Novel methodologies and technologies to assess mid-palatal suture maturation: a systematic review

A technique to assess palatal suture maturation includes the use of multi-slice low-dose CT to capture axial slices of the maxilla and quantitatively measure the bone density at a particular ROI in HU [11]. It is known that CT is an excellent modality to evaluate the localized architecture of cancellous and cortical bone of the jaws; [17] however, less is known regarding the quantitative measurement of bone density, the HU scale. Hus were first utilized in dentistry to evaluate the pre-surgical bone density of implant sites [17‐19]. The HU scale is a linear transformation of tissue attenuation coefficients where air is defined as −1000 HU, distilled water at standardized conditions equal to 0 HU and very dense bone defined as ≥1000 HU [17]. Consequently the authors considered and utilized the calculated Hus as an applicable unit of measurement to quantitatively assess mineralization at the palatal suture [11].

Franchi et al. [11] utilized the Houndsfield quantitative scale to evaluate the radiodensity of four previously mentioned ROI in the maxilla, 2 sutural and 2 bony areas. Pre-expansion (T0) statistical analysis noted a significant difference between the anterior and postural sutural regions (563.3 ± 183.29 HU and 741.7 ± 167.1 HU) and anterior and posterior bony areas (1057.5± 129.4 HU and 1102.8 ± 160.9 HU) (P < 0.05) (Table 2). Further statistical analysis yielded a significant difference between the anterior sutural and posterior sutural landmarks at T0 (P < 0.05), but no significant differences of these sutural areas at T1 or T2 (P > 0.05, Mann-Whitney). A significant difference between the radiodensity of the anterior and postural sutural ROIs between T0 and immediately post-expansion (T1), but no difference between their radiodensities when comparing pre-treatment (T0) and the post-expansion retention phase (T2) readings (P < 0.05) (Table 2).

Throughout the course of the study, trends in bone density measurements at the suture and its comparison to lateral bony sites followed conventional expectations of successful RME. Pre-expansion the measured HU at the anterior sutural region was significantly smaller than that of the posterior sutural site, and the applied expansion protocol introduced differential sutural opening with greatest opening at the anterior sutural region consistent with the pre-expansion HU scores. Additionally, the results measured at T2 at the end of the 6 month RME retention protocol, were congruent with previous histologic findings, namely post-expansion evidence of reorganization and sutural interdigitation [20].

An inherent advantage of using a low-dose CT protocol, where the voltage was decreased to 80 kV (KV), is subjecting the patient to a lower absorbed dose required for children undergoing radiologic evaluation [21]. Additionally, when the kilovoltage is reduced, image contrast of anatomical structures increases while still acceptable for assessing bone quality via this protocol [21]. Future areas of interest relating to the findings and protocol of this study would include further studies to define an anterior sutural HU: postural sutural HU ratio that best predicts the success of RME treatment. Conversely, further studies could elucidate specific ratios comparing sutural radiodensity to maxillary bony radiodensity that may predict an improved expansion prognosis.

It has to be noted that the reliability of using HU between subjects and within the same subjects the same day has not been demonstrated. Therefore some variation could be due to such factors. Also, not all studies specified patient orientation when taking the images thus effect of patient positioning on the image and HU or grey values is another aspect that should be tested.

Korbmacher et al. [10] proposed assessing palatal suture maturation via micro-CT scanning and calculation of a number of developed indices, namely the obliteration index (%) and mean obliteration index (%) in the frontal plane, as well as, suture length [μm], linear sutural distance [μm] and interdigitation index in the axial plane.

Korbmacher et al. [10] evaluated 28 human palate specimens in the frontal and axial planes. In the frontal plane there was no demonstrated age dependent difference in the mean obliteration index between specimens (P = 0.244). The specimens were classified into one of three age groups (<25 years of age (yo), ≥25 to <30 yo and ≥30 yo) and results demonstrate that the frontal plane obliteration index varied across age groups between a minimum index of 0% to a maximum interdigitation of 7.3% (44 yo patient) (Table 2). Although the ≥25 to <30 yo age group consistently had a higher obliteration index in the frontal plane compared to other age groups, the results were not significant. Across all age groups, each subject had at least one frontal sutural cross-section that was devoid of interdigitation (mean obliteration index of 0%), with the oldest patient exhibiting a frontal plane mean obliteration index of 0% being a 71yo female. Investigation into the degree of interdigitation in the axial plane demonstrated no significant age-dependent differences in the calculated interdigitation index (P = 0.633). The authors did report a large standard deviation in the interdigitation index in the axial plane in the youngest and oldest age groups, and considerably less variation in the calculated index in the middle (<25 yo group and >30 yo) group [10] (Table 2).

