Photomedicine and Laser Surgery

Volume XX, Number XX, 2016 Original Research

ª Mary Ann Liebert, Inc.
Pp. 1–5
DOI: 10.1089/pho.2016.4117

Role of Low-Level Laser Therapy as an Adjunct

to Initial Periodontal Treatment in Type 2 Diabetic Patients:
A Split-Mouth, Randomized, Controlled Clinical Trial

Oya Demirturk-Gocgun, DDS, PhD,1 Ulku Baser, DDS, PhD,1 Gokce Aykol-Sahin, DDS, PhD,2
Nevin Dinccag, MD, PhD,3 Halim Issever, MD, PhD,4 and Funda Yalcin, DDS, PhD1

Abstract

Objective: In this split-mouth clinical trial, we evaluated the clinical benefits of low-level laser therapy (LLLT)
as an adjunct to nonsurgical periodontal treatment in patients with type 2 diabetes mellitus (DM). Background
data: The impaired wound healing seen in diabetic patients may affect the results of periodontal treatment and
may require an additional approach. Materials and methods: In total, 22 chronic periodontitis patients with type
2 DM were included. Applying a split-mouth design, two quadrants were treated with only scaling and root
planing (SRP) as the control and those in the other two were treated with SRP + LLLT as the test sites in each
patient. An 808 nm GaAlAs diode laser was performed in the test sites at the energy density of 4.46 J/cm2 on days
1, 2, and 7 after SRP. Plaque index (PI), probing depth (PD), bleeding on probing (BOP), and clinical attachment
level were measured at baseline and again at 1 and 3 months after treatment. Deep periodontal pockets (PD
‡4 mm) were evaluated separately. Results: Test sites showed significant improvement in PI and BOP in deep
pockets at the 1-month follow-up period ( p < 0.001 and <0.001, respectively), whereas no difference was found
between the control and the test sites in other periodontal parameters. Conclusions: LLLT during periodontal
treatment offered minimal short-term additional benefit in deep pocket healing in patients with type 2 DM.

Keywords: diabetes mellitus type 2, laser therapy, low level, periodontology

Introduction LLLT as an adjunct to nonsurgical periodontal treatment in

periodontal wound healing in smokers and nonsmokers with
D iabetes Mellitus (DM) is caused by a deficiency in
insulin or its action, resulting in hyperglycemia and
hyperlipidemia, which are involved in the development of
chronic periodontitis were evaluated and that previous study
indicated that LLLT, as an adjunctive therapy to nonsurgical
periodontal treatment, improved periodontal healing.12 In
many systemic and oral complications, such as period-
the current study, because DM, like smoking, affects the
ontitis.1,2 Studies have shown that type 2 DM is a risk factor
healing of periodontal tissues, it was hypothesized that LLLT
for periodontal diseases and those patients with diabetes
would have positive effects on the microcirculation and on
show higher prevalence and severity of periodontitis com-
collagen and cytokine production, which are negatively af-
pared with nondiabetic patients.3–5 The impaired wound
fected by DM. Thus, the purpose of this split-mouth, ran-
healing seen in diabetic patients may affect the results of
domized, controlled clinical trial was to evaluate the effects
periodontal treatment and may require an additional ap-
of LLLT as an adjunct to nonsurgical periodontal treatment
proach. According to the evidence, low-level laser therapy
on clinical parameters in type 2 DM patients with chronic
(LLLT), also called photobiomodulation, has potential utility
periodontitis.
in accelerating the inflammatory response of wound healing.
Clinical studies have shown that LLLT promotes wound
Materials and Methods
healing by accelerating inflammation,6 collagen synthesis,7
healing time, and strength acquisition.8 Further, the studies The study was approved by the Research Ethics Com-
about LLLT as a treatment method for diabetic ulcers have mittee of Istanbul University of Medical Sciences (protocol
shown positive results.9–11 In a previous study, the effects of number: 2012/1467/1223). Written informed consent was
1
Department of Periodontology, Faculty of Dentistry, Istanbul University, Istanbul, Turkey.
2
Department of Periodontology, Faculty of Dentistry, Okan University, Istanbul, Turkey.
3
Division of Diabetes, Department of Internal Medicine, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkey.
4
Department of Biostatistics, Istanbul Faculty of Medicine, Istanbul University, Istanbul, Turkey.

