High-frequency ultrasound measurement for assessing post-thrombotic syndrome and monitoring compression therapy in chronic venous disease
Article Outline
- Abstract
- Materials and methods
- Results
- Discussion
- Conclusion
- Author contributions
- Acknowledgment
- References
- Copyright
Background
The purpose of this study was to validate high-frequency ultrasound (HFU) measurement of dermal thickness for quantification of edema in patients with different severities of chronic venous disease.
Methods
HFU measurements of dermal thickness were made with a 17-MHz probe (Philips iU22 Ultrasound scanner, Bothell, Wash) or a 20-MHz medium-focus probe (DermaScan-C, Cortex Technology, Denmark), 7.5 cm above the medial malleolus. For validation, 20 patients with venous leg ulcers who were not receiving compression therapy, 20 patients with previous deep vein thrombosis (DVT) and symptoms of post-thrombotic syndrome (PTS) without ulceration, and 31 age-matched healthy controls were measured on a single occasion. To investigate the effect of compression on dermal thickness, the leg ulcer patients from the validation study were treated with compression therapy for 7 weeks and measured after 1, 3, 5, and 7 weeks. The association between dermal thickness and the clinical (C) component of the CEAP classification was examined in a cross-sectional analysis of 157 patients with a confirmed history of DVT ≥3 years ago.
Results
Dermal thickness in patients with venous leg ulcers before compression therapy (median, 2.56 mm; interquartile range [IQR], 2.31-2.82 mm) was significantly greater (P = .002) than that in patients with symptoms of PTS without ulceration (median, 2.16 mm; IQR, 1.90-2.36 mm). Dermal thickness in both groups was significantly greater (P < .0001) than the control group (median, 1.34 mm; IQR, 1.29-1.44 mm). Compression therapy caused a steady and significant decrease in dermal thickness during the first 5 weeks until normal control levels were achieved. Dermal thickness increased with increasing CEAP category. In 121 patients with a positive diagnosis of DVT ≥3 years ago from Radiology Department records, a hypothetical test cutoff of 1.985 mm for the prediction of severe PTS noted as C4b, C5, and C6 (lipodermatosclerosis or leg ulceration) had a positive predictive value of 46.9% and a negative predictive value of 90.3%.
Conclusion
HFU measurement of dermal thickness enables the monitoring of edema reduction by compression therapy. A prospective study is required to determine the temporal dynamics of dermal thickness changes after DVT and the relationship to the development of PTS. This test has the potential to be beneficial in the follow-up of patients after a DVT and provide clinical evidence for using graduated elastic compression stockings to control edema and prevent the development of more advanced skin changes.
Objective methods for early detection of skin changes in chronic venous disease are not readily available; similarly, means for assessing the effect of compression therapy are limited. Assessment is frequently based on the gross appearance of subjective skin changes, such as edema, pigmentation, and induration, and in the case of venous ulceration, monitoring changes in ulcer size.
In patients with venous insufficiency, increased venous pressure is transmitted to the capillary bed, increasing hypostatic pressure and predisposing to edema.1 Edema is frequently the first clinical sign of venous insufficiency but may be hard to detect visually or by palpation in the early stages. The severity of edema, from slight to very marked, is often reported on a subjective 4-point scale.2 Existing methods of edema quantification in venous disease such as leg circumference measurements and leg volume method for measurement of edema are insensitive, and only relative changes can be determined. A sensitive, quantitative measure of edema would be a valuable tool to assess the early development of skin changes in patients who have had a previous deep vein thrombosis (DVT) or to assess the effect of compression therapy.
High-frequency ultrasound (HFU) imaging provides a noninvasive means to quantify edema, because edema produces an increase in dermal thickness3 and HFU produces high-resolution images of the dermis that allow accurate measurements.4, 5, 6, 7 A previous study used dermal thickness assessed by HFU to quantify edema in venous leg ulceration.8 Other studies have shown a relationship between dermal echogenicity and chronic venous disease,9, 10, 11 but this parameter requires careful calibration and is more difficult to calculate than the relatively simple measurement of dermal thickness.
