|Year : 2022 | Volume
| Issue : 4 | Page : 531-540
Are preoperative CT parameters predictive for the outcome of septoplasty?
N Sari1, F Cankal2, E Uysal3
1 Mardin Otolaryngology Clinic, Mardin Training and Research Hospital, Mardin, Turkey
2 Ankara Medipol University, Visart Medical Imaging Center, Ankara, Turkey
3 Department of Technique, Vocational School of Technical Sciences, Dicle University, Diyarbakir, Turkey
|Date of Submission||25-Sep-2021|
|Date of Acceptance||31-Jan-2022|
|Date of Web Publication||19-Apr-2022|
Dr. N Sari
Otolaryngology Clinic, Mardin Training and Research Hospital Artuklu/Mardin
Source of Support: None, Conflict of Interest: None
| Abstract|| |
Background: Despite successful septoplasty surgery, some patients still complain of nasal obstruction (NO). Aim: Our aim in the present study is to determine whether preoperative computed tomographic (CT) parameters have prognostic significance for the success of septoplasty. Material and Methods: Retrospective data from 61 patients in a secondary care hospital who had undergone septoplasty met the inclusion and exclusion criteria. The effects of demographic and preoperative CT parameters (internal nasal valve [INV], external nasal valve area, angle of septal deviation, choana area, aperture pyriformis area, high septal deviation [HSD], transverse diameter of midnose, anterior/posterior deviation, concha and meatus diameter [superior, middle, inferior]) were studied in relation to the change in NOSE scores and the success of surgery. Results: Of the 61 patients studied, 31 were male (51%) aged 18–55 years with a mean age ± SD (26.59 ± 9.41). It was found that the changes in NOSE scores were significantly different from each other (P < 0.01). Male gender, trauma history, moderate septal deviation, HSD, mucosal pathology, posterior deviation, bullous turbinate (right and left middle, right superior), and in the absence of allergy, paradoxical turbinate (right and left middle and superior), S-shaped deviation were found to have a significant correlation in the change of NOSE scores (P < 0.05). Age and internal valve differed with respect to outcome (P < 0.05). Variables in multiple linear regresion models of all parameters were found to be insignificant (P < 0.05). Conclusion: Preoperative CT parameters are informative, but do not predict the postoperative success of septoplasty.
Keywords: Computed tomography, predictor, radiologic, septoplasty
|How to cite this article:|
Sari N, Cankal F, Uysal E. Are preoperative CT parameters predictive for the outcome of septoplasty?. Niger J Clin Pract 2022;25:531-40
| Introduction|| |
Nasal septal deviation (NSD) is a deformity of the nasal septum that occurs with a frequency of 70% in adults and 90% in overall. NSD is one of the most common causes of nasal obstruction (NO) with a prevalence of 26.7%. The treatment of NSD is septoplasty, one of the most common surgeries in otolaryngology practice. The success rate of septoplasty is 43–85%. Many anatomical, physiological, and psychological factors have an influence on NO. Despite successful septoplasty, many patients may still experience NO postoperatively which may even lead to medico-legal problems between the physician and patient, or third-party payers who do not provide financial support due to lack of objective evidence. Sometimes the cause may be inadequate surgery, allergy, nasal valve collapse, lateral wall insufficiency, or simply a psychological disorder.
Since the introduction of functional endoscopic sinus surgery (FESS) in 1985, CT has become a mandatory method to understand the paranasal sinus anatomy and its variations, to identify the vital structures, and has become the gold standard in preoperative diagnosis. In septoplasty, our goal is also to evaluate the preoperative radiological indicators that affect the outcome of septoplasty and to inform the patient about the possible outcomes of surgery through the preoperative use of CT. By identifying prognostic factors, we can provide information to patients who are dissatisfied postoperatively despite successful septoplasty, recommend rhinoplasty or nasal valve surgery, or prevent surgery by specifying these prognostic factors and providing objective evidence to payers for patients.
