Key Points
Question
Can multispectral optoacoustic tomography be used to distinguish between arteriovenous and venous vascular malformations by quantitative analysis of oxygenated and deoxygenated hemoglobin?
Findings
In this pilot study of 12 patients with arteriovenous or venous vascular malformations, multispectral optoacoustic tomography showed significant differences between types of vascular malformations in the ratio of oxygenated to deoxygenated hemoglobin. Moreover, multispectral optoacoustic tomography enabled quantitative assessment of therapeutic response after embolization or sclerotherapy.
Meaning
Multispectral optoacoustic tomography may become a promising and versatile tool for noninvasive diagnosis and monitoring of vascular malformations before and after treatment.
This pilot study evaluates the use of multispectral optoacoustic tomography for assessment of biomarkers among patients with arteriovenous and venous malformation.
Abstract
Importance
Differential diagnosis of congenital vascular anomalies is challenging, and misdiagnosis is frequent. Vascular malformations are considered one of the most difficult vascular diseases to treat. A new imaging approach that visualizes anatomical features and quantitatively assesses molecular biomarkers noninvasively would aid diagnosis and monitoring of treatment response of vascular malformations.
Objective
To evaluate multispectral optoacoustic tomography (MSOT) for noninvasive assessment of molecular biomarkers for diagnosis and therapeutic monitoring of vascular malformations.
Design, Setting, and Participants
This pilot study examined 6 patients with arteriovenous malformation (AVM) and 6 patients with venous malformation (VM) diagnosed according to the classification system of the International Society for the Study of Vascular Anomalies. All patients underwent clinical hybrid MSOT/ultrasonographic (US) imaging before and after treatment at an interdisciplinary vascular malformations clinic by trained MSOT and US examiners. Examiners were blinded to the patient history and stage of disease. Data were collected from April 11 to 25, 2017, and analyzed from May 1 to October 31, 2017.
Interventions
Clinical hybrid MSOT/US imaging was performed before or within 1 week after endovascular embolization (for AVM) or percutaneous sclerotherapy (for VM).
Main Outcomes and Measures
Region-of-interest analysis of the lesion and contralateral healthy tissue revealed quantitative values for oxygenated (HbO2) and deoxygenated (HbR) hemoglobin by spectral unmixing of optoacoustic data acquired at multiple wavelengths. The HbO2:HbR ratio was calculated for healthy tissue and for AVM and VM before and after treatment.
Results
Twelve patients (9 female and 3 male; mean [SD] age, 23 [18] years; age range, 6-59 years) with vascular malformations (6 with AVMs and 6 with VMs) were included. Significantly higher HbO2:HbR ratios for AVMs (mean [SEM], 1.82 [0.08] vs 0.89 [0.03]; P < .001) and for VMs (mean [SEM], 1.12 [0.04] vs 0.89 [0.03]; P = .001) were found on MSOT tissue compared with healthy tissue. Significantly higher HbO2:HbR ratios for AVMs compared with VMs (mean [SEM], 1.82 [0.08] vs 1.12 [0.04]; P < .001) were also found. Therefore, MSOT provided intrinsic biomarker patterns to distinguish both vascular malformations. After therapy, the HbO2:HbR ratio dropped in correlation to treatment success validated by magnetic resonance imaging or angiography.
Conclusions and Relevance
This study suggests that different types of vascular malformations are clearly distinguished by MSOT-based, noninvasive assessment of hemoglobin levels in vascular malformations. The therapy effects found in this study could be instantly visualized, and this may offer a new tool for noninvasive diagnosis and monitoring of vascular malformations.
Introduction
Vascular malformations constitute a broad spectrum of angiogenic and vasculogenic disorders affecting children and adolescents. They are classified into low-flow (including venous malformation [VM]) and fast-flow (arteriovenous malformation [AVM]) malformations.1 Vascular malformations are always present at birth, never regress spontaneously, and can be asymptomatic for a long time.1 However, with increasing growth, sometimes triggered by mechanical or hormonal influence, they can cause pain, extensive functional impairment, ischemia, or even cardiovascular failure.2 The therapeutic strategy differs fundamentally based on the underlying type of malformation. Although misdiagnosis is common, consecutive false therapeutic interventions can trigger rapid disease progression.1 Therefore, correct classification of vascular malformations according to the International Society for the Study of Vascular Anomalies scheme3 is crucial for patient outcome. Current diagnostic imaging approaches include ultrasonography (US) and magnetic resonance imaging. Both techniques address only anatomical information but do not provide tissue-specific biomarkers. For safe lesion classification, those additional biomarkers are needed; moreover, biomarkers will enable immediate monitoring of therapy. Hence, with respect to the young patient age and necessity of repetitive monitoring, an easy-to-use and noninvasive clinical diagnostic approach is needed not only to address anatomical information but also functional biomarkers. Multispectral optoacoustic tomography (MSOT) is based on the photoacoustic effect. Absorption of pulsed, high-energy light causes thermoelastic expansion of the target tissue, inducing the transmission of US waves that can be detected and used for image reconstruction. Owing to specific absorber properties of intrinsic biomarkers like hemoglobin, melanin, or lipids, image reconstruction can quantitatively reveal their distribution within the tissue of interest. Preclinical models of disease and the initial clinical applications of MSOT have shown the potential of the modality.4,5,6,7,8,9,10,11 With MSOT, we present a novel noninvasive approach that provides quantitative imaging biomarkers for diagnosis and monitoring of therapy for vascular malformations.