The purpose of this study was to validate HFU measurement of dermal thickness as a tool to quantify edema in patients with different severities of chronic venous disease; specifically, patients with chronic venous leg ulcers, patients with other symptoms of post-thrombotic syndrome (PTS), and healthy controls. Two potential clinical applications of the technique were then investigated: (1) The effect of compression therapy on dermal thickness in patients with chronic venous leg ulcers, and (2) the relationship between dermal thickness and the clinical (C) component of the CEAP classification of venous disease in patients with a history of DVT.
Materials and methods
This study was approved by the South Metropolitan Area Health Service Human Research Ethics Committee. All individuals who participated in this study were recruited at Fremantle Hospital, and all participants provided written informed consent.
Three separate studies were performed to evaluate dermal thickness as a tool for the assessment of chronic venous disease. These consisted of:
Validation study
This study assessed three groups: patients with chronic venous ulcers, patients with signs of PTS but without ulceration, and control participants.
The chronic venous ulcer group comprised 20 patients with an ulcer on the leg in the gaiter region, with confirmation of chronic venous disease by photoplethysmography (venous refilling time <25 seconds) and exclusion of arterial disease by Doppler ankle-brachial index (ABI) >0.8. This group is C6 on the CEAP classification.
The PTS group comprised 20 patients with a history of DVT of the leg ≥3 years previously, confirmed by DU venous scanning or venography, and one or more of the following symptoms: lower limb edema, hyperpigmentation, venous eczema, or lipodermatosclerosis. Individuals were excluded from this group if they had a history of leg ulceration. This group was C3, C4a, and C4b on the CEAP classification.
The control group of 31 individuals, with a similar age distribution to the PTS and venous ulcer groups, was recruited from employees and volunteers at Fremantle Hospital. These individuals had no clinical evidence of varicose veins or telangiectases, no history of DVT or chronic venous leg ulceration, and no evidence of venous disease on photoplethysmography (venous refilling time >25 seconds). Arterial disease was excluded (Doppler ABI >0.8).
Individuals with a medical history of diabetes, rheumatoid arthritis, and renal or cardiac problems were excluded from all groups.
Compression therapy study
After the validation study, the 20 patients with venous leg ulcers were entered into the compression therapy study. The dermal thickness was measured, and then these patients commenced compression therapy for the treatment of their venous ulcer. A simple nonadherent dressing was applied to the wound, and the compression therapy consisted of a layer of padding with orthopedic wool (Velband; Smith & Nephew, Hull, United Kingdom), two inelastic compression bandages (Comprilan, Smith & Nephew), and a retention layer of elasticized tubular bandage (Tubigrip, ConvaTec, Victoria, New South Wales, Australia). Bandages and primary dressings were changed two to three times weekly, depending on the amount of exudate. The dermal thickness was remeasured after 1, 3, 5, and 7 weeks of compression therapy.
CEAP classification study
A total of 157 individuals were entered into the study, comprising 121 patients with a history of DVT ≥3 years previously, confirmed by duplex venous scan or venography, who were identified through the Fremantle Hospital Radiology Department database, and 36 with DVT and venous ulceration who were identified through the Fremantle Hospital Vascular Research Laboratory. Patients were invited to attend a study visit at which the CEAP classification was determined by clinical examination.12 The study excluded individuals who were unable to attend a study visit or who had malignant disease.
Scanning protocol
HFU imaging was performed with using a 17-MHz linear probe on a B-mode diagnostic ultrasound scanner (Phillips iU22, Bothell, Wash) or with a 20-MHz medium-focus probe on a portable B-mode scanner (DermaScan C, Cortex Technology, Denmark).
The HFU assessment of the leg was performed with the patient in a side position. Three independent transverse measurements were taken at 7.5 cm above the medial malleolus, and the mean of the three measurements was used in subsequent analysis. Dermal thickness was measured between the surface of the epidermis and the interface of the dermis with the subcutaneous tissue layer. Measurements with the Phillips iU22 scanner consisted of the dermal thickness at a single point (Fig 1). Measurements with the DermaScan-C consisted of the mean of 224 A-mode scans along a 12.1-mm continuous scanning length (Fig 2).