| Patients and Method|| |
Study design and participants
This retrospective study was conducted at a single center hospital between November 2018 and March 2020. About 120 patients are operated for NSD unresponsive to medical management and 61 of them were enrolled in this study (of these, 61 patients with CT were included in this study). This study was conducted in accordance with the ethical principles stated in the Declaration of Helsinki and was approved by the local ethics committee for clinical research (2021/635). Institutional ethics committee approval was obtained before the start of the study. The demographic, clinical, surgical, and follow-up data and photographs were collected in the clinical archive of the first author after each operation. Medical records and CT images were reviewed retrospectively. Inclusionary criteria are as follows: at least 18 years old, to have preoperative paranasal sinus CT (PNS CT) performed after diagnosis of NSD by nasal endoscopic examination, medical records of clinical data consisting of NOSE scores. Exclusion criteria: history of underlying disease that interfere with wound healing like diabetes, immunologic disease, previous history of nasal surgery, coagulation disorders, sinonasal malignancy, need of nasal surgery other than septoplasty (rhinoplasty, FESS, nasal valve surgery), craniofacial malformation, septal perforation.
NO was measured by NOSE score. Patients filled in NOSE scores before the operation and 3 months after the operation. Allergy, trauma, and smoking history was recorded.
After 3 months patients were asked whether they were satisfied with the surgery. Some of them have no complaints (pass group). Some of them said that they have few and sometimes complaints (fail group) and the others said that they were not satisfied at all (fail group). Continous variables were given in [Table 1] and categorical variables were given in [Table 2].
|Table 1: Distribution of parameters (continous variables) in patients between groups|
Click here to view
|Table 2: Distribution parameters (categorical variables) according to success and result of Chi-square test|
Click here to view
All surgical procedures were done by the same surgeon under general anesthesia. No patient was given premedication. About 2% lidocaine with 1:100.000 adrenaline was infiltrated submucosally before incision. Hemi-transfixion incision followed by mucoperichondrium elevation in both sides, adressing areas of deviation and reshaping and removing the deviated part of cartilage. Then mucoperichondrial flap is sent back to its original place and sutured by 4.0 vicryl. Inferior concha were cauterized for size reduction under 00 endoscope. If bullous middle turbinate were present, lateral resection of middle turbinate under visualisation was performed. Bilateral silicone Doyle type nasal splint nasal was routinely used. They were removed on the second postoperative day. Ringer lactate and almond oil was used after the operation to clean and soften the crusts.
All patients were scanned with the same spiral CT scan (Aquilion 16 CFX; Toshiba Medical System, Tokyo, Japan). Extended images of all paranasal sinuses at a detector configuration of 16 × 0.5 mm within a range of 0.5 mm and a reconstruction width of 0.35 mm. Secondary coronal and sagittal multiplanar reconstructions (MPRs) were constructed from original axial sections. No contrast agent was used.
An image-processing software dedicated to digital imaging and communications in medicine (DICOM) images, was used to visualize the digital three-dimentional volumetric image on axial, coronal, sagittal planes. Images were analyzed by a single radiologist on two seperate dates.
A total of 61 PNS CT were analyzed according to radiological parameters that play a role on passage of air according to literature and clinical experience. The presence and angle of high septal deviation (HSD) and the angle of septal deviation were measured using coronal images from the level where the deviation is most prominent. Angle of internal nasal valve (INV) and external nasal valve was measured on the axial sections where these structures were clearly distinguished. Inferior/middle turbinate diameters, inferior/middle/superior meatus thickness, thickness of anterior cartilage, and vomer are measured. SD side, SD angle grading and presence of HSD, area of apertura pyriformis and choana were measured in coronal sections. The area of the external nasal valve was measured atthe at the level of the nasal aperture in the axial sections, and the area of the nasal valve was measured at the level of the head of the inferior turbinate as seen in the [Figure 1] in the axial sections. Presence of mucosal pathology, bullous middle and superior turbinate, paradoxical middle and superior turbinate, S-shaped deviation, A/P deviation, and midnose transverse diameter was interpreted by evaluating axial and coronal sections together.
|Figure 1: Preoperative perimeter and area of INV in axial plane. P: perimeter|
Click here to view
In this study, the normality distribution assumption for continous variables of the obtained data was investigated by Shapiro–Wilk test and the homogenity was investigated with the Levene test. Independent t-test is used to compare patients between pass and fail groups for continous variables and Chi-square test for categorical variables. Paired samples statistics test is used to evaluate the differences of preoperative and postoperative measurements for different parameters, Pearson correlation analysis in examining the relationships of the parameters and multiple linear regression test analysis techniques were used, taking into account the assumption of whether there is a linear relationship between the independent variables (collinearity). Intra-observer reliability coefficients were determined using Pearson correlation coefficients. Reliability coefficients >0.7 were considered acceptable.