Methods
Patients and Study Design
In this preliminary cross-sectional study, we included 12 patients ranging in age from 6 to 59 years with vascular malformations (6 with AVM and 6 with VM). Patients with AVM underwent analysis with regard to the status of embolotherapy with ethylene vinyl alcohol copolymer. Two patients with VM were investigated before and after successful sclerotherapy with polidocanol. Data were collected from April 11 to 25, 2017. This study protocol was reviewed by Ethikkommission der Westfälischen Wilhelms Universität, and all patients, or parents if the child is a minor, gave written and informed consent before study enrollment.
Technical Aspects of MSOT Imaging Device
The technical details of the clinical MSOT imaging system (MSOT Acuity Echo, iThera Medical GmbH) have been described in detail elsewhere.4,6 The pulse energy was attenuated to ensure adherence with American National Standards Institute limits of maximum permissible exposure. Maximum penetration depth of the laser was 3 cm. Reflection US computed tomographic mode images were generated accordingly.
MSOT Image Acquisition
Examiners experienced in US and optoacoustic scanning performed MSOT/US and were blinded to patient history and clinical data. The scanner probe was placed centered above the vascular malformation and at the equivalent region on the contralateral healthy side. The MSOT images were acquired at 6 different wavelengths ranging from 700 to 850 nm. Regions of interest were drawn in consensus by 2 examiners (M.M. and A.H.) based on the US image in the upper half of anomalous vasculature or in equivalent depth in healthy tissue. Individual contributions of oxygenated (HbO2) and deoxygenated (HbR) hemoglobin were calculated from acquired data based on their spectral absorption characteristics after spectral unmixing. The HbO2:HbR ratio was used for comparison of AVM and VM.
Statistical Analysis
Data were analyzed from May 1 to October 31, 2017, using cLabs software (iThera Medical GmbH) and MATLAB (version R2017b; MathWorks, Inc). Statistical analysis was performed using GraphPad Prism (version 4, GraphPad Software, Inc). The 2-sided t test was used for group comparisons. Results are shown as mean (SEM). P <.05 was considered to be significant.
Results
Among the 12 patients (9 female [75%] and 3 male [25%]; mean [SD] age, 23 [18] years) included in the study, simultaneous acquisition of MSOT and US images enabled exact anatomical colocalization of vascular malformations, followed by spectral unmixing of reconstructed data for HbO2 and HbR quantification (Figure 1A). The clinical hybrid MSOT/US imaging system enabled noninvasive, contrast agent–free, real-time imaging of intralesional HbO2 and HbR, revealing a characteristic pattern to differentiate among AVM, VM, and healthy tissue (Figure 1B). Compared with healthy tissue, MSOT showed significantly higher mean HbO2:HbR ratios for AVMs (1.82 [0.08] vs 0.89 [0.03]; P < .001) and VMs (1.12 [0.04] vs 0.89 [0.03]; P = .001). Moreover, MSOT was able to differentiate AVMs and VMs based on significantly higher mean HbO2:HbR ratios for AVMs compared with VMs (1.82 [0.08] vs 1.12 [0.04]; P < .001).
Figure 1. Noninvasive and Quantitative Differentiation of Vascular Malformations.
A, Principle of multispectral optoacoustic tomography (MSOT) and image reconstruction for quantitative analysis of deoxygenated (HbR) and oxygenated (HbO2) hemoglobin levels. MSOT is based on the photoacoustic effect where high-energy light pulses at different wavelengths that induce repetitive thermoelastic expansion of the scanned tissue, which thereby emits ultrasonographic (US) waves. Owing to different absorber properties (HbR and HbO2), spectral unmixing enables quantitative analysis of tissue biomarkers. B, The HbO2:HbR ratio shows significantly higher mean values for arteriovenous malformations (AVM) compared with venous malformations (VM) or healthy tissue at equivalent depth. MSP indicates microspectrophotometry; data points, calculated ratio of each patient.