1
2 DEMIRTURK-GOCGUN ET AL.

obtained from all patients before their enrollment in the Table 1. Parameters of LLLT
study. The study protocol was approved and registered un- Energy per point/per tooth 1.25/5 J
der International Clinical Trials Registry Platform with the Power density 0.89 W/cm2
Thai Clinical Trials Registry, TCTR20150730001. Energy density/per point 4.46 J/cm2
Between October 2013 and February 2014, 80 type 2 DM Duration of each point 5 sec
patients being followed as outpatients in the Department of Four points/per gingiva 20 sec
Endocrinology, Istanbul Medical Faculty, Istanbul Uni- of root surface
versity, were screened. Patients who had type 2 DM were Frequency of the treatment 3 times (days 1, 2, and 7)
selected and classified based on the criteria of the American Cumulative dose for 3 days 13.38 J/cm2
Diabetic Association (2011) and glycated hemoglobin levels LLLT, low-level laser therapy.
(HbA1C).13 All diabetic patients were followed up regularly
in the department of endocrinology and all had stable dia-
betes. Among the type 2 DM patients, 22 patients with
reference point. A trained dentist, who was blinded to the
moderate to severe chronic periodontitis were selected for
treatment received by the patients, conducted periodontal
the study,14 and periodontal disease severity was defined
examinations.
according to Page and Eke.15
Exclusion criteria were any kind of periodontal treatment
during the last 6 months or use of antibiotics during the last 3 Statistical analysis
months, smoking, acute oral or systemic infection, periodontal
abscess, hemorrhagic disorders, autoimmune diseases, preg- For sample size calculation, BOP% was used as the pri-
nancy, or <15 teeth and partial dentures. Teeth with fixed mary outcome variable. According to the results of the
prosthodontics, and a mobility grade of III or pockets deeper power analysis, a sample size of 80 quadrants in 20 par-
than 10 mm in the studied areas were not evaluated. ticipants was identified for 80% statistical power, b = 0.20,
All patients received oral hygiene instructions and su- and a = 0.05 (to detect D = 10%).
pragingival scaling in two appointments, 1 week apart, be- The data collected were analyzed using a statistical soft-
fore root planing treatment. Full-mouth subgingival scaling ware package (SPSS, version 15.0; SPSS, Chicago, IL). Sites
and root planing (SRP) under local anesthesia (Ultracain were divided into two subgroups according to initial PD23:
DS; Aventis Pharmacheuticals, Istanbul, Turkey) was per- shallow pockets (0–3 mm); moderate and deep pockets
formed by the same operator in a single appointment for (4–10 mm). Differences between test and control sites and
each patient in all groups using hand instruments (Gracey between different time points were analyzed using the Mann–
Curets; Hu-Friedy, Chicago, IL) and ultrasonic devices Whitney U test and Wilcoxon signed-rank test, respectively.
(Cavitron; Dentsply, York, PA). Statistical significance was set at the 99% confidence level
The study was performed according to a split-mouth de- ( p < 0.01) for the Mann–Whitney U test and 95% ( p < 0.05)
sign, and each patient was included in a clinical protocol for the Wilcoxon signed-rank test.
consisting of two different modalities: teeth were treated by
SRP alone (control site) and by SRP followed by LLLT (test Results
site). Two quadrants were randomly assigned to the control
and two to the test group for each patient. Randomization Descriptive results
was performed using a computer-generated random alloca- Twenty-two participants attended the baseline examina-
tion table. Each selected segment first received a code tion, treatment, and the follow-up appointments over a 3-
number, and one of the study coordinators used a computer- month period. The total numbers of test and control sites
generated table to randomly allocate each quadrant to the were 1518 and 1548, and the test and control sites with PD
SRP (control, n = 22) or SRP + LLLT (test, n = 22) group. ‡4 mm were 698 and 669, respectively. Demographic vari-
A GaAlAs diode laser (k = 808 nm, Fotona XD-2; FOTONA ables described in Table 2.
D.O.O., Ljubljana, Slovenia) was used in this study. The
laser (output power, 0.25 W; continuous mode; handpiece,
R24-B; spot diameter, 6 mm; spot size, 0.28 cm2) was applied Table 2. Descriptive Parameters of the Study Group
with noncontact technique to the root surfaces through the
No. of patients 22
labial and oral gingiva at the test sites after the periodontal Mean age (min/max years) 41/72, 50.50 – 8.72
treatment. The parameters16–20 applied in the protocol are (mean – SD)
described in Table 1. Female (n, %) 15, 68.2
Clinical periodontal parameters were bleeding on probing Male (n, %) 7, 31.8
(BOP), plaque index (PI),21 gingival index (GI),22 probing Pocket depth ‡4 mm (%) 44.5, 22.7 (test)/21.8 (control)
depth (PD), and clinical attachment level (CAL). The pri- Diabetes duration (min/ 3/13, 5.40 – 2.55
mary outcome was BOP changes in deep pockets of the test max years) (mean – SD)
and control sites over the 3-month follow-up period. PI, GI, HbAIc at baseline (%) 5.50/10.90, 7.45 – 1.43
PD, and CAL were our secondary outcomes. All measure- (min/max) (mean – SD)
ments were recorded at six sites per tooth (mesio-, mid-, and Oral antidiabetic 15, 68.2
therapy (n, %)
disto-vestibular, and mesio-, mid-, and disto-palatal) at base- Oral antidiabetic + insulin 7, 31.8
line and at 1 and 3 months after treatment with a periodontal therapy (n, %)
Williams probe (PW6; Hu-Fredy, Chicago, IL), calibrated in
millimeters. The cemento–enamel junction was used as a SD, standard deviation.
LLLT ON DIABETIC PATIENTS 3