Descriptive statistics and test analyzes were performed using R version 3.2.3 (2015-12-10), Copyright (C) 2015 The R Foundation for Statistical Computing free software computer software. Results for P < 0.05 were considered statistically significant.
| Results|| |
A total of 61 patients were included in the study because they met the eligibility criteria. About 31 (51%), of them are man, 30 (49%) of them are women with age range of 18–55 and mean age ± SD (26 ± 9.41). Follow-up of patients were 3 months. Demographic data are given in [Table 1] and [Table 2].
In [Table 1], only age and left internal valve area (INV) [Figure 1] were found to be significantly different between the groups (P < 0.05). In [Table 2] none of the categorical parameters were found to be significantly different (P > 0.05).
Change of preoperative and postoperative NOSE scores were found to be significantly different from each other (P < 0.01). Mean of preoperative NOSE score is 19.3 ± 1.1 and postoperative NOSE score is 5.22 ± 4.73. Minimum changed NOSE (cNOSE) score was − 20 and maximum cNOSE was − 2. Mean cNOSE is − 14.09 ± 4.74.
Evaluation of the differences in preoperative and postoperative measurements of NOSE scores for parameters; no significant differences were found between the groups (P > 0.05) [Table 3].
|Table 3: Summary of NO symptom evaluation (NOSE) total score before and after operation with regard to characteristics|
Click here to view
In [Table 4], statistically, positive correlation were found between change of NOSE scores in male gender (r = 0.499; P = 0.004), moderate septum deviation (r = 0.421, P = 0.026), posterior deviation (r = 0.428, P = 0.003), in positive subgroups; trauma history (r = 0.447, P = 0.022), high septum deviation (r = 0.416, P = 0.022) [Figure 2] and [Figure 3], mucosal pathology (r = 0.372, P = 0.007), right middle bullous turbinate (r = 0.404, P = 0.016), left middle bullous turbinate (r = 0.354, P= 0.018), right superior bullous turbinate (r = 0.448, P= 0.032), in negative subgroups; in allergy (r = 0.452, P= 0.026), paradoxical turbinate (right middle (r = 0.317, P= 0.023) and left middle (r = 0.3, P= 0.008), right superior (r = 0.336, P = 0.015), left superior (r = 0.329, P= 0.018), S shaped deviation (r = 0.388, P= 0.023).
|Figure 3: Positive correlation x axis: change of preoperative NOSE scores and y axis: postoperative change of NOSE scores seen on the graphic in patients with HSD (+). One example of graph of positive correlations on [Table 4]|
Click here to view
|Table 4: Correlations in parameters between pre-operatif and post-operatif NOSE scores among parameters|
Click here to view
We performed multiple linear regression analysis to determine significant predictive factors of the iNOSE. Variables in each model were found to be insignificant (P > 0.05) [Table 5]. R2 signifies independent factors explaining and describing percentage of dependent variable iNOSE, a low value is found R2 = 0.03652, approximately 3.7%.
The reliability of the measurements was tested, the intra-observer reliability were found between 0.95 and 0.98. Results were considered safe and measurements were averaged.
| Discussion|| |
There is a gap in knowledge regarding the diagnostic tool that objectively measures the presence of NO after septoplasty. Anterior rhinoscopy and nasal endoscopy are golden standards in the assessment of NO. Other assessment methods include question-based studies; Nose Obstruction Symptom Evaluation (NOSE), Sinonasal Outcome Test-22 (SNOT-22) rhinomanometry and acoustic rhinometry. CT is also used as an assessment method. Some authors advocate CT examination while others are against it. Some of them advise CT examination in special situations. For example, CT is recommended for in middle and posterior deviations of the septum. In cases of anterior septal deviations preventing the passage of the endoscope, inability to evaluate the middle meatus and posterior nasal cavity, obstructive hypertrophy of the middle turbinate, the use of CT is recommended. In clinical practice, both patients and insurance companies may use CT to confirm the success of surgery although its use is controversial. To resolve this ambiguity, we investigated whether preoperative radiological parameters predict the outcome of septoplasty.