With the hybrid MSOT/US system, real-time optoacoustic imaging with US-based anatomical coregistration of vascular malformations such as AVMs was possible at the patient’s bedside (eFigure 1A in the Supplement). Three patients with AVMs presented with a significantly higher mean HbO2:HbR ratio (1.82 [0.08]) before treatment than after complete AVM embolization (1.06 [0.03]; P = .001) (Figure 2A) or than in healthy tissue from 6 patients (0.94 [0.04]). Although complete embolization of AVM showed a distinct reduction in HbO2:HbR ratio (Figure 2B, green label), partial embolization of AVM (n = 1) resulted in an only moderate reduction (Figure 2B, red label). Accordingly, angiography confirmed complete treatment (eFigure 1B in the Supplement) or showed persistent arteriovenous shunting in case of partial embolization (eFigure 1C in the Supplement). Pseudocolor-coded MSOT images revealed lower HbO2 and HbR values, depending on the therapy success between complete (Figure 2C) and partial (Figure 2D) AVM embolization.
Figure 2. Multispectral Optoacoustic Tomography (MSOT)–Based Quantification of Therapy Success in Patients With Arteriovenous Malformations (AVMs).
A, The ratio of oxygenated (HbO2) to deoxygenated (HbR) hemoglobin values shows higher mean values for patients with AVM before compared with after complete endovascular embolization of the malformation and healthy tissue at equivalent depth. B, The HbO2:HbR ratios of patients with AVM before and after endovascular embolotherapy (lines connect same patient). Patient 1 experienced a major drop in HbO:HbR ratio after therapy nearly reaching the level of healthy tissue. Patient 2 revealed only a moderate drop in HbO2:HbR ratio after partial lesion embolization. C, Pseudocolor-coded MSOT images of patient 1 in part B show lower hemoglobin values (especially HbO2) after complete AVM embolization of the malformation. D, Pseudocolor-coded MSOT images of patient 2 in part B show only slightly lower hemoglobin values after partial AVM embolization. Data points indicate calculated ratio of each patient.
In contrast to endovascular embolization of AVMs, VMs are treated using percutaneous sclerotherapy (eFigure 2A in the Supplement). We analyzed hemoglobin values before and after sclerotherapy in 2 patients with VM (Figure 3A). The mean HbO2:HbR ratio dropped from 1.07 before therapy to 0.82 (23% reduction) after successful sclerotherapy, equal to the ratio of the healthy tissue on the contralateral side of these patients (0.79 [0.02]) (Figure 3A). Magnetic resonance imaging showed successful VM treatment (eFigure 2B in the Supplement). Pseudocolor-coded MSOT images revealed a distinct difference in hemoglobin signal before and after sclerotherapy as well (Figure 3B). Despite the small patient number, these data suggest that MSOT enables monitoring of therapy response in patients with vascular malformations.
Figure 3. Therapy Success in Patients With Venous Malformations (VMs).
A, The ratio of oxygenated (HbO2) to deoxygenated (HbR) hemoglobin values in patients with VM before and after percutaneous sclerotherapy (n = 2) compared with healthy tissue (n = 4) at equivalent depth (lines connect same patient; healthy tissue was obtained from the nondiseased contralateral side). After successful sclerotherapy, both treated patients displayed HbO2:HbR ratios comparable to healthy tissue. B, Pseudocolor-coded multispectral optoacoustic tomography (MSOT) images of the VM before and after treatment show a distinct lower hemoglobin (especially HbO2) signal in treated venous malformation after compared with before sclerotherapy. Data points indicate calculated ratio of each patient.
