The inuence of gingival phenotype in the morphology of the central maxillary interdental papilla

Preservation of interdental papilla is an essential part of the functional and esthetic rehabilitation of dental treatment. It was described that thicker gingival tissues are more resistant to recession. The main objective of this investigation was to analyze if the gingival phenotype represents a potential risk factor affecting interdental papilla ll, height, or width in an esthetic region between central maxillary incisors. The secondary goals were: 1) analyze parameters describing papilla and the inuence of a wide papilla basis for the vertical papillary dimension; 2) the correlation between different non-invasive measurements of gingival thickness; 3) comparison among both sexes. Methods was probe - and and the degree of was


Introduction
The patient's demands on dental treatment are often very high. Patients do not seek only for functional rehabilitation, but also for the natural esthetic result. A balanced size, shape, position, and color of teeth are essential components of successful esthetic outcome and should be in harmony with surrounding soft tissues [1]. However, in some situations, speci c problems with so-called "pink esthetic" occur, comprising the mucogingival conditions, such as excessive gingival display, uneven gingival contours, exposure of root surfaces or the loss of interdental papillae. In such cases, it is also essential to analyze the patient's smile line [2].
In exposed regions, the interdental papilla plays a vital role in the nal esthetic outcome, especially if a high smile line is present. Physiologically, the interdental papilla in the anterior region has a pyramidal shape and lls the entire space under the contact point between two adjacent teeth. If a papilla does not ll the whole interdental space, a black triangle occurs. It is considered as an esthetic impairment, and it can also cause phonetic problems or food retention, which can adversely affect periodontal health [3]. Thus, clinicians should be able to adequately analyze the factors related to the interproximal papilla to prevent its loss.
Various situations can in uence the morphology of interdental papillae, mainly periodontal attachment loss resulting in the recession and impairment of the volume of the alveolar bone relative to the interproximal contact [3]. Tarnow et al. observed that critical distance from the contact point/area to the alveolar bone crest is 5 mm [4]. Other investigations also revealed a signi cant correlation between increasing interdental distance and papilla recession [5][6][7][8][9] Soft tissue thickness in relation to interdental papillae was investigated in very few studies. There is an assumption that thicker gingival tissues are more resistant to physical trauma and have a lower risk of recession due to the better blood supply and adequate amount of dense brous tissue [10].
Some authors found that the thick periodontal phenotype was observed with signi cantly greater papillary ll [11,12], but with decreased papillary height [13,14]. On the other hand, there are also opposite results published, where the thin periodontal phenotype presented a signi cantly higher presence of entire papilla ll [15]. Only the recent study found papilla width as an independent in uencing factor of periodontal biotype [16]. Many other investigations revealed no statistically signi cant correlation between the gingival phenotype and the morphology of interdental papillae [1,17,18].
To date, it is not clear whether the gingival phenotype represents a signi cant factor associated with the morphology of interdental papilla and if there are any differences among both sexes. The main objective of this investigation was to analyze if the gingival phenotype represents a potential risk factor affecting interdental papilla ll, height, or width in an esthetic region between central maxillary incisors. The secondary goals were: 1) analyze parameters describing papilla and the in uence of a wide papilla basis for the vertical papillary dimension; 2) the correlation between different non-invasive measurements of gingival thickness; 3) comparison among both sexes. We decided to include only central maxillary incisors as reference area because differences between phenotypes are the most explicit for these teeth and because this region is the most exposed part of the dental arch and act as the most challenging area in the way of esthetic [19][20][21].

Participants
All clinical measurements were performed between April 2017 to August 2017 in the Department of Periodontology and Oral Medicine at the Institute of Dentistry and Oral Sciences in Olomouc, Czech Republic. Together 57 undergraduate students of dentistry (32 females, 22 males, in the range 20-30 years of age) were enrolled in this study. All participants were thoroughly educated in the eld of oral hygiene. All subjects were required to have state of Gingival health on intact periodontium according to new classi cation scheme [22]: no clinical attachment level loss, probing pocket depth (assuming no pseudo pockets) ≤3 mm, no bleeding on probing at examined sites [23]. Exclusion criteria were following: 1. medication intake or suffering from any disorder classi ed at Systemic diseases and conditions that affect the periodontal supporting tissues [24]; 2. pregnant or lactating females; 3. severe smokers (10 and more cigarettes per day); 4. lack of keratinized tissue width (≤ 2 mm) in the region of upper central.
Three subjects were excluded because of insu cient data obtained due to absence on the second appointment.