Table 3. Differences in Clinical Parameters

Parameter Difference Difference
mean – SD Baseline 1 Month (0- to 1-month) p* 3 Months (0- to 3-month) p{
PI
Test 1.81 – 0.36 0.91 – 0.39 0.89 – 0.38 0.000 0.74 – 0.26 1.06 – 0.31 0.000
Control 1.78 – 0.36 1.06 – 0.32 0.71 – 0.34 0.000 0.81 – 0.20 0.97 – 0.34 0.000
p* 0.122 0.364
BOP (mm)
Test 0.88 – 0.11 0.39 – 0.17 0.49 – 0.15 0.000 0.33 – 0.15 0.54 – 0.17 0.000
Control 0.87 – 0.14 0.44 – 0.16 0.43 – 0.16 0.000 0.33 – 0.16 0.54 – 0.19 0.000
p* 0.225 0.954
PD (mm)
Test 3.94 – 0.68 2.80 – 0.58 1.13 – 0.31 0.000 2.74 – 0.54 1.20 – 0.53 0.000
Control 3.86 – 0.57 2.79 – 0.56 1.06 – 0.23 0.000 2.69 – 0.46 1.16 – 0.30 0.000
p* 0.395 0.769
CAL (mm)
Test 4.30 – 1.01 3.10 – 0.89 1.20 – 0.34 0.000 3.14 – 0.98 1.16 – 0.89 0.000
Control 3.91 – 1.06 2.99 – 0.73 0.92 – 0.75 0.000 2.89 – 0.62 1.02 – 0.78 0.000
p* 0.123 0.573
*Mann–Whitney U test, p < 0.01.
{
Wilcoxon test, p < 0.05.
BOP, bleeding on probing; CAL, clinical attachment level; PD, probing depth; PI, plaque index.

Clinical parameters
There was no statistically significant difference in the
clinical baseline measurements (PI, BOP, PD, and CAL)
between the test and control sites (Table 3). Within the sites,
statistically significant improvements ( p < 0.05) were noted
for PI, BOP, PD, and CAL when baseline values were
compared with the 3-month follow-up evaluation (Table 3).
The improvements in these clinical parameters were slightly
better in the test sites, but the differences were not statisti-
cally significant ( p > 0.01).
The clinical measurements of deep pockets were analyzed
and are shown in Table 4. BOP and PI scores showed sta-
tistically significant improvements in the test sites between
baseline and the first month. The reduction in PI was also
significantly greater in the test than in the control sites be-
tween baseline and the third month. There was no significant
difference between control and test sites in PD or CAL
changes at follow-up.
Discussion
LLLT has been shown to be an effective modality in
patients who suffer from diabetic wounds, especially for
diabetic foot ulcers.10 Research findings to date, based on
in vitro animal and human studies, have shown that LLLT
can play a useful role in healing chronic diabetic ulcers

