Quantifying Skin Stiffness in Graft-Versus-Host Disease, Morphea, and Systemic Sclerosis Using Acoustic Radiation Force Impulse Imaging and Shear Wave Elastography

There is a need for objective, sensitive, and reproducible methods of assessing skin sclerosis. We successfully adapted ultrasound-based, noninvasive quantifying tissue stiffness use on the skin: acoustic radiation force impulse (ARFI) imaging shear wave elasticity (SWEI) (Lee et al., 2015Lee S.J. Wolff D. Kitko C. Koreth J. Inamoto Y. Jagasia M. al.Measuring therapeutic response in chronic graft-versus-host disease. National Institutes Health consensus development project criteria clinical trials disease: IV. The 2014 Response Criteria Working Group report.Biol Blood Marrow Transplant. 2015; 21: 984-999Abstract Full Text PDF PubMed Scopus (180) Google Scholar). ARFI generates images micron-scale displacement, whereas SWEI measures speed at which sound waves travel across target (shear [SWS]) meters per second (Palmeri Nightingale, 2011Palmeri M.L. Nightingale K.R. Acoustic force-based methods.Interface Focus. 2011; 1: 553-564Crossref (140) Fibrotic, stiffer would be expected to have lower displacement relatively higher SWS than healthy, more compliant performed an Institutional Review Board?approved, pilot, cross-sectional, noninterventional study from July 2012 June 2016 (ClinicalTrials.gov identifier NCT01758250) evaluate utility after obtaining written informed consent healthy volunteers (HVs) patients with cutaneous disease, morphea, systemic In this study, we describe relationship between graded sclerosis scoring different diseases, anatomic regions, patient characteristics. obtained 805 measurements using 6 HVs, 11 2 (Table 1 Supplementary Table S1). Before imaging, each was evaluated by dermatologist expertise fibrosing disorders (ARC), site that measured received Clinical Sclerosis Score (CSS) ranging 0 3(Clements 1995Clements P. Lachenbruch Siebold White B. Weiner S. Martin R. al.Inter intraobserver variability total thickness score (modified Rodnan TSS) sclerosis.J Rheumatol. 1995; 22: 1281-1285PubMed Scholar; Skrzypek-Salamon 2018Skrzypek-Salamon A. Lis-?wi?ty Ranosz-Janicka I. Brzezi?ska-Wcis?o L. Localized Scleroderma Cutaneous Assessment Tool (LoSCAT) adult patients: report initial validation study.Health Qual Life Outcomes. 2018; 16: 185Crossref (7) A 230 (28.6%) were back, 184 (22.9%) arms, 97 (12%) legs, 103 (12.8%) thighs, 105 (13%) abdomen. rest other body sites: chest, sides trunk, ankles, hips, medial knee, elbows, hands, wrists, pannus.Table 1Measurements Tissue Displacement (ARFI, ?m) (in m/s) Distributed According CSS (Range = 0–3, 3 Being Severely Sclerotic)Sclerosis ScoreTissue CSS0 (Normal Skin)123HVGVHDMorpheaSScARFI N12416533827595438 Mean (SD)5.7 (1.3)5.8 (1.7)5.0 (1.6)5.0 (1.4)4.2 (1.7)3.6 (1.8)2.8 (1.8) Median5.75.64.95.33.83.22.2 Q1, Q34.9, 6.74.7, 6.93.9, 6.04.0, 5.92.9, 5.12.0, 4.81.3, 3.8 Range(2.7–8.8))(1.1–10.7)(1.2–11.5)(2.7–8.1)(2.0–9.7)(1.1–7.8)(0.9–7.2)SWS N12416433727595539 (SD)2.9 (0.8)3.2 (1.5)3.4 (1.2)3.4 (1.7)4.6 (1.7)6.4 (2.4) Median2.82.83.12.93.84.26.2 Q32.3, 3.52.3, 3.62.6, 4.02.4, 4.03.0, 4.93.1, 5.94.7, 8.7 Range(1.3–5.8)(1.3–10.1)(1.4–8.2)(1.8–6.9)(2.0–12.0)(2.1–9.6)(1.9–10.4)Abbreviations: ARFI, impulse; CSS, Score; GVHD, disease; HV, volunteer, first quartile; Q3, third SSc, sclerosis; SWS, speed.N indicates number sites imaged. Open table new tab Abbreviations: speed. N mean HVs’ 5.7 ?m, 2.9 m/s 1). negative correlation positive (Supplementary Figure S1a S2) univariate analysis (P < 0.001). multivariable regression model identify independent factors impact SWS. After adjusting sex, age, mass index (BMI), diagnosis type, region, average equal 1, 2, 1.00, 1.66, 2.45 ?m those normal appearing 0.001), respectively, 0.72, 1.26, 3.10 (all P respectively 2). Diagnosis BMI, region affected displacement. controlling morphea less HVs 1.33 0.005), suggesting clinically among HVs. With one-unit increase increased 0.04 type 0.047), minimal but statistically significant reduction increasing BMI. leg thigh 1.79 0.001) 0.66 0.009) type. Aside only factor abdomen arm significantly back 0.006) 0.58 0.007), 2) good excellent inter-rater (Shrout Fleiss, 1979Shrout P.E. Fleiss J.L. Intraclass correlations: uses rater reliability.Psychol Bull. 1979; 86: 420-428Crossref (16207) Scholar) two evaluators (intraclass coefficient 0.82, 95% confidence interval 0.67–0.90) 0.90, 0.8–0.95; S1b).Table 2Multivariable Regression Models Association Each Adjusting Diagnosis, Sex, Age, Anatomic RegionEffectARFISWSParameterEstimate (?m)95% CI1P-values determined simulation method multiple pairwise comparisons reference values as appropriate.P-valueParameterEstimate (m/s)95% appropriate.P-valueSclerosis 0Reference——Reference—— 1–1.00–1.41 –0.58<0.0010.720.33–1.110.001 2–1.66–2.09 –1.22<0.0011.260.86–1.66<0.001 3–2.45–2.97 –1.92<0.0013.102.62–3.58<0.001Diagnosis HVReference——Reference—— GVHD–0.18–0.78 0.410.8630.27–0.27 0.810.590 Morphea–1.33–2.07 –0.580.0050.56–0.11 1.230.236 SSc–0.92–1.83 –0.010.1230.10–0.72 0.920.988Sex MaleReference——Reference—— Female0.47–0.14 1.080.1240.02–0.53 0.570.948Age0.00–0.02 0.010.5010.00–0.01 0.010.712BMI0.040.00 0.080.047–0.01–0.05 0.020.463Anatomic BackReference——Reference—— Abdomen0.35–0.16 0.860.558–0.66–1.06 –0.270.006 Arm–0.18–0.65 0.290.923–0.58–0.93 –0.230.007 Thigh–0.66–1.16 –0.160.044–0.41–0.79 –0.020.153 Leg–1.79–2.30 –1.28<0.0010.11–0.28 0.500.980 Other–0.92–1.57 –0.280.0260.22–0.26 0.700.858Abbreviations: index; CI, interval; volunteer; speed.The pseudo R2 models are 0.52 0.43, respectively.1 P-values appropriate. respectively. outcome limitations detection early prediction outcome, or evaluation therapy (Aki 2016Aki S.Z. Carpenter P.A. Storer B.E. Sandmaier B.M. Lee al.Confounding affecting (NIH) disease organ-specific global severity.Bone 2016; 51: 1350-1353Crossref (10) Baird 2013Baird K. Steinberg S.M. Grkovic Pulanic Cowen E.W. Mitchell S.A. al.National staging severely organ correlate established indicators severity prognosis.Biol 2013; 19: 632-639Abstract (62) Kelsey Torok, 2013Kelsey C.E. Torok K.S. Tool: responsiveness change pediatric population.J Am Acad Dermatol. 69: 214-220Abstract (60) Inter-rater agreement can poor, even specialists (Cardones 2019Cardones A.R. Sullivan K.M. Green Chao N.J. Rowe-Nichols Bañez L.L. al.Interrater reliability grading graft-vs-host disease.JAMA 2019; 155: 833-837Crossref (3) Clements Therefore, great exists quantifiable evaluating sclerotic S.Y. Cardones Doherty Palmeri Preliminary results feasibility ARFI/SWEI assess diseases.Ultrasound Med Biol. 41: 2806-2819Abstract (50) High-frequency ultrasound, magnetic resonance optical coherence tomography and/or elastography shown some promise substantial detecting active resolution cost, depth, (Abignano 2013Abignano G. Aydin Castillo-Gallego Liakouli V. Woods Meekings al.Virtual biopsy tomography: quantitative biomarker scleroderma.Ann Rheum Dis. 72: 1845-1851Crossref (64) Du 2016Du Liu C.H. Lei Singh Li Hicks M.J. al.Rapid, quantitation elastography.J Biomed Opt. 46002Crossref (21) Nezafati 2011Nezafati K.A. Cayce R.L. Susa J.S. Setiawan A.T. Tirkes T. Bendeck S.E. al.14-MHz ultrasonography measure (localized scleroderma).Arch 147: 1112-1115Crossref (25) Shahidi-Dadras 2018Shahidi-Dadras Abdollahimajd F. Jahangard Javinani Ashraf-Ganjouei Toossi Magnetic linear treated methotrexate high-dose corticosteroid.Dermatol Res Pract. 