Deciding on septal surgery based on a radiological report can be misleading. Mild, moderate, and severe NSD do not have the same degree of NO. Poor correlation has been found with the degree of deviation on CT and NOSE scores. In our study, no significant correlation was found between deviation site and degree of deviation, but some weak correlations were found among HSD, moderate septal deviation (SD), S deviation positive patients, and those who are negative. HSD is the septal deviation touching the nasal bridge at the level of the posterior margin of the frontal recess on the coronal and sagittal planes. Manipulations at perpendicular plate of the ethmoid bone are avoided because of the fear of disturbing the keystone area, risk of damage to the olfactory area, and the possibility of developing cerebrospinal fluid fistula. Deviation of perpendicular plate of ethmoid bone was observed in 44% of patients who required revision surgery. In addition, HSD-positive patients are candidates for rhinoplasty. high-level horizontal chondrotomies and osteotomies performed during septoplasty or rhinoplasty in such patients may result in damage to the K-area causing saddling. To avoid this, proper placement and adequate amount of L-strut is required. In our study, 49% of cases were HSD positive patients. Obstruction may be observed in patients with HSD as the incidence of deviation increases in the upper part of residual septum and surgeon may avoid high-level chondrotomy as this area is vulnerable to complications.
Narrowing or obstruction of the nasal valve leads to NO. There are nonsurgical and surgical treatment options for nasal valve pathologies. Surgical treatment of the nasal valve involves the placement of grafts as part of a functional rhinoplasty. Frequency of INV collapse is observed at 48%. External nasal valve collapse at 40%, INV narrowing at 95%, and external valve narrowing at 45% were seen. Rhinomanometry and acoustic rhinometry have been used to assess INV obstruction but are limited and therefore not reliable. CT has been proposed as a non-invasive, safe, objective tool to measure INV collapse. INV angle and area measurements have been suggested prior to septoplasty and rhinoplasty surgeries to manipulate the surgery when the problem was seen in CT especially in nasal base view where CT scans were reformatted to a plane perpendicular to the estimated acoustic axis in one of the studies. In our study, no correlation was found between the change in NOSE scores and the INV angle and area in CT, but the left INV area in successful operations differed from patients with complaints [Table 1]. Nasal valve dysfunction is an underdiagnosed entity and should be considered in all patients with NSD before undergoing septoplasty.
The diameter of choana, midnose, and pyriform aperture, as well as the degree of NSD, may play a role in NO. The angle of deviation, cross-sectional areas at the ostiomeatal unit (OMU), and choana level were found to correlate significantly with NOSE scores, whereas no correlation was found for the nasal valve area. CT was suggested as an adjunct to rhinometry for the middle and posterior nasal regions in one of the studies. A moderate degree of NSD angle affects the change in NOSE scores in our study. However, another study found a poor correlation between the degree of deviation on CT and NOSE scores. A nasal floor slope of 3° or more and a nasal floor height difference of 1.5 mm is associated with NO. In addition, a pyriform aperture width of 22 mm or less can be considered narrow. In our study, choana, mid-nose, and pyriform aperture did not affect the change in NOSE scores and no difference was found in the success of septoplasty [Table 1].
According to the evaluation of PNCTs, 8.3% of patients who underwent septoplasty underwent other surgical procedures such as concha bullosa or some minor endoscopic procedures. The incidence of middle concha bullosa is 14-54% causing NO. Therefore, any obstructing pathology in the middle meatus or enlargement of the middle concha causes NO. In our study, patients who had bullous middle turbinates on both the right and left side on CT were correlated with a change in NOSE score but there was no differences between successful surgeries and others in terms of bullous middle turbinate. In literature, the incidence of superior pneumatized turbinates ranges from 12.2% to 50%, which is consistent with our results, [Table 1]. While right superior bullous turbinate was statistically correlated with the change in NOSE score, patients without paradoxical turbinate were found to be correlated with change in NOSE, and no difference was found in terms of success (P > 0.05).
NO is a multifactorial entity. Many parameters causing NO affect septoplasty outcomes. Allergy, mucosal thickening, age, sex, and smoking play a role in NO postoperatively after septoplasty. Age, sex, smoking, and allergy had no effect on NO after septoplasty. In another study, allergic rhinitis, was found to have effect on NO, while body mass index (BMI) do not. Our results are shown in [Table 1], [Table 2], [Table 4], and [Table 5], and no significant effect was found on NO (P < 0.05).