Table 4. Differences in Clinical Parameters (>4 mm PD)

Parameter Difference Difference
mean – SD Baseline 1 Month (0- to 1-month) p* 3 Months (0- to 3-month) p{
PI
Test 1.98 – 0.60 1.02 – 0.86 0.95 – 0.93 0.000 0.87 – 0.80 1.11 – 0.88 0.000
Control 1.96 – 0.60 1.25 – 0.82 0.71 – 0.90 0.000 0.97 – 0.78 0.99 – 0.92 0.000
p* 0.000 0.000
BOP (mm)
Test 0.94 – 0.23 0.48 – 0.50 0.46 – 0.51 0.000 0.40 – 0.49 0.54 – 0.52 0.000
Control 0.93 – 0.26 0.59 – 0.49 0.33 – 0.52 0.000 0.40 – 0.49 0.52 – 0.54 0.000
p* 0.000 0.657
PD (mm)
Test 5.31 – 1.08 3.37 – 1.28 1.93 – 1.09 0.000 3.17 – 1.19 2.14 – 1.14 0.000
Control 5.30 – 0.96 3.41 – 1.23 1.88 – 1.14 0.000 3.24 – 1.15 2.05 – 1.11 0.000
p* 0.808 0.337
CAL (mm)
Test 5.73 – 1.45 3.90 – 1.65 1.94 – 1.13 0.000 3.60 – 1.51 2.13 – 1.19 0.000
Control 5.53 – 1.15 3.68 – 1.42 1.90 – 1.15 0.000 3.48 – 1.34 2.04 – 1.14 0.000
p* 0.854 0.255
*Mann-Whitney U test, p < 0.01.
{
Wilcoxon test, p < 0.05.
4 DEMIRTURK-GOCGUN ET AL.
resistant to conventional treatments.9,24 The use of LLLT in
periodontal treatment is still controversial. One study has
shown no additional benefit,25 whereas others have shown
that LLLT has demonstrable clinical efficacy, including our
previous study that showed additional improvement in
smokers.12,26–29 The present study was conducted from the
perspective that more dramatic changes may occur among
patients with impaired wound healing, such as those with
diabetes mellitus (DM).
Limited clinical research was found in the literature
when the search was focused on adjunctive effects of LLLT
during the periodontal therapy in patients with DM2.30,31
Although there were methodological differences with our
study, two studies reported that the effects of LLLT on
periodontal tissues were mediated by DM.30,31 A histolog-
ical study conducted by Obradović et al.30 reported that
more pronounced signs of gingival tissue healing were ob-
served after applying LLLT (670 nm, 2 J/cm2) as an adjunct
to nonsurgical periodontal treatment in patients with type
1 and type 2 DM. A subsequent study evaluated the effects
of LLLT by exfoliative cytology in patients with DM
and gingival inflammation; more pronounced GI reduction
and lower cellular parameters were noticed on the laser-
treated side of the jaw than on the nonlaser side.31
The present study was designed to evaluate the role of
LLLT as an adjunct to nonsurgical periodontal treatment in
deep pockets among patients with type 2 DM. When deep
pockets were compared, test sites showed significant im-
provements in BOP between baseline and 1 month. LLLT
has been shown to be effective in the treatment of impaired
microcirculation, reducing inflammation, and swelling.
Moreover, it can facilitate collagen synthesis, angiogenesis,
and reepithelialization, which eventually accelerate wound
healing.32–34 Thus, the greater reduction in gingival bleed-
ing in deep pockets at the test sites can be explained by the
anti-inflammatory effect of vascular stimulation by the
LLLT. Considering that the improvement was present at the
1-month follow-up, but not later, it is thought that LLLT
may have reduced inflammation in the initial healing of the
wounds and attachment of junctional epithelium.
When all pocket depths were analyzed, the test sites did
not show any additional improvement compared with the
control sites, as measured by any of the periodontal clinical
parameters examined (PI, BOP, PD, and CAL). Some
studies have investigated the use of LLLT as an adjunct in
nonsurgical periodontal treatment in systemically healthy
patients.12,27–29 These studies reported that LLLT as an
adjunct to periodontal treatment showed significant clinical
improvements. Insufficient irradiation may have been the
cause of poor results in some studies due to the lack of
consensus about dosage.35 In this study, the LLLT was ad-
