Discrepancies in Initial Clinical and Radiological Diagnoses of Vascular Malformations and the Role of the ISSVA Classification : Journal of Vascular Anomalies

Secondary Logo

Journal Logo

Clinical study (Prospective, Retrospective, Case Series)

Discrepancies in Initial Clinical and Radiological Diagnoses of Vascular Malformations and the Role of the ISSVA Classification

Bolt, Janne W.a; Raphael, Martine F.b; Pasmans, Suzanne G. M. A.c; Langeveld, Hesterd; de Graaf, Nankoe; de Laat, Peter C. J.f

Author Information
Journal of Vascular Anomalies 4(1):p e057, March 2023. | DOI: 10.1097/JOVA.0000000000000057
  • Open



Patients with vascular anomalies are best treated in centers of expertise by a team of specialized physicians, working in a multidisciplinary setting.1,2 The success rate of diagnosis, classification and treatment of vascular anomalies improves with expertise and the quality of interdisciplinary collaboration and communication.

Currently, the most frequently used and accepted classification of vascular anomalies is the International Society for the Study of Vascular Anomalies (ISSVA) Classification. It originates from a proposal of Mulliken and Young in 1982, further specified by ISSVA in 1996 and 2014, and updated most recently in 2018.3–5 This classification divides vascular anomalies into vascular tumors and vascular malformations by natural history, clinical findings, histopathology, and genetics, and facilitates multidisciplinary communication by serving as a uniform language. The use of the ISSVA classification nomenclature in clinical practice has been poorly explored.6,7 These studies showed considerable differences in diagnosis of vascular anomalies between specialties which were explained by differing levels of expertise, lack of a shared classification system, and differences in diagnostic criteria.

In this study, our objective was to investigate the level of agreement between clinicians and radiologists for vascular malformation diagnoses, and the contribution of the ISSVA classification.

Materials and methods

We retrospectively evaluated data of pediatric patients (0–18 years) who were diagnosed with a vascular malformation from 1996 until 2015 in the Vascular Anomalies Center Rotterdam (WEVAR) of the Erasmus MC-Sophia Children’s Hospital in Rotterdam, The Netherlands. The 2014 ISSVA classification was used to determine final diagnoses retrospectively based on the conclusions reached by three members of the WEVAR (S.G.M.A.P, P.C.J.d.L., and H.L.).8 We evaluated the level of agreement in clinical and radiological diagnoses and ISSVA classification use in the initial diagnostic process (P.C.J.d.L. and J.W.B.). Clinical and radiological diagnoses were made by clinicians, such as pediatricians, dermatologists and surgeons, and radiologists that used their level of experience from that moment to make the initial diagnosis. We evaluated the diagnoses as following; for each vascular malformation, we analyzed whether the clinician and radiologist described the same diagnosis in the medical chart. This was evaluated independently from the accurate use of the ISSVA classification. When, for example, the final diagnosis following ISSVA was a venous malformation this could be described by a clinician as “venous malformation” and by a radiologist as “vascular malformation.” We know from experience that these physicians both suspected a venous malformation following the ISSVA classification and thus described the same diagnosis. For this reason, we will refer to this diagnosis as a concordant diagnosis. If the clinician and radiologist described a different type of vascular malformation (or other diagnoses), this is referred to as discordant as we assume that the clinician and radiologist described a different diagnosis. This was for example the case if a clinician described a “venous malformation” and a radiologist described the same lesion as a “lymphatic malformation.” We next evaluated the use of the ISSVA classification terminology independent of the concordant diagnosis. If a vascular malformation was described following the ISSVA terminology by both the clinician and the radiologist, the diagnosis was classified as adherent to the ISSVA classification, if this was not the case it was considered non-adherent. When for example, the clinician described a vascular malformation as a “lymphangioma” and the radiologist described a “lymphatic malformation,” this diagnosis was non-adherent to the ISSVA classification, as the clinician did not use the correct ISSVA terminology.

We used SPSS Statistics version 26 for binary logistic regression analyses.


In total, we identified 158 vascular malformations in 157 patients diagnosed after initial clinical assessment and diagnostic imaging. We included 73 (46%) male and 84 (54%) female patients ranging from 0 to 17 years with a median age of 2 years old. Table 1 shows an overview of the vascular malformations diagnosed and the radiological procedures used in the initial diagnostic process. The use of ultrasound over the consecutive 5-year cohorts decreased (96%, 87%, 77%, and 74%, respectively), while the use of magnetic resonance imaging increased over the 5-year periods (4%, 10%, 19%, and 21%, respectively). The overall use of computed tomography, angiography and lymphoscintigraphy was low.