Fig 1.
Image of dermal thickness measurement obtained using the Phillips iU22 scanner.
Fig 2.
Image of dermal thickness measurement obtained using the DermaScan-C scanner.
Statistical analysis
Nonparametric statistics were used to analyze the data due to the nonsymmetric distribution of dermal thickness in the group with venous leg ulcers. Pair-wise comparisons of dermal thickness between the three groups in the validation study (venous leg ulcer, PTS, and healthy controls) were tested for significance using a two-sided Mann-Whitney U test. For individuals in the compression therapy study, pair-wise comparisons of dermal thickness between compression treatment time points were tested for significance using the two-sided Wilcoxon signed rank test for paired samples.
In the CEAP classification study, two stages of analysis took place. First, dermal thickness levels were compared among the CEAP categories using the Jonckheere-Terpstra test for a priori ordered categories.13 Second, pair-wise comparisons of dermal thickness between categories C0, C1, C2, and C3 and the healthy controls from the validation study were tested for significance using a two-sided Mann-Whitney U test.
To test whether gender and type of scanner were significant sources of variation, regression analysis was performed using the R statistical package (R Foundation for Statistical Computing, Vienna, Austria).14 Gender and disease group were fitted as factor variables in the validation study, and CEAP category and scanner were fitted as factor variables in the CEAP classification study.
To determine the appropriate cutoff value of dermal thickness for a test for severe post-thrombotic outcomes (≥C4b) based on data from the CEAP classification study, receiver-operating characteristic (ROC) curve analysis was performed using SPSS 17.0 software (SPSS Inc, Chicago, Ill). The ROC curve is a representation of the trade-off between sensitivity and specificity, and the area under the curve (AUC) is a measure of the test's discriminatory power (0.5 = chance response; 1.0 = perfect test).15 The level of significance for all statistical tests was P = .05. The Bonferroni correction was applied when multiple pair-wise tests were performed.
Results
All significant results presented were significant after the Bonferroni correction was applied.
Validation study
Table I summarizes the demographics of individuals in the validation study. As expected because of age-matching, the difference in age between the three groups was not statistically significant. The control group contained a higher proportion of women than the other two groups, and gender was examined as a potential source of variation. When disease group and gender were fitted as factor variables in regression analysis, the effect of gender on dermal thickness was small (estimated effect size, −0.017 mm) and not significant (P = .07).
Table I. Demographics of participants in the validation study
| Group | No. | Gender | Age, median (IQR), y | |
|---|---|---|---|---|
| M | F | |||
| Post-DVT (without ulceration) | 20 | 15 | 5 | 69 (64-76) |
| Venous ulcer | 20 | 12 | 8 | 72(56-78) |
| Healthy controls | 31 | 13 | 18 | 65(50-74) |
The median dermal thickness in the control group was 1.34 mm (interquartile range [IQR], 1.30-1.44 mm). The PTS group without ulceration showed significantly increased dermal thickness compared with the controls (P < .0001), with a median of 2.16 mm (IQR, 1.90-2.36 mm). Dermal thickness in patients with venous leg ulcers was 2.56 mm (IQR, 2.31-2.82 mm), which was significantly higher than healthy skin (P < .0001) and also significantly higher than the PTS group without ulceration (P < .002; Fig 3).
Fig 3.
Bars show the comparison of dermal thickness between the three groups in the validation study (Mann-Whitney U-test). IQ, Interquartile; PTS, post-thrombotic syndrome.
Compression study
These patients were the venous ulcer group whose demographics are summarized in Table I. After 1 week of compression bandaging, the dermal thickness in the venous ulcer group (median, 2.56 mm; IQR, 2.31-2.83 mm) reduced significantly (P < .0002; median, 1.82 mm; IQR, 1.64-2.14 mm; Fig 4). Dermal thickness continued to decrease significantly from week 1 to week 3 (P < .0004) and from week 3 to week 5 (P = 0.002), at which point dermal thickness (median, 1.39 mm; IQR, 1.23-1.45 mm) was similar to the normal control levels in the validation study.