In examining the predictive utility of preoperative CT, the deleterious effects of radiation particularly eye and thyroid, must also be considered. In recent years, high-quality images can be produced with low doses using the protocols of Cone beam CT. It is recommended to use low doses, especially in children, and it has been reported that images of diagnostic quality are obtained with conventional fan-beam CT even with ultra-low voltage (70 kVp). If the method becomes accepted, there will be fewer side effects and it can be widely used to obtain safer information before surgery.
Prior to septoplasty, patients should be carefully evaluated to determine if they need rhinoplasty, nasal valve surgery, or drug treatment for allergies and sinusitis, etc. We made an effort to detect these situations with the parameters of CT but in our work in [Table 4], some weak correlations were found statistically. If we increase the number of patients, these correlations could be significant in a larger group of patients. The small sample size and lack of control (postoperative CT) because of radiation risk are the weaknesses of our study. Further studies are needed for further evaluation with larger groups of patients. Another limitation of our study is the fact that it is a retrospective study and the NOSE score where patient-reported outcomes have the potential for reporting bias. Further studies may be helpful, especially with new models that measure NO differently than the existing models, through innovations in technology such as artificial intelligence, etc., and the causes of septoplasty failure may be better understood.
| Conclusion|| |
Preoperative CT parameters are informative, but do not predict the postoperative success of septoplasty.
We would like to thank our colleagues and the nursing staff at Kızıltepe Goverment Hospital, Mardin, Turkey, for their kind cooperation and timely help.
The study was conducted and completed in accordance with the ethical principles outlined in the Declaration of Helsinki and was approved by the Ethics Committee of Gazi Yaşargil Education and Research Hospital in Diyarbakır, Turkey (2021/635).
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Mladina R, Cujić E, Subarić M, Vuković K. Nasal septal deformities in ear, nose, and throat patients: An international study. Am J Otolaryngol 2008;29:75-82.
Bezerra TF, Stewart MG, Fornazieri MA, Pilan RR, Pinna Fde R, Padua FG, et al
. Quality of life assessment septoplasty in patients with nasal obstruction. Braz J Otorhinolaryngol 2012;78:57-62.
Derin S, Sahan M, Deveer M, Erdogan S, Tetiker H, Koseoglu S. The causes of persistent and recurrent nasal obstruction after primary septoplasty. J Craniofac Surg 2016;27:828-30.
Kahveci OK, Miman MC, Yucel A, Yucedag F, Okur E, Altuntas A. The efficiency of Nose Obstruction Symptom Evaluation (NOSE) scale on patients with nasal septal deviation. Auris Nasus Larynx 2012;39:275-9.
Janovic N, Janovic A, Milicic B, Djuric M. Is computed tomography ımaging of deviated nasal septum justified for obstruction confirmation? Ear Nose Throat J 2021;100:131-6.
Cantone E, Ricciardiello F, Oliva F, De Corso E, Iengo M. Septoplasty: Is it possible to identify potential “predictors” of surgical success? Acta Otorhinolaryngol Ital 2018;38:528-35.
Cashman EC, Macmahon PJ, Smyth D. Computed tomography scans of paranasal sinuses before functional endoscopic sinus surgery. World J Radiol 2011;28:199-204.
Onerci Altunay Z, Onerci TM. The relationship of high septal deviation, the depth of olfactory fossa, and gera angle: Is high septal deviation associated with any anatomic abnormalities in the anterior skull base? Ear Nose Throat J 2020;29:145561320926406.
Patel B, Virk JS, Randhawa PS, Andrews PJ. The internal nasal valve: A validated grading system and operative guide. Eur Arch Otorhinolaryngol 2018;275:2739-44.
Buyuklu F, Cakmak O, Hizal E, Donmez FY. Outfracture of the inferior turbinate: A computed tomography study. Plast Reconstr Surg 2009;123:1704-9.
Ardeshirpour F, McCarn KE, McKinney AM, Odland RM, Yueh B, Hilger PA. Computed tomography scan does not correlate with patient experience of nasal obstruction. Laryngoscope 2016;126:820-5.