Results indicated a generally low obliteration index amongst all subjects as well as an age-independent degree of interdigitation in the axial plane; however, across all measured indices there was significant intra-sutural and inter-subject variation [10]. This was the first time micro-CT was used on human samples and although this methodology was not implemented as part of an active expansion study, its principles can still be important to evaluate the pre-expansion maturity of the palatal suture. Additionally, it could be applied during mid-expansion protocol to evaluate the efficacy of treatment via calculation of the above noted indices and evaluation of the sutural architecture.

A limiting feature of the Korbmacher et al. [10] modality is the fact cadaver specimens were used, making direct translation of this study’s findings poorly applicable to clinical practice [15]. Considering the limitations of the gantry size of the micro-CT unit, and maximum scanning time used (200 min), micro-CT is best used on ex-vivo samples, and very small in-vivo samples to avoid an excessive absorbed dose emitted to patients [22]. Consequently, the use of micro-CT for in-vivo radiologic evaluation of the palate is impractical at this time. Therefore continued improvements to micro-CT technology including decreasing the emitted radiation while maintaining superior resolution, is necessary prior to implementation of such a technique on active RME patients.

An area of interest is the development of a CT-based strain assessment of peri-sutural and maxillary tissues; the development of which the authors believe will help facilitate predicting the success of RME treatment [10].

Sumer et al. [9] utilized US to evaluate sutural mineralization at five time points during the SARME and retention protocol for three patients, scoring each patient’s palatal suture calcification via assignment of semi-quantitative bone fill scores (0–3).

US findings in the Sumer et al. [9] study demonstrated that immediately post-expansion all subjects had a bone fill score = 0. (Table 2) Two of the three subjects at 2 and 4 months post-expansion were identified as having a bone score = 1, while the remaining subject was determined to have a bone fill score = 2 for these same time periods. Following the removal of the tooth-borne appliance at 6 months and 2 months subsequent to that during continued fixed appliance therapy, the bone scores for two of the subjects demonstrated increased mineralization and identification of echogenic material, having bone fill scores =2. The remaining patient received a bone fill score = 3 due to incomplete transmission of the waves and 100% echogenicity measured at these respective time points [9]. (Table 2) It should be noted that no statistics were reported by the authors.

The results of this study follow those of a similar animal study, [23] such that there was a statistically significant increase in bone fill scores that were directly related to the length of time the patient has been in retention post expansion. A major advantage to US is its low cost and non-invasiveness, [9, 23, 24] as well as improved usability compared to other methodologies, with the ability to perform real-time chair side evaluations with smaller hand held units. Additionally, US is a reliable method to image early bone formation as demonstrated by previous studies involving distraction osteogenesis [9, 23, 24]. A study comparing US to normal panoramic radiography, demonstrated that the efficacy of US to measure an osteotomy gap during distraction osteogenesis is equal to that of conventional radiography [9, 25]. US also demonstrated increased reliability compared to panoramic radiography to evaluate the maturation of early bone formation [9, 25] in the distraction gap. A disadvantage to US is its inability to penetrate cortical bone [9] However, following SARME or successful RME the osteotomy gap and its margins are easily visualized [9]. An area of significant future interest is to ascertain whether this technology can penetrate an immature midpalatal suture prior to the start of RME treatment, and allow the clinician to perform a chair side subjective evaluation of the bone maturity and interdigitation along the whole length of the suture. Limitations to this study included a very small sample size of three patients and lack of a gold standard (histology) or CT to validate the findings. Consequently, an area of future research is the use of this technology and bone fills scores in a similar larger sample size study in conjunction with a gold standard methodology to support the findings [9].

Angelieri et al. [3] utilized a standardized methodology to capture axial CBCT cross-sections of the palatal suture to provide individual staging of midpalatal suture maturation from the authors’ proposed maturation stages (A-E).

As it relates to Angelieri et al. [3] a validation study performed reported a weighted Kappa statistic for intra- and interexaminer reliability to be κ =0.75 (95% Confidence Interval (CI), 0.64–0.99) and be κ =0.79 (95% CI, 0.60–0.97) (no P-value reported), respectively. Due to a lack of an histologic or micro-CT gold standard, the authors also reported examiner reliability compared to the “ground truth”, a descriptor used to represent consensus among examiners with the principal investigator’ radiographic evaluations or other interpretations. Examiner reliability with ground truth ranged from κ = 0.82 (95% CI, 0.64–0.99) to κ =0.93 (95% CI, 0.86–1.00) (no P-value reported) [3].