Fig 4.
Effect of compression therapy on dermal thickness in patients with venous leg ulceration. IQ, Interquartile; Normal, Median dermal thickness of healthy controls in the validation study.
CEAP classification study
All CEAP groups contained participants who were measured by the two scanners. When CEAP classification and scanner were fitted as factor variables in regression analysis, the effect of scanner was not significant (P >.10).
Table II summarizes the CEAP classification profile and demographics of participants in the CEAP classification study. There was an increasing trend in dermal thickness with increasing CEAP category (Fig 5), and this increase was highly significant using the Jonckheere-Terpstra test for ordered categories (P < .0001). When categories C0, C1, C2, and C3 were compared, pair-wise with the healthy controls from the validation study (no history of DVT), C2 and C3 were each significantly different from healthy controls (P = .0008 and P < .0002, respectively), but not C0 or C1.
Table II. CEAP classification profile and demographics of participants in the CEAP classification study
| CEAP category | Clinical features | No. (%) | Gender | Age Median (IQR) y | |
|---|---|---|---|---|---|
| M | F | ||||
| C0 | No visible or palpable venous disease | 9(5.7) | 3 | 6 | 61(56-61) |
| C1 | Telangiectasias | 10(6.4) | 5 | 5 | 54(38.5-68.5) |
| C2 | Varicose veins | 5(3.2) | 1 | 4 | 71(60-71) |
| C3 | Edema | 29(18.5) | 9 | 20 | 65(60-73) |
| C4a | Hyperpigmentation, eczema | 41(26.1) | 18 | 23 | 69(61-76) |
| C4b | Lipodermatosclerosis | 13(8.3) | 12 | 1 | 68(61-74) |
| C5 | Healed leg ulcers | 27(17.2) | 16 | 11 | 69(57.5-78) |
| C6 | Active (open) leg ulcers | 23(14.6) | 17 | 6 | 62(49.5-70.5) |
| Total | 157(100) | 81 | 76 | ||
Fig 5.
Dermal thickness of subjects classified according to the clinical component of the CEAP classification. IQ, Interquartile; Normal, dermal thickness of healthy controls in the validation study.
To test the clinical utility of a test based on dermal thickness to predict patients with severe post-thrombotic outcomes documented as C4b-6(lipodermatosclerosis, healed or active leg ulcers), we used ROC curves to analyze the data from the 121 patients with a positive diagnosis of DVT ≥3 years ago from Radiology Department records. These patients were a subgroup of the 157 patients in the CEAP classification study. We assumed that the increase in dermal thickness occurs as a result of the DVT and pre-dates the development of skin changes. A ROC curve constructed from the dermal thickness measurements was significant (P < .0001) for the prediction of severe post-thrombotic outcomes (≥C4b), with an AUC of 0.751 (95% confidence interval [CI], 0.655-0.848), which indicates a test of fair clinical utility. The optimal cutoff value of 1.985 mm was selected as that which gave the highest sensitivity (sensitivity, 76.7%; 95% CI, 57.3%-89.4%), with a specificity of ≥70% (specificity, 71.4%; 95% CI, 60.8-80.2%). Of the 46 patients with a dermal thickness >1.985 mm by positive test result, 46.9% (95% CI, 32.8%-61.6%) had clinical evidence of severe PTS, resulting in a positive predictive value of 46.9%. Of the 75 of patients who had a negative test result, 9.7% had severe PTS, for a negative predictive value of 90.3% (95% CI, 80.4%-95.7%).
Discussion
Two different HFU scanners were used to enable a high throughput of patients, and analysis confirmed that the type of scanner was not a significant source of variation in this study. In practice, a range of suitable scanners could be used, and all would require further validation and the development of standard operating procedures before use in monitoring and diagnostic testing.