2018: 8391218Crossref Measuring through inaccurate when compared histologic does not necessarily reflect activity (Nezafati organ’s mechanical properties rather just relative density (Doherty 2013Doherty J.R. Trahey G.E. diagnostic ultrasound.IEEE Trans Ultrason Ferroelectr Freq Control. 60: 685-701Crossref (201) Controlling required our Ultrasound-based technology allows larger field view tomography, ability focus refined spatial interrogation. These techniques noninvasive, portable, quick perform. As demonstrated pilot there differences sensitivity SWEI, these likely biochemical aspects technique. Our limited sample size lack correlation. also did control status treatment patients. Nevertheless, promising stiffness. It remains seen whether one combination ultimately yield most useful relevant information outcomes. group intends address current future studies. Datasets related article found https://doi.org/10.7924/r4h995b4q, hosted Duke Research Data Repository. Adela Cardones: http://orcid.org/0000-0001-7489-2046 Russell Hall III: http://orcid.org/0000-0001-7621-4935 Keith Sullivan: http://orcid.org/0000-0002-1379-9216 Joanna Hooten: http://orcid.org/0000-0002-6451-7012 Seung Yun Lee: http://orcid.org/0000-0001-7091-2441 Beiyu Liu: http://orcid.org/0000-0002-6706-9729 Cynthia Green: http://orcid.org/0000-0002-0186-5191 Nelson Chao: http://orcid.org/0000-0001-6725-7220 Krista Rowe Nichols: http://orcid.org/0000-0003-2722-8972 Lionel Bañez: http://orcid.org/0000-0002-2360-200X Ankoor Shah: http://orcid.org/0000-0002-6315-8089 Nicole Leung: http://orcid.org/0000-0001-7565-1009 Mark Palmeri: http://orcid.org/0000-0001-8995-0050 ARC unpaid consultant MicroElastic Ultrasound Systems. She principal investigator investigator-initiated grant Pfizer. AS has research support Boehringer Ingleheim Reata Pharmaceuticals. MLP Chief Scientific Advisor Systems receives in-kind technical Siemens Medical University inventors intellectual property associated been licensed Medical, Samsung, This funded Dermatology Foundation (Career Development Award ARC) Duke?Coulter Grant. funders had no role conceptualization, design, data collection, analysis, decision publish, preparation manuscript. Conceptualization: ARC, RPH III, KMS, NJC, MLP; Curation: SYL, NL, BL; Formal Analysis: CLG, BL, LLB, Fund Acquisition: Investigation: JH, KRN, Methodology: RPH, Project Administration: KRN; Supervision: Writing - Original Draft Preparation: ARC; Editing: AS, S1Demographic Information PatientsDemographic InformationTotal (N 25)Diagnosis HV6 (24%) GVHD6 Morphea11 (44%) SSc2 (8%)Sex Female19 (76%) Male6 (24%)Age (y) N25 (SD)47.1 (17.8) Median54.0 Q329.0, 62.0 Range(20.0–72.0)BMI (kg/m2) (SD)30.2 (7.4) Median28.3 Q325.7, 34.2 Range(19.8–54.8)Abbreviations: S2Univariable Displacements (ARFI SWS) ScoreEffectARFISWSParameterEstimate values.P-valueParameterEstimate values.P-valueSclerosis 1–0.99–1.44 –0.54<0.0010.720.35–1.10<0.001 2–1.73–2.19 –1.27<0.0011.300.92–1.69<0.001 3–2.61–3.15 –2.06<0.0013.182.73–3.63<0.001Abbreviations: 0.16 0.24, values.