Results of the validation study demonstrated “almost perfect” inter-examiner reliability with the “ground truth”, however, the authors did not report appropriate P-values with their statistics. As was mentioned before, there was no reference standard utilized during the validation study, but rather utilized what the authors termed the “ground truth”, [3] the professional opinion of the principal investigator when utilizing their own proposed maturation stages to classify each patient’s sutural maturation. Due to the lack of a gold standard, nor listed P-values, the results of the validation should be interpreted with caution. An additional limitation of this methodology is the proposed novel palatal suture maturation classification system itself. The authors developed the stages (A-E) based on comparison of CBCT axial cross-sections of the palatal suture to the perceived likeness of this radiographic morphology to the histological morphology of the suture as determined by previous studies [13‐15]. Theoretically direct comparison of the histological morphology to the CBCT morphology of the suture is incompatible due to the histological assessment being on the microscopic scale as compared to the macro or eye level scale of sutures depicted in the CBCT axial slices. Consequently, any inference or direct translation of the sutural histological appearance and subsequent development of CBCT based sutural maturation stages is not possible. Therefore, the findings and developed maturational stages should be used with caution, and should not drive clinical decision making. Rather, at best, this maturational staging may be used as part of an extended protocol to subjectively assess palatal suture maturity during the treatment planning process. Future studies to thoroughly validate the proposed maturation stages to an available gold standard are advised.

Kwak et al. [12] utilized CBCT imaging in conjunction with quantitative fractal analysis to ascertain if this analysis can be correlated to the maturational stage of each subjects palatal suture. Conceptually fractal analysis is based on the observation that cranial sutures can be visualized as a fractal pattern, [16] the dimensions of which are directly related to localized stresses experienced [12]. Additionally, the closer the approximation of two articulating bones, the more complex sutural morphology [12] suggestive of a more mature suture. Conceptually sound, fractal analysis has demonstrated its applicability in various areas dental research [26].

Fractal dimension intra- and inter-reliability results from the Kwak et al. [12] study demonstrated agreement with calculated weighted kappa coefficients of 0.84 (95% CI 0.74–0.93) and 0.67 (95% CI 0.38–0.95) to 0.72 (95% CI 0.48–0.97), respectively (Table 2). The CVM index inter- and intra-examiner reliability demonstrated agreement with weighted kappa coefficients from 0.69 (95% CI 0.53–0.86) and 0.71 (95% CI 0.56–0.86), respectively. The authors reported that none of the patients investigated possessed a CVM 1-IV nor was any subject classified as having palatal suture maturational stage A. It was found that 13 of 21 subjects with CVM V were classified as having maturational stage B or C (61.9%; males 77.8%, females 50.0%). Additionally, 42 of 110 subjects with CVM VI were classified as having maturational stage B or C (38.2%; males 41.6%, females 34.0%). Post-hoc analysis demonstrated that maturational stages B, C, D and E were related to differences in mean fractal dimension (P < 0.05). A negative correlation existed between fractal dimension and maturation stage (−0.623, P < 0.001). Male and female correlation coefficients were determined to be −0.649 (P < 0.001) and −0.569 (P < 0.001) respectively. A ROC curve was generated and determined the boundary between dichotomous maturation stages A–C and D or E, allowing for fractal dimension to be used to identify midpalatal suture fusion. Predictive statistical analysis noted that fractal dimension is a statistically significant indicator capable of predicting dichotomous maturation stages ((A, B, & C) vs. (D or E) (area under ROC curve [AUC] = 0.794, P < 0.001) [12] (Table 2).

The study notes a significant correlation between fractal patterning and degree of maturation of the midpalatal suture, and consequently the authors feel that fractal analysis can provide an objective and quantitative methodology to assess palatal suture maturity [12].

Disadvantages of this methodology include requiring significant training and proficiency in classifying the maturation stage of palatal sutures as proposed by Angelieri et al. [3]. Another disadvantage is requiring the clinician to have significant familiarity with image processing and possessing necessary software. Consequently, the time, cost and resources to do so may be prohibitive to clinicians. Additionally, this modality relies on complex statistical analyses to determine the variable (optimal cut-off value) to predict the dichotomous maturation stage of the patient’s palatal suture. Kwak et al. [12] argue that if an individual’s fractal dimensions can be compared, it may provide a straightforward and clinically viable method to assess the maturation of the palatal suture and aid in clinical decision making as it relates to the modality of expansion at the diagnostic record visit [12]. Conversely, the authors do note a variety of methods to calculate fractal dimensions and the fact these varying techniques produce different fractal dimension values. Consequently, Kwak et al. [12] argue for a more agreed upon method for its calculation to be utilized clinically.

Performing and interpreting these analyses requires significant advanced knowledge of statistics. Ultimately it is the view of the authors that this methodology is impractical in terms of time, cost, resources and knowledge required to complete this methodology for each patient as part of their diagnostic work up in day-to-day clinical practice.

Furthermore, as was stated previously in the discussion, utilization of the crudely proposed maturational staging as defined by Angelieri et al. [3] should be used with caution and lacks validation to a reference standard as does this study as mentioned by Kwak et al. [12]. Further areas of interest include the development of a ratio comparing the fractal dimensions of a mature coronal suture to that of the midpalatal suture [12]. Additionally, improvement in the accuracy of the methodology may be gained by refinement and minimization of the number of actions needed to determine fractal dimensions [12].