The validation study confirmed that dermal thickness measurements obtained by HFU imaging are potentially useful for discriminating between individuals with different severities of venous disease. The range of normal values was narrow, and the venous ulcer group before compression therapy showed consistently high dermal thickness. The dermal thickness in the patients in the PTS group without a history of ulceration was between these two extremes.
The standard method of treatment for healing ulcers and preventing ulcer recurrence is compression therapy. With the compression system used in the study, applied by experienced practitioners, dermal thickness measured by HFU imaging reduced to normal levels during the first 5 weeks of treatment. The median dermal thickness increased by a small amount at week 7 in this study, but this difference was not significant.
Because the principle of compression therapy is to reduce edema, HFU imaging provides the opportunity to make direct measurements of the effectiveness of this mechanism. In clinical practice it is very difficult to establish in the short-term whether the level of compression is appropriate for a given patient, whether the compression has been adequately applied, and whether it is having any benefit. We propose that serial measurements of dermal thickness by HFU imaging can be used to monitor the effectiveness of compression therapy and allow adjustments in the level of compression to be made throughout the entire course of treatment. Serial measurements of dermal thickness during compression treatment would require access to a portable HFU device in the clinic.
Further studies are required to demonstrate (1) the effect of monitoring dermal thickness during the course of compression therapy on ulcer healing rates, and hence, (2) cost-effectiveness. HFU imaging may also be a useful tool to assess other venous interventions, for example, perforator surgery.
Previous studies have indicated that skin changes and chronic venous ulceration will develop in a proportion of patients after DVT. The incidence of ulceration after a DVT in previous studies has been evaluated at between 5% and 10% of patients.16 A Cochrane systematic review concluded that compression stockings reduce the occurrence of symptoms of PTS, including leg ulceration, by >50%.17 The most recent edition of the “American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition)” now recommends compression stockings alongside conventional anticoagulation (grade 1A recommendation) in all patients with acute symptomatic proximal DVT for the prevention of PTS, where feasible.18 Early identification of skin changes in patients after DVT will assist the identification of patients for whom compression therapy may prevent the development of severe PTS.
Most PTS symptoms have been shown to occur ≤2 years of DVT.19 The inclusion of patients with a DVT at ≥3 years previously in this study ensured that symptoms had reached the highest level of CEAP categorization at the time of the analysis. Our results demonstrate an increase in dermal thickness with increasing CEAP category and confirm that edema is quantitatively more severe in the higher CEAP categories. Dermal thickness was also significantly higher in the C2 category (varicose veins) than in healthy controls, one level lower than the C3 level of clinically detectable edema. The information gained from a test based on a 1.985 mm cutoff for severe PTS (positive predictive value of 46.9%) would be a powerful incentive for the application of graduated compression stockings. No means are currently available to predict the 5% to 10% of individuals who will present with severe PTS after DVT.20 According to these data, 53.1% of patients identified by the test would not go on to develop severe post-thrombotic changes of lipodermatosclerosis or ulceration. A proportion of patients, however, would still be likely to benefit from wearing compression stockings with the reduction of other signs and symptoms of PTS, such as pain, edema, hyperpigmentation, and venous eczema. In post-DVT patients a once-off scan could be performed using a suitable hospital diagnostic scanner (≥17 MHz) at an interval after DVT (to be defined in a prospective study), and the cost of this scan would be small compared with the savings from reduced incidence of leg ulceration and other symptoms of PTS.
Some of the symptomatic individuals in the CEAP classification study were bandaged or were wearing compression stockings at the time of the study. Because compression reduces dermal thickness, this will have reduced the ability of the dermal thickness measurements to discriminate between adjacent CEAP categories. We argue that the positive and negative predictive values of a test based on a cutoff value of dermal thickness would be improved if this variable had been controlled. Also, the study does not show at what time after a DVT the increases in dermal thickness are predictive of the development of serious skin changes. Prospective studies will be required to determine the dynamics of dermal thickness changes after DVT, the relationship to the development of PTS, and accurate cutoff values for clinical application.