Dizdar D, Bozan A, Dizdar SK, Göde S, Alpay HC. Evaluation of nasal symptoms in septoplasty patients using SNOT-22. Acta Otorhinolaryngol Ital 2019;39:98-102.
Lee DC, Shin JH, Kim SW, Kim SW, Kim BG, Kang JM, et al
. Anatomical analysis of nasal obstruction: Nasal cavity of patients complaining of stuffy nose. Laryngoscope 2013;123:1381-4.
Günbey E, Günbey HP, Uygun S, Karabulut H, Cingi C. Is preoperative paranasal sinus computed tomography necessary for every patient undergoing septoplasty? Int Forum Allergy Rhinol 2015;5:839-45.
Kayabasoglu G, Nacar A, Yilmaz MS, Altundag A, Cayonu M, Guven M. A novel method for nasal dorsal reconstruction: Permanent fixation using Kirschner wire–guided hidden sutures. J Craniofac Surg 2015;26:881-4.
Shafik AG, Alkady HA, Tawfik GM, Mohamed AM, Rabie TM, Huy NT. Computed tomography evaluation of internal nasal valve angle and area and its correlation with NOSE scale for symptomatic improvement in rhinoplasty. Braz J Otorhinolaryngol 2020;86:343-50.
Samra S, Steitz JT, Hajnas N, Toriumi DM. Surgical management of nasal valve collapse. Otolaryngol Clin North Am 2018;51:929-44.
Chambers KJ, Horstkotte KA, Shanley K, Lindsay RW. Evaluation of improvement in nasal obstruction following nasal valve correction in patients with a history of failed septoplasty. JAMA Facial Plast Surg 2015;17:347-50.
Cakmak O, Coşkun M, Celik H, Büyüklü F, Ozlüoğlu LN. Value of acoustic rhinometry for measuring nasal valve area. Laryngoscope 2003;113:295-302.
Gillman GS, Egloff AM, Rivera-Serrano CM. Revision septoplasty: A prospective disease-specific outcome study. Laryngoscope 2014;124:1290-5.
Riechelmann H, Widmann G, Kofler B, Arminger R, Url C, Giotakis AI. Nasal floor asymmetry is associated with nasal obstruction. J Oral Maxillofac Surg 2020;78:1833.e1-9.
Stallman JS, Lobo JN, Som PM. The incidence of concha bullosa and its relationship to nasal septal deviation and paranasal sinus disease. AJNR Am J Neuroradiol 2004;25:1613-8.
İla K, Yilmaz N, Öner S, Başaran E, Öner Z. Evaluation of superior concha bullosa by computed tomography. Surg Radiol Anat 2018;40:841-6.
Ozcan KM, Selcuk A, Ozcan I, Akdogan O, Dere H. Anatomical variations of nasal turbinates. J Craniofac Surg 2008;19:1678-82.
Stewart MG, Smith TL, Weaver EM, Witsell DL, Yueh B, Hannley MT, et al
. Outcomes after nasal septoplasty: Results from the Nasal Obstruction Septoplasty Effectiveness (NOSE) study. Otolaryngol Head Neck Surg 2004;130:283-90.
Mondina M, Marro M, Maurice S, Stoll D, de Gabory L. Assessment of nasal septoplasty using NOSE and RhinoQoL questionnaires. Eur Arch Otorhinolaryngol 2012;269:2189-95.
Karatzanis AD, Fragiadakis G, Moshandrea J, Zenk J, Iro H, Velegrakis GA. Septoplasty outcome in patients with and without allergic rhinitis. Rhinology 2009;47:444-9.
Schulz B, Potente S, Zangos S, Friedrichs I, Bauer RW, Kerl M, et al
. Ultra-low dose dual-source high-pitch computed tomography of the paranasal sinus: Diagnostic sensitivity and radiation dose. Acta Radiol 2012;53:435-40.
Feragalli B, Rampado O, Abate C, Macrì M, Festa F, Stromei F, et al
. Cone beam computed tomography for dental and maxillofacial imaging: Technique improvement and low-dose protocols. Radiol Med 2017;122:581-8.
Chi J, Xu D, Yin S, Li M, Shen L, Ding N, et al
. Reducing the radiation dose of pediatric paranasal sinus CT using an ultralow tube voltage (70 kVp) combined with iterative reconstruction: Feasibility and image quality. Medicine (Baltimore) 2020;99:e21886.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]