ministered on days 1, 2, and 7 after the initial periodontal
treatment at an energy density of 4.46 J/cm2. Although we
found significant enhancement in BOP parameters between
the baseline and 1 month, the effects of treatment on other
clinical parameters were not significant. This might have
been due to insufficient irradiation. Although their patients
were not diabetic, Pejcic et al. reported that the effect of
LLLT was more pronounced after the fifth application of the
laser;29 further, Obradović et al. achieved positive effects
after five consecutive applications of LLLT in DM pa-
tients.31 This limitation might also explain why significant
BOP reduction in the first month of the test sites was not
maintained to the third month. Thus, given these results,
there is a need to evaluate the effects of different dosages on
periodontal tissues in diabetic patients. Further studies are
also required to establish optimal treatment protocols, such
as wavelength, fluency, intensity, exposure time, and total
duration of the treatment.
Conclusions
This split-mouth, randomized clinical trial indicated that
the use of a low-level laser as an adjunct to SRP showed a
minor short-term additional benefit on gingival bleeding, but
it did not significantly enhance other clinical parameters.
Acknowledgments
This work was supported by the Scientific Research
Projects Coordination Unit of Istanbul University, project
number; T-39791. The authors declare that there are no
conflicts of interest in this study. No external funding, apart
from the support of the authors’ institutions, was available
for the study.
Author Disclosure Statement
No competing financial interests exist.
References
1. Mealey BL, Ocampo GL. Diabetes mellitus and periodontal
disease. Periodontol 2000 2007;44:127–153.
2. Taylor GW, Borgnakke WS. Periodontal disease: associa-
tions with diabetes, glycemic control and complications.
Oral Dis 2008;14:191–203.
3. Khader YS, Dauod AS, El-Qaderi SS, Alkafajei A, Batayha
WQ. Periodontal status of diabetics compared with non-
diabetics: a meta-analysis. J Diabetes Complications 2006;
20:59–68.
4. Taylor GW, Burt BA, Becker MP. Non-insulin dependent
diabetes mellitus and alveolar bone loss progression over 2
years. J Periodontol 1998;69:76–83.
5. Tsai C, Hayes C, Taylor GW. Glycemic control of type 2
diabetes and severe periodontal disease in the US adult
population. Community Dent Oral Epidemiol 2002;30:
182–192.
6. Mester E, Mester AF, Mester A. The biomedical effects of
laser application. Lasers Surg Med 1985;5:31–39.
7. Pereira AN, Eduardo Cde P, Matson E, Marques MM.
Effect of low-power laser irradiation on cell growth and
pro-collagen synthesis of cultured fibroblasts. Lasers Surg
Med 2002;31:263–267.
8. Broadley C, Broadley KN, Disimone G, Riensch L, Da-
vidson JM. Low-energy helium-neon laser irradiation and
the tensile strength of incisional wounds in the rat. Wound
Repair Regen 1995;3:512–517.
9. Kaviani A, Djavid GE, Ataie-Fashtami L, et al. A ran-
domized clinical trial on the effect of low-level laser
therapy on chronic diabetic foot wound healing: a prelim-
inary report. Photomed Laser Surg 2011;29:109–114.
10. Beckmann KH, Meyer-Hamme G, Schroder S. Low level
laser therapy for the treatment of diabetic foot ulcers: a
critical survey. Evid Based Complement Alternat Med
2014;2014:626127.
LLLT ON DIABETIC PATIENTS
11. Yu W, Naim JO, Lanzafame RJ. Effects of photostimula-
tion on wound healing in diabetic mice. Lasers Surg Med
1997;20:56–63.
12. Aykol G, Baser U, Maden I, et al. The effect of low-level
laser therapy as an adjunct to non-surgical periodontal
treatment. J Periodontol 2011;82:481–488.
13. Diagnosis and classification of diabetes mellitus. Diabetes
Care 2011;34:62–69.
14. Armitage GC. Development of a classification system for
periodontal diseases and conditions. Ann Periodontol
1999;4:1–6.