Conclusion
Dermal thickness measured by HFU imaging is a sensitive and quantitative method for evaluating dermal edema in patients with venous disease and enables the monitoring of edema reduction by compression therapy in the treatment of chronic venous leg ulcers. A prospective study is required to determine the temporal dynamics of dermal thickness changes after DVT and the relationship to the development of PTS. This noninvasive and objective test has the potential to be beneficial in the follow-up of patients after a DVT and could provide good clinical evidence for using graduated elastic compression stockings to control edema and prevent the development of more advanced skin changes.
Author contributions
The authors wish to acknowledge Sue Hoskin, Lorraine Linacre, and Gail Brunt for their valuable clinical support in the Vascular Research Laboratory, Fremantle Hospital; Dr Dermot Kearney for providing access to the ultrasound facility and electronic databases in the Fremantle Hospital Radiology Department; Clifford Brouwer for sonography support in the Fremantle Hospital Radiology Department; Michael McDonnell for assistance with the Fremantle Hospital Radiology Department electronic database; and Peter Henderson for providing information technology and database support. Nicole Warrington of the Western Australian Institute for Medical Research (Laboratory for Genetic Epidemiology) is acknowledged for her statistical advice, and appreciation is expressed to Gunnar Svendsen and Cortex Technology (Denmark) for the generous loan of the DermaScan-C ultrasound scanner.
References
- . Peripheral edema. Am J Med. 2002;113:580–586
- . Examination and history taking. Philadelphia: J.B. Lippincott; 1991;
- . Influence of body water distribution on skin thickness: measurements using high-frequency ultrasound. Br J Dermatol. 2001;144:947–951
- . The determination of skin thickness using conventional diagnostic ultrasound equipment. Clin Exp Dermatol. 1984;9:280–285
- . Echographic assessment of age dependent variations of skin thickness: a study on 162 subjects. Skin Res Technol. 1995;1:81–85
- Echographic measurement of skin thickness in adults by high frequency ultrasound to assess the appropriate microneedle length for intradermal delivery of vaccines. Vaccine. 2007;25:6423–6430
- . Assessment of an ultrasonic dermal scanner for skin thickness measurements. Med Eng Phys. 2008;30:804–807
- . Dermal oedema assessed by high frequency ultrasound in venous leg ulcers. Br J Dermatol. 1998;138:815–820
- . Dermal oedema in lipodermatosclerosis: distribution, effects of posture and compressive theraphy evaluated by high-frequency ultrasonography. Acta Derm Venereol. 1995;75:120–124
- . Assessment of dermal water by high-frequency ultrasound: comparative studies with nuclear magnetic resonance. Br J Dermatol. 1996;135:218–224
- . Age-related diurnal changes of dermal oedema: evaluation by high-frequency ultrasound. Br J Dermatol. 1994;131:849–855
- Revision of the CEAP classification for chronic venous disorders: consensus statement. J Vasc Surg. 2004;40:1248–1252
- . Non-parametric statistical methods. 2nd ed. New York: John Wiley; 1999;202-12
- . R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; 2006;http://www.R-project.org
- . The interpretation of diagnostic tests. Stat Methods Med Res. 1999;8:113–134
- . Relationship between deep venous thrombosis and the postthrombotic syndrome. Arch Inter Med. 2004;164:17–26
- . Non-pharmaceutical measures for prevention of post-thrombotic syndrome. Cochrane Database Syst Rev. 2004;1:CD004174
- . Antithrombotic therapy for venous thromboembolic disease: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th edition). Chest. 2008;133(suppl):454–545S
- . Post-thrombotic syndrome: prevalence, prognostication and need for progress. Br J Haematol. 2009;145:286–295
- Predictors of the post-thrombotic syndrome during long-term treatment of proximal deep vein thrombosis. J Thromb Haemost. 2005;3:718–723
This study was supported by grants from the Australian National Health and Medical Research Council, the Fremantle Hospital Medical Research Foundation, and the Western Australian State Health Research Advisory Council.
Competition of interest: none.
PII: S0741-5214(09)01196-3
doi:10.1016/j.jvs.2009.05.060
© 2009 Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