Discussion
Vascular malformations pose a diagnostic and therapeutic challenge owing to their heterogeneous composition and their potential for significant morbidity as early as childhood.2 Because current techniques have various drawbacks for vascular malformation imaging, especially in children, and one may observe a high rate of misdiagnosis of up to 47% and false treatment in a usually young patient population,1 a new tool for diagnosis and treatment monitoring would be desirable to ensure the correct diagnosis and improve patient outcome. To our knowledge, we present the first study of MSOT imaging of vascular malformations. Imaging with MSOT proved versatile in various preclinical studies as well as in first clinical applications in oncology,4,5,7,8 inflammatory bowel disease,6 and vasculature.9,10,11
We demonstrate in a first patient cohort, to our knowledge, that MSOT/US enables differentiation of AVMs from VMs and healthy tissue via noninvasive, quantitative measurement of HbO2 and HbR values. Multispectral optoacoustic tomography showed a significantly higher HbO2:HbR ratio for AVMs compared with VMs, most likely caused by highly oxygenated arterial influx and arteriovenous shunting of these high-flow lesions. Although Doppler US depicts intravascular flow variables and patterns, MSOT quantitatively depicts functional tissue biomarkers such as the HbO2:HbR ratio of the involved tissue. Multispectral optoacoustic tomography allowed for quantification and grading of therapeutic success and safe identification of patients in need of additional treatment. Notably, penetration depth of the laser was limited to 3 cm. Owing to high absorbance of light by the vasculature, imaging of deeper vascular structures requires further technical improvements. Despite this limitation, hemoglobin content and oxygenation measured by noninvasive MSOT can serve as a biomarker for differentiation of vascular malformations.
Conclusions
With its similarities in handling to conventional US, MSOT is easy to use and does not require use of a contrast agent or ionizing radiation, especially favorable for pediatric patients. Thereby, MSOT offers a versatile, noninvasive tool for diagnosis and therapy monitoring of vascular malformations.
eFigure 1. Results of Embolotherapy for Arteriovenous Malformation (AVM)
eFigure 2. Results of Sclerotherapy for Venous Malformation (VM)
References
- 1.Greene AK, Liu AS, Mulliken JB, Chalache K, Fishman SJ. Vascular anomalies in 5,621 patients: guidelines for referral. J Pediatr Surg. 2011;46(9):1784-1789. doi: 10.1016/j.jpedsurg.2011.05.006 [DOI] [PubMed] [Google Scholar]
- 2.Ricci KW. Advances in the medical management of vascular anomalies. Semin Intervent Radiol. 2017;34(3):239-249. doi: 10.1055/s-0037-1604297 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.International Society for the Study of Vascular Anomalies ISSVA classification for vascular anomalies. http://www.issva.org/UserFiles/file/ISSVA-Classification-2018.pdf. Published April 2014. Revised May 2018. Accessed August 29, 2018.
- 4.Becker A, Masthoff M, Claussen J, et al. Multispectral optoacoustic tomography of the human breast: characterisation of healthy tissue and malignant lesions using a hybrid ultrasound-optoacoustic approach. Eur Radiol. 2018;28(2):602-609. doi: 10.1007/s00330-017-5002-x [DOI] [PubMed] [Google Scholar]
- 5.Diot G, Metz S, Noske A, et al. Multi-spectral optoacoustic tomography (MSOT) of human breast cancer. Clin Cancer Res. 2017;23(22):6912-6922. doi: 10.1158/1078-0432.CCR-16-3200 [DOI] [PubMed] [Google Scholar]
- 6.Knieling F, Neufert C, Hartmann A, et al. Multispectral optoacoustic tomography for assessment of Crohn’s disease activity. N Engl J Med. 2017;376(13):1292-1294. doi: 10.1056/NEJMc1612455 [DOI] [PubMed] [Google Scholar]
- 7.McNally LR, Mezera M, Morgan DE, et al. Current and emerging clinical applications of multispectral optoacoustic tomography (MSOT) in oncology. Clin Cancer Res. 2016;22(14):3432-3439. doi: 10.1158/1078-0432.CCR-16-0573 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Neuschmelting V, Lockau H, Ntziachristos V, Grimm J, Kircher MF. Lymph node micrometastases and in-transit metastases from melanoma: in vivo detection with multispectral optoacoustic imaging in a mouse model. Radiology. 2016;280(1):137-150. doi: 10.1148/radiol.2016160191 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Taruttis A, Timmermans AC, Wouters PC, Kacprowicz M, van Dam GM, Ntziachristos V. Optoacoustic imaging of human vasculature: feasibility by using a handheld probe. Radiology. 2016;281(1):256-263. doi: 10.1148/radiol.2016152160 [DOI] [PubMed] [Google Scholar]
- 10.Taruttis A, Wildgruber M, Kosanke K, et al. Multispectral optoacoustic tomography of myocardial infarction. Photoacoustics. 2012;1(1):3-8. doi: 10.1016/j.pacs.2012.11.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Karlas A, Reber J, Diot G, et al. Flow-mediated dilatation test using optoacoustic imaging: a proof-of-concept. Biomed Opt Express. 2017;8(7):3395-3403. doi: 10.1364/BOE.8.003395 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eFigure 1. Results of Embolotherapy for Arteriovenous Malformation (AVM)
eFigure 2. Results of Sclerotherapy for Venous Malformation (VM)