15. Page RC, Eke PI. Case definitions for use in population-
based surveillance of periodontitis. J Periodontol 2007;78:
1387–1399.
16. Enwemeka CS. Standard parameters in laser phototherapy.
Photomed Laser Surg 2008;26:411–412.
17. Enwemeka CS. Intricacies of dose in laser phototherapy for
tissue repair and pain relief. Photomed Laser Surg 2009;
27:387–393.
18. Huang YY, Chen ACH, Carroll JD, Hamblin MR. Biphasic
dose response in low level light therapy. Dose Response
2009;7:358–383.
19. Jenkins PA, Carroll JD. How to report low-level laser
therapy (LLLT)/photomedicine dose and beam parameters
in clinical and laboratory studies. Photomed Laser Surg
2001;29:785–787.
20. WALT. Standards for the design and conduct of systematic
reviews with low-level laser therapy for musculoskeletal
pain and disorders. Photomed Laser Surg 2006;24:759–
760.
21. Silness J, Loe H. Periodontal disease in pregnancy. II.
Correlation between oral hygiene and periodontal condi-
tions. Acta Odontol Scand 1964;22:121–135.
22. Loe H, Silness J. Periodontal disease in pregnancy. I.
Prevalence and severity. Acta Odontol Scand 1963;21:533–
551.
23. Brown LJ, Oliver RC, Loe H. Periodontal diseases in the
U.S. in 1981: prevalence, severity, extent, and role in tooth
mortality. J Periodontol 1989;60:363–370.
24. Houreld NN, Sekhejane PR, Abrahamse H. Irradiation at
830 nm stimulates nitric oxide production and inhibits pro-
inflammatory cytokines in diabetic wounded fibroblast
cells. Lasers Surg Med 2010;42:494–502.
25. Ribeiro IW, Sbrana MC, Esper LA, Almeida AL. Evalua-
tion of the effect of the GaAlAs laser on subgingival scaling
and root planing. Photomed Laser Surg 2008;26:387–391.
26. Qadri T, Miranda L, Tunér J, Gustafsson A. The short-term
effects of low-level lasers as adjunct therapy in the treat-
ment of periodontal inflammation. J Clin Periodontol 2005;
32:714–719.
5
27. Makhlouf M, Dahaba MM, Tunér J, Eissa SA, Harhash TA.
Effect of adjunctive low level laser therapy (LLLT) on
nonsurgical treatment of chronic periodontitis. Photomed
Laser Surg 2012;30:160–166.
28. Dukić W, Bago I, Aurer A, Roguljić M. Clinical effec-
tiveness of diode laser therapy as an adjunct to non-surgical
periodontal treatment: a randomized clinical study. J Peri-
odontol 2013;84:1111–1117.
29. Pejcic A, Kojovic D, Kesic L, Obradovic R. The effects of
low level laser irradiation on gingival inflammation. Pho-
tomed Laser Surg 2010;28:69–74.
30. Obradović R, Kesić L, Mihailović, D et al. A histological
evaluation of a low-level laser therapy as an adjunct to
periodontal therapy in patients with diabetes mellitus. La-
sers Med Sci 2013;28:19–24.
31. Obradović R, Kesić L, Mihailović D, Jovanović G, Antić S,
Brkić Z. Low level lasers as an adjunct in periodontal
therapy in patients with diabetes mellitus. Diabetes Technol
Ther 2012;14:799–803.
32. Marques MM, Pereira AN, Fujihara NA, Nogueira FN,
Eduardo CP. Effect of low-power laser irradiation on pro-
tein synthesis and ultrastructure of human gingival fibro-
blasts. Lasers Surg Med 2004;34:260–265.
33. Hawkins D, Abrahamse H. Effect of multiple exposures of
low-level laser therapy on the cellular responses of woun-
ded human skin fibroblasts. Photomed Laser Surg 2006;
24:705–714.
34. Silveira PC, Streck EL, Pinho RA. Evaluation of mito-
chondrial respiratory chain activity in wound healing by
low-level laser therapy. J Photochem Photobiol B 2007;
86:279–282.
35. Posten W, Wrone DA, Dover JS, Arndt KA, Silapunt S,
Alam M. Low-level laser therapy for wound healing:
mechanism and efficacy. Dermatol Surg 2005;31:334–340.
Address correspondence to:
Ulku Baser
Department of Periodontology
Faculty of Dentistry
Istanbul University
34093 Capa
Istanbul
Turkey
E-mail: baserulku@hotmail.com
Received: March 18, 2016.
Accepted after revision: August 1, 2016.
Published online: November 17, 2016.