Data collection
All measurements were performed by one experienced and previously calibrated examiner (Š.B.). Intraexaminer reproducibility was achieved by reassessing 20 random subjects to nd the accuracy between repeated measures.
First, the gingival thickness was measured using a non-invasive Pirop Ultrasonic Biometer (Echo-Son, Krancowa, Poland) with the A-scan probe (tip diameter 1.7 mm) with 20 MHz frequency and 1540 m/s ultrasonic impulse velocity and accuracy up to 0.01 mm. A chlorhexidine gel was applied on the tip of the probe, which was gently applied on the reference point on the intersection between the mid-facial longitudinal axis of the left central upper incisor and the horizontal axis of the keratinized mucosa at the midpoint of mucogingival and para-marginal grove ( Figure 1). Each assessment was based on ten automatic measurements. These were averaged and displayed on the screen of the device. The standard deviation of the mean value from ten automatic measurements did not exceed 0.05 mm.
On the next appointment scheduled a month later, intraoral photography of the anterior maxillary region was taken in a standardized manner -uni ed shooting conditions and camera setting parameters. A pressure-sensitive periodontal probe, with a controlled (∼0.25 N) force to the apical end (Carl Martin 973/SP, Solingen, Germany), was placed in the center of the facial aspect of gingival sulcus of left central maxillary incisor to assess the gingival probe transparency [25]. If the periodontal probe was visible through gingival sulcus, the phenotype was categorized as a thin. If the periodontal probe was not visible, the phenotype was assessed as a thick (Figure 2,3).
Papilla ll between central maxillary incisors was assessed from photographs using classi cation proposed by Nordland and Tarnow [26]. The height and width of interdental papilla between central maxillary incisors were measured from the intraoral photography. Image calibration was done using markers from a periodontal probe placed parallelly in the gingival sulcus of the tooth 21, in Planmeca Romexis dental imaging software (Planmeca, Helsinki, Finland). Papilla height was measured as a distance from the tip of the papilla to the connecting line of the gingival zeniths of central maxillary incisors (Figure 1). The width of the interdental papilla was determined as the distance between the intersections of the longitudinal axis, which divides the central incisors into a mesial and a middle third and marginal gingiva ( Figure 1).
Statistical analysis IBM SPSS Statistics version 22 software was used to analyze the data. Quantitative variables were reported as mean ± standard deviation (SD); categorical variables were reported as absolute and relative frequencies. The correlation between quantitative data was assessed using Spearman's correlation analysis. The relationship between enumeration data was evaluated using the Chi-square test. The correlation between quantitative and categorical variables was assessed using the Mann-Whitney U test. The remeasurements of papilla height and gingival thickness were controlled by the Dahlberg formula and the Intraclass correlation coe cient (ICC). Cohen's kappa coe cient was used to measure intra-rater agreement for the transparency and papilla ll parameters. The normality of the data was veri ed using the Shapiro-Wilk test. A signi cance level of less than 0.05 was considered statistically signi cant.

Results
The Cohen kappa value showed a substantial agreement (0.61) between original and control measurements of gingival probe transparency. ICC for the gingival thickness parameter (0.933) also showed an excellent match, but the Dahlberg error rate of variation was higher than 5%. The ICC coe cient for papilla height (0.985) and papilla width (0.935) revealed an excellent match, and also low Dahlberg error rate indicates a very good match of both measurements. Papilla ll showed absolute agreement in both measures. There was no statistically signi cant systematic shift between measurements.
The age of the participants was 26 ± 1.5 years. Table 1 lists distribution, the mean values, and standard deviation of the clinical data of 54 participants included in this study.
Papilla characteristics in different gingival phenotype based on gingival probe transparency are shown in Table 2. In the thick phenotype group, papilla recession was seen in 34.3% of cases, while in the thin phenotype group, it was 60%. However, a statistically signi cant correlation was not con rmed, also for papilla height or width. Table 3 shows the correlation of gingival thickness with papilla characteristics. No statistically signi cant relationship between papilla height, width, or ll was found.
The relationships among papilla characteristics are shown in Table 4 and Table 5. There was a signi cant correlation between papilla ll and papilla height (P = 0.028). Papillae classi ed as normal, lling entire interdental space seem to be shorter than papillae from class 1. No signi cant relationship between papilla ll and width was found. Spearman's correlation analysis revealed a strong positive correlation between papilla height and papilla width (r = 0.738, P <0.0001).
The relationship between different assessments of the gingival phenotype (Table 6) was statistically signi cant (P < 0.001). Mean gingival thickness for the group assessed as the thick was greater compared to the thin phenotype group.
Differences in measured parameters among gender were statistically signi cant only in papilla height (P = 0.01). Mean papilla height was greater in the male group compared to the female group. Other parameters revealed no statistical difference (Table 7).

