defining characteristics. (2) Cadaver dissection study: bilateral nasal NPD fossae were found in every specimen probed to maximum approximate depth of 8 mm. No buccal pits associated with patent NPD were detected. Putative VNOs superior and just anterior to NPD fossae were detected in fewer than half the specimens.
The nasal opening of the NPD is bilateral and symmetrical, located at the base of the nasal septum. Unusually contradictory anatomical descriptions in the human putative VNO literature may be attributable to inexact descriptions or misidentification of structures. The function of NPD remains controversial.
IN MANY MAMMALS, the nasopalatine duct (NPD) passes through the incisive canal, from the incisive papillae in the mouth up to the lower floor of the nasal cavity, providing direct communication between the oral and nasal cavities. The NPD may serve to deliver pheromones and social chemosignals entering the oral cavity to the vomeronasal organ (VNO). Three general anatomical relations between the VNO and NPD are seen among mammals: (1) the VNO opens directly into the nasal cavity with the NPD close by, eg, rodents and bats; (2) the VNO opens directly into the NPD, eg, cats; and (3) the VNO opens directly into the oral cavity, eg, cattle. Together, these 2 ducts form the vomeronasal system, which also includes neural projections from the VNO to the vomeronasal bulb (accessory olfactory bulb). This vomeronasal system can play a role in tracking prey, sex recognition, and courtship.1 The NPD also may be involved in carrying food odors from the mouth to the sensory epithelium of the primary olfactory system. In humans, however, the function and even existence of the NPD as well as a complete vomeronasal system remain unascertained.
We wanted to systematically study the nasal opening of the human NPD to categorize its physical characteristics and frequency of detection, not only for the inherently interesting anatomical and evolutionary aspects of this structure, but also to promote an increased awareness of the structure, particularly in the otolaryngology literature where the NPD is rarely mentioned. We quantified the physical characteristics and gross location of the human NPD through 2 independent studies. In the first general study, we categorized the physical characteristics and frequency of the NPD's nasal fossae in vivo in young adults without nasal abnormalities. We also evaluated 3 types of observer bias as a source of discrepant reports of frequency of detection of ducts in the vomeronasal system. In the second study of cadaver dissections, we looked at the gross anatomy of the NPD and examined the nasal septum more closely for the VNO. We will use "NPD" and "VNO" throughout the text, but do not wish to imply function, particularly in reference to humans.
Interestingly, modern anatomy books, such as Netter's 1989 classic,2 typically illustrate the anterior portion of the human nasal septum as smooth mucosa that lacks any nasal ducts or structures, while older texts such as Potiquet's 1891 anatomical illustrations3 show the openings of both the NPD and VNO (Figure 1). Moreover, in a 1920 publication, Schaeffer4 took the NPD's nasal opening for granted, in his coherent description of the human NPD, and stated that each fossa, one per nostril, leads into the nasopalatine canal,
courses obliquely caudalward and with its fellow of the opposite fossa converges toward the nasal septum, descends almost vertically, and passes through the Y-shaped incisive foramen (anterior palatine canal) in the hard palate. . . . The canals end on the roof of the mouth at the side of the papilla palatina or incisive pad.
Controversy in the literature prevents forming rigorous conclusions concerning NPD patency in human embryos and neonates.5-7 Farmer and Lowton5 believed that the ducts are patent in the fetus and close just before birth. Bellairs,6 however, believed that the ducts are represented by solid cords that may canalize or regress in places to give rise to epithelial pearls and that complete patency on one or both sides may occur but is rare in embryonic, neonatal, and postnatal individuals. Noyes7 examined serial sections of the lower third of the nasal cavity and premaxilla of newborn infants and found NPDs in all specimens, but no continuous passage between the oral and nasal cavities.
In the human dental literature,8-35 there are a number of case reports and less systematic studies of palatal fossae just behind the papilla palatina associated with a patent NPD (Table 1). The NPD is often implicated in such pathological case reports as the nasopalatine canal is a common site for inflammation, infection, tumors, and cysts.36 Patent NPDs are usually discovered by complaints of troublesome symptoms such as squeaky noises, debris collection, discharge from the roof of the mouth, passage of food from the mouth through the nose, and pain or swelling in the premaxilla area associated with the condition. The patency can present itself as a bilateral, unilateral, or central opening on both sides or one side of the papilla palatina, or as central on the midline posterior to the incisive papilla respectively. Such nasopalatine patency is regarded as atypical but an estimate of the true frequency of patency cannot be made without a systematic, in vivo study in human adults. Therefore, we use the literature's term "NPD" without assuming that it is either patent or blind.
Controversy exists in the human literature not only on the NPD, but also on the VNO. Studies of a putative human VNO, the second duct of the vomeronasal system, vary in reported frequency of occurrence, ranging from 39%37 to 100%.38 Our experienced otolaryngologists, however, previously had not appreciated either the VNO or the NPD. Moreover, following each of the contradicting anatomical descriptions in the VNO literature (Table 2), we had difficulty locating it during preliminary nasal explorations. In fact, the contradictory literature made unequivocal identification of the VNO virtually impossible. Much more easily, we detected the nasal fossa of the NPD and discovered that, despite the fact that the 2 structures lie within centimeters of each other, there is very little mention of the NPD's nasal opening in VNO or current general otolaryngology literature. Therefore, we chose to focus on the NPD because (1) it is an easily identifiable structure, which potentially could be useful as an anatomical landmark, and (2) it is rarely mentioned in the current literature.
The clinical implications of the existence of the human NPD, whether patent or blind, make it an important subject for study. Although the putative vomeronasal system in humans is gaining interest in the scientific community, the NPD has remained relatively ignored.35 Evidence suggests that the NPD in human adults is rarely patent and likely exists as a vestigial remnant. Vestigial remnants are often the sites of pathologic conditions—the human appendix is an example—and therefore are important to clinical practice. In an evolutionary context, the existence of a human NPD is interesting because the ducts serve a functional role in the vomeronasal system of other mammals, including some primate species.44,45
Our participants were 125 healthy volunteers from a university community, with a mean±SEM age of 23±0.4 years (range, 18-35 years). Detailed examination interviews found that 3 of the participants had undergone prior endoscopic sinus surgery procedures that did not affect our regions of interest.
Because the focus of McClintock and colleagues' general research was the effects of ovarian hormones on chemical senses, we had a larger sample of women (n=80) than men (n=45). There was ethnic diversity in our participant group (58% were white, 22% were East and South Asian; 12% were black; 2% were Hispanic, and 6% were mixed or other), which differs from some previous studies of the VNO that may have focused on more homogeneous populations.41,42 All participants provided written informed consent that was approved by the institutional review board.
Nasal endoscopy was performed by otolaryngologists (A.G., D.A., or R.N.) and took about 20 minutes of the participant's time. First, the participant's nose was sprayed with 1% xylometazoline hydrochloride, a nasal decongestant. Local anesthesia was obtained with 2% lidocaine spray.
The otolaryngologist examined each nostril of the 125 participants using a 30° Storz endoscope. The otolaryngologist focused the examination on the base regions of the anterior one third of the septum and noted whether septal deviations obstructed the view of the NPD region. Images of the intranasal examinations were recorded on videotape and at least 2 endoscopic photographs were taken of the region of interest in each nostril.
The photographs were examined to validate the clinical diagnosis of the presence or lack of fossae at the nasal end of the NPDs. Based on the endoscopic examination, photographs, and videotape, an anatomical description sheet was filled out describing the location, visible characteristics, and any other physical observations of the fossae.
Because some participants had obstructions precluding NPD visualization, nostrils were examined independently, providing a total of 221 nostrils unobstructed in our region of interest. Participants were placed into 3 categories based on having bilateral, unilateral, or no visible NPD.
We found that it was necessary to differentiate between 2 distinct morphologic features: (1) the structure's typically larger fossa or craterlike indentation and (2) its smaller aperture (within the fossa, a circular opening found off center, on the fossa floor). The NPD was coded as detectable only if it was in a specific location (approximately 2 cm dorsal to the nostril opening and <0.5 cm above the junction of the nasal floor and septum) and at least 1 of the 2 definitive features was identifiable: a fossa or an aperture. The characteristics of the fossa including its shape, depth, or the paleness of its floor in relation to the surrounding mucosa were defined and their frequencies were determined.
To analyze the bilaterality of NPD detection, we included only the 100 participants with no obstructions in either nostril that would preclude visualization of the NPD. Of these, 90 participants had bilaterally detected structures and we compared the shapes of fossae in both nostrils to assess if the NPD's shape is also symmetrical. Fossae were coded as bilateral if they were or were not visible in both nostrils, and the shapes of fossae were coded as symmetrical if they were the same in both nostrils (ie, both oval).
To assess potential sources of observer bias and confusion in NPD detection, we had 3 types of viewers: a "neutral" otolaryngologist, a research scientist, and a "critical" otolaryngologist. These types were chosen to distinguish between detection criteria and the limitations of viewing a 2-dimensional videotape as sources of observer bias. Before the study, the examining otolaryngologist was neutral on the possibility of a human NPD and was not familiar with the animal or dental literature on the NPD and vomeronasal system. We had a critical otolaryngologist view the videotapes of the examinations independently of the neutral physician who performed the examinations. The critical otolaryngologist was skeptical of the NPD's existence because he had never noted an NPD during his decades of surgical experience with the nasal cavity, and was unfamiliar with the dental or animal literature. He did not perform the endoscopies and was unaware of our results when he coded the videotapes using the same anatomical description data sheet that the examining otolaryngologist had used. For a third perspective, a research scientist, who did know the animal and dental literature, coded the videotapes using the same type of data description sheet and did so unaware of the other viewers' categorizations.