Discussion
Preservation of interdental papilla is an essential part of the functional and esthetic rehabilitation of dental treatment. It has been described that the morphology of interdental papilla is strongly related to bone volume in the interproximal space [4][5][6][7][8][9]. In addition to the recession of interdental papilla related to periodontal disease, recession can also occur in healthy gingiva due to anatomical and physiological predispositions [1]. Therefore, the gingival phenotype and characteristics of interdental papilla were examined only in periodontally healthy patients. Also, contact points may vary in different regions, which may in uence the shape of interdental papilla. For this reason, we studied the only papilla between central maxillary incisors to have a homogeneous sample.
Two different non-invasive methods of phenotype assessment were performed, the most commonly used in similar studies-gingival probe transparency [14,[17][18][19]27,28] and also less standard method-ultrasonic measurement [29][30][31] of an accurate thickness of gingival tissues. A comparison of both methods showed a signi cant correlation. Thus, such a straightforward way of phenotype assessment by gingival probe transparency is as reliable as different methods, which are often more time-consuming or require some additional costs for appliances.
Despite the fact that the papilla recession was present in 34,3% cases with thick phenotype and in 60% cases with thin phenotype, there was no statistically signi cant correlation between interdental papilla of central maxillary incisors and gingival phenotype. This result supports previously published studies by Kim et al. [1] and Singh et al. [17]. Some authors assume that a thick phenotype is more resistant to physical trauma and has a lower risk of papilla recession due to the better blood supply and adequate amount of dense brous tissue [10]. Thick phenotype is also associated more with square-shaped tooth crowns with contact point located more apically and requires less tissue to ll the interproximal space [20,21]. This assumption con rmed Chow et al. [12], who observed that gingival tissues were signi cantly thicker when the papilla was competent. Opposite results published De Lemos et al. [15], who noted a signi cantly higher presence of papillae in the thin phenotype group. However, in this study, the phenotype was assessed only visually, what may have introduced unnecessary method error by subjective opinion. Most of the other authors studied the correlation between phenotype and papilla height as the only papilla descriptive parameter. Results found that increased papillary height is associated with a thin phenotype [13,14,27,28], what may be in uenced by different tooth shape [13,19,21,28,32]. As the tooth becomes triangular, what is more typical for thin periodontal phenotype subjects, the contact point can be seen more coronally, and longer papillae appear. This study failed to nd an in uence on papilla width with different gingival phenotype, in both types of phenotype assessment. Yin et al. [16] recently published that papilla width has a signi cant in uence on phenotype, making the gingival papilla of the maxillary central incisor of the thin biotype narrower. They assessed papilla width as the distance between the gingival zeniths of the two adjacent teeth. The incongruity in measurement methods of papilla width may be the major reason for different results. However, there are few studies on the correlation between the phenotype and papilla width, and more research needs to be provided.
We have also compared papilla characteristics -papilla ll, height and width between each other. Results showed that papillae assessed as normal, which ll whole interdental space seem to be shorter than papillae from class 1, where a slight reduction of papilla ll is present. Papilla height has previously been found to be signi cantly greater in the group, where papilla was present in the study by Chang et al. [33]. However, this result was not con rmed in the study by Kim et al. [34]. Both of them measured papilla height on radiographs using radiopaque material as the distance from the crest of the bone to the tip of the papilla. It can be speculated if different measurement method of papilla height may be the major factor contributing to the discrepancy in these results, or as reported by Chow et al. [12], who con rmed that papilla height decreased 0.012mm with each year of increasing age, there may play some role the enrolment of participants from different age groups. In our study, there was no signi cant relationship between papilla ll and papilla width, but a signi cant correlation was found between papilla height and width. It seems that there is a positive in uence of a wide papilla basis for its vertical dimension.
Mean papilla height was greater in the male group compared to the female group. Chow et al. have reported the same results [12], what is in contrast with the study by Joshi et al. [14]. Many other authors observed thin phenotype more frequently in females [14,19,35]; however, in this study, no correlation was found. We assume that the greater height of interproximal papillae found in the male group was due to different tooth forms and position of the contact point, which could be the reason for the only difference among gender in this study.
The small sample size without any groups with different age limits the assessment of gingival phenotype and its correlation with papilla characteristics. Therefore, in future studies, it is recommended to expand the sample size. Also, it is advisable to evaluate other potential risk factors, such as the tooth form or tooth angulation, which seem to be one of the signi cant factors in uencing interdental papillae by the different shape and position of the contact point. Finally, another potential factor, buccolingual tooth position, which may affect the gingival phenotype and thickness of alveolar bone, should be added in future studies to provide more convincing evidence.

Conclusion
The appearance of interdental papillae could be in uenced by various factors. Within the limitations of this study, results showed no signi cant correlation between the morphology of interdental papillae and soft tissue thickness. It seems that there is an in uence of a wide papilla basis for its vertical dimension. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
All risks and bene ts of the associated procedures were explained to all participants who signed informed consent before enrollment into the study. The study was approved by the Ethical Committee of Palacky University and University Hospital in Olomouc with reference number 29/17 and was conducted following the Helsinki Declaration.
The authors would like to thank professor Ricardo Faria Almeida for assistance with the preparation of this paper.
ŠB and MS participated in conception and design of the work and data collection. Statistical analysis was done by JZ. MŠ participated in collecting data. The original draft was written by ŠB and RŽ. Review and editing of the draft did JŠ. All authors made a substantial contribution to this study and/or manuscript, and all have reviewed the nal paper prior to its submission. The authors read and approved the nal manuscript.
No potential con ict of interest relevant to this study was reported.
This study was self-funded by the authors and their institutions. Data are reported as N (%) out of total N=54 (100 %) or as mean ± standard deviation.  Data are reported as Spearman's coe cient of rank correlation (r) Data are reported as mean ± standard deviation.