In the second part of our study, we examined the nasal septa of 8 human cadavers (all male) that had been kept frozen and then thawed. We removed the nasal septum in its entirety by a superior transseptal cut and by parallel sagittal cuts through the hard and soft palates using a Gigli saw. Each section thus comprised the nasal opening of the NPD and the portion of the premaxilla encompassing nasopalatine pathology as described in the dental literature. With the cadaver specimens we were able to easily probe the NPD from the nasal opening with a gutta-percha point. We chose this classic anatomical probe because it was useful for visualizing the proximity and orientation of the 2 organs on the nasal septum and has been used in contemporary anatomical studies.31,34 Septal specimens were immediately examined for NPDs and VNOs. Nasopalatine ducts were probed with gutta-percha points for gross depth, patency, and direction of the canal. Vomeronasal organs were also probed with gutta-percha points. Nasopalatine duct fossae shapes were recorded and the specimens were photographed and fixed in 4% formaldehyde. After fixation, the upper palates in the area of the papilla palatina were examined for obvious nasopalatine abnormalities.
For an illustrative model of the anatomy of the medial nasal wall, the critical otolaryngologist mapped out our findings of NPD location on Netter's2 classic drawing along with locations of published text descriptions of the putative human VNO (Figure 2 and Table 2).
The nasal opening of the NPD was consistently located 1.9 cm (SEM, ±0.02) dorsal to the columella nasi (anterior part of the septum), 0.2 cm (SEM, ±0.01) above the junction of the nasal floor and septum, where Potiquet3 had depicted it in 1891 (see Figure 1 and area X in Figure 2). The nasal opening of the NPD presented an easily detectable fossa in this location. Our reported measurements of size are estimates rather than true measurements and were based on other anatomical landmarks. Most fossae were approximately 2 mm in diameter but ranged from 1 mm to 20 mm, with an average diameter of 3.6 mm (SEM, ±0.20).
We detected an NPD in 94% of 221 nostrils unobstructed in our region of interest; 52% of the NPD were in the left nostril and 48% were in the right. The fossa was the most obvious feature of the NPD and could be seen easily even from an oblique angle (from 30° to less than 90°). The aperture of the NPD, however, was harder to find. When a view normal to the fossa floor was achieved with the endoscope, the small, round aperture was seen off center on the fossa's floor in 30% of the NPD. Septal deformities and the confines of our participants' nostrils often prevented attaining a view normal to the fossa floor. In addition to inadequate viewing angles, nasal secretions or air bubbles occasionally added to the difficulty in detecting the aperture.
In 70% of the cases where the fossa's floor was viewable, the floor appeared paler than the surrounding mucosa. This, however, might be affected by the endoscope's lighting or might be associated with depth rather than evidence of a distinct morphologic characteristic.
Most individual variation of the NPD lay in the shape of the fossa. Fossa shape descriptions are based on eccentricity from round to tubular (Figure 3). The oval fossa was the most common shape (57%), but the less common round (18%) and spindle-shaped (18%) fossae were also detected. More rarely, we detected tubular (7%) fossae in the appropriate region and this shape may be a consequence of minor septal abnormalities. Apertures had more uniform features because they were always small (approximately ≤1 mm in diameter) and circular.
Of the 100 participants with no nasal obstructions, 90% had bilateral, 9% had unilateral, and 1% had no visible NPD. Frequency of detection was similar in both sexes. Of the 61 women, 90% had bilateral NPD and 10% had unilateral NPD. Of the 39 men, 90% had bilateral NPD, 8% had unilateral NPD, and 2% had no visible NPD. There were neither sex-specific anatomical differences in symmetry (χ2=1.69; P=.43) nor obvious race-specific differences in symmetry (χ2=2.94; P=.82).
We found that 69% of the participants with bilaterally detectable structures (n=90) had fossae of similar shape in both nostrils (χ2=12.32; P<.001). A person was more likely to have 2 typical fossae (oval) or 2 atypical fossae (circle, ellipse, groove) rather than one typical and the other atypical.
The otolaryngologist examiner and the research scientist detected the NPD at similar frequencies, 93% and 94%, respectively. The critical otolaryngologist, however, reported that only close to 40% of the septal regions contained an NPD. Using the same anatomical description sheet as the other observers, his lower detection rate resulted from the use of more stringent criteria: the presence of the small aperture as well as the fossa as a definitive feature and the exclusion of atypical variants in shape—the elliptical or groove-shaped fossae.
We examined the excised nasal septa from 8 cadavers and found bilateral nasopalatine fossae with apertures in every specimen (area X in Figure 2). The NPDs were found in the same location as in our general study. Cadaver specimens with fossae of the tubular shape showed that this shape was likely a result of septal deviations. Without the restriction of external nasal structures, the fossae could be viewed from all angles and tubular fossae sometimes opened posteriorly into more typically shaped fossae. In addition, deviations that could hide the fossae in vivo were easily manipulated in the specimens to provide better views. Twenty-five percent of the NPD fossae were round, 37% were oval, 19% were spindle shaped, and 19% were tubular.
We could easily insert gutta-percha points into the apertures to a minimum depth of 2 mm and a maximum of 8 mm. The width and degree of pliancy of the gutta-percha points present obvious limitations in this preliminary exploration of depth and patency of the NPD, and underestimated the length of the duct. All canals angled obliquely forward with a downward turn close to the oral cavity in accordance with published anatomical descriptions.
We examined the premaxilla area of the fixed specimens for pits or slits associated with nasopalatine pathology as described in literature. Without dissecting tissue or histological examination, we did not find any obvious bilateral, unilateral, or central oral openings indicative of patent NPD.
We found VNO in less than half of the 16 septal regions studied. We probed the VNO with gutta-percha points to show its location on the nasal septum in relation to the NPD (Figure 4, area A in Figure 2).
In our systematic characterization of the human NPD's nasal opening, we found that it is an easily detectable bilateral and symmetrical structure consistently located near the juncture of the nasal septum and the floor of the nasal cavity. Although oral openings of human NPD have been reported rarely and usually in the context of pathology, the nasal opening is common and relatively uniform in its morphologic characteristics.
Comparing the location of the NPD with the published locations of the putative VNO revealed that four38,41-43 of the 7 articles located the putative VNO precisely where we have found the NPD fossae (area X in Figure 2). Area C in Figure 2 correctly locates the VNO anterior to the NPD but places it near the nasal floor, whereas area B in Figure 2 vaguely locates it on the septum posterior to the location Potiquet described and where we observed it in cadaver specimens (compare Figure 2 area B with Figure 1 and Figure 4). Further investigation could determine if inexact anatomical descriptions are to blame or if some investigators mistakenly identified the NPD as a putative VNO. Future descriptions of nasal and septal structures such as the VNO would benefit from using stable anatomical landmarks. Because of its common occurrence and consistent location, the nasal fossa of the NPD is a good candidate.
We showed that our observer bias in detecting the NPD did not depend on the method of visualization but rather on using the more stringent criteria of visualizing the aperture within the NPD fossa. Such differences in coding criteria could be a source of the different VNO detection frequencies in the current literature.
The cadaver study allowed us to make a more thorough investigation of the NPD than with the general study alone. We were able to easily insert the gutta-percha probes into the fossae to a maximum of 8 mm. We did not examine the depth of the NPD of the participants in our general study, and in the cadaver study we were unable to detect buccal fossae associated with patent NPD. Probing the nasal opening of the NPD in a large population is the best way to determine the percentage of fossae that have nasal apertures to a patent NPD or are just blind fossae with no apertures. In addition, systematic investigations focusing on patency to the oral cavity would add to our knowledge of NPD and general nasal anatomy. Perhaps patency exists on a more microscopic or molecular level than can be detected with large, typical probes.
We were unable to find a VNO in all of the specimens we examined and do not know whether this reflects true frequencies or if, perhaps, the freezing process somehow altered the appearance of some of the ducts in such a way as to make their detection more difficult. Whether functional or not, we were able to detect VNOs and NPDs. Although it may be that a functional vomeronasal system does not exist in humans,46 our data confirm that anatomical remnants of this system certainly do.
The NPD could have more purpose than as a site of pathologic conditions. We noticed air bubbles and mucus trapped in NPD fossae, suggesting a mechanism by which chemicals could pool in the area for increase absorption into the bloodstream. Fossae of different shapes and sizes could differ in pooling effectiveness. Mucus retention in the human NPD is interesting given that the elephant's NPD has recently been shown to capture chemosignals via mucociliary protein carriers en route to the VNO.47 We made no investigations of pooling as a functional mechanism for absorption of chemosignals, unassociated with a neural vomeronasal system, and suggest it primarily to pique curiosity and stimulate further thinking. Since so little is known about the NPD in humans, there remain many questions to explore.
Accepted for publication January 12, 2000.
This work was supported by the Mind-Body Network of the John D. and Catherine T. MacArthur Foundation, and the National Institutes of Health MERIT award R37 MH41788 to Dr McClintock.
We thank Sheila Garcia and Davinder Hayreh for their valuable assistance in data collection and subject recruitment.
Corresponding author: Martha K. McClintock, PhD, Department of Psychology, 5730 S Woodlawn Ave, Chicago, IL 60637 (e-mail: email@example.com).