Table 1. - Final Diagnosed Vascular Malformations Following ISSVA Classification Nomenclature and Specification of Imaging Techniques Used for Diagnostic Evaluation
ISSVA Diagnoses No. VM Diagnostic Imaging Performed
US MRI Other
n n n n
Total 158 129 24 5
Simple vascular malformations 133 110 18 5
 Capillary malformations*
 Lymphatic malformations 71 64 5 2
 Venous malformations 56 42 12 2
 Arteriovenous malformations 5 4 0 1
 Arteriovenous fistulas 1 0 1 0
Combined vascular malformations 20 15 5 0
Malformations associated with other anomalies 5 4 1 0
Abbreviations: ISSVA, International Society for the Study of Vascular Anomalies; MRI, magnetic resonance imaging; US, ultrasound; VM, vascular malformations.
*Capillary malformations were not included as in most patients no imaging is performed.
†Other imaging techniques were: computed tomography (3), lymphangiography (1), or angiography (1).

In Table 2, the concordance of diagnoses by clinicians and radiologists and their adherence to the ISSVA classification in the initial diagnostic process is shown in total and divided into four consecutive 5-year cohorts. In total, the percentage of concordant diagnoses by clinicians and radiologists was 64%. We observed no significant variation in the specificity of the vascular malformation descriptions from the clinicians and radiologists between the 5-year cohorts. Adherence to the ISSVA terminology by clinicians and radiologists was associated with a higher percentage of concordant diagnoses compared to the diagnoses without adherence to the ISSVA (71% versus 52%).

Table 2. - Use of ISSVA Classification Nomenclature and Comparability of Initial Diagnoses by Clinicians and Radiologists
Period of Diagnostic Imaging (y)
1996–2000 2001–2005 2006–2010 2011–2015 1996–2015
Use of ISSVA Classification Nomenclature Total VM n = 24 Concordant Diagnoses Total VM n = 38 Concordant Diagnoses Total VM n = 53 Concordant Diagnoses Total VM n = 43 Concordant Diagnoses Total VM n = 158 Concordant Diagnoses
n = 9 38% n = 22 58% n = 37 70% n = 33 76% n = 101 64%
n/% n % n/% n % n/% n % n/% n % n/% n %
Nonadherent use 12 3 25 15 10 67 18 8 44 15 10 67 60 31 52
Adherent use 12 6 50 23 12 52 35 29 83 28 23 82 98 70 71
% adherent use 50 60 66 65 62
Abbreviations: ISSVA, International Society for the Study of Vascular Anomalies; VM, vascular malformations.

Considering the 4 consecutive 5-year cohorts, there was a clear increase in concordant diagnoses by clinicians and radiologists (38%, 58%, 70%, and 76%, respectively).

Adherence to the ISSVA terminology by clinicians and radiologists was associated with an increase in concordant diagnoses in the 5-year cohorts (50%, 52%, 83%, and 82%, respectively).

In cases which did not adhere to the ISSVA terminology, the concordance of diagnosis was variable (25%, 67%, 44%, and 67%, respectively).

The increase in the percentage of adherence to the ISSVA terminology by clinicians and radiologists was found to be less remarkable over the four consecutive 5-year cohorts (50%, 60%, 66%, and 65%, respectively).

In Table 3, we show additional univariable and multivariable binary logistic regression for the presence of concordant diagnoses. The presence of concordant diagnoses were significantly positively correlated to a diagnostic evaluation in the recent years and ISSVA adherence.

Table 3. - Univariable and Multivariable Binary Logistic Regression Analyses for Concordant Diagnoses
Univariable Multivariable*
Independent Factor OR 95% CI P value OR 95% CI P value
Consecutive cohorts of 5-year periods 1.888 1.344, 2.651 0.000 1.810 1.22, 2.575 0.001
ISSVA adherence 3.750 1.905, 7.383 0.000 3.464 1.711, 7.015 0.001
Abbreviations: CI, confidence interval; ISSVA, International Society for the Study of Vascular Anomalies; OR, odds ratio.
*Order of inclusion following the table.


In this retrospective analysis, we found a difference between clinical and radiological diagnoses in the initial diagnostic process of vascular malformations in around one-third of our patients. This may be explained by the additional information that imaging techniques add to the initial clinical diagnosis and eventually result in a different vascular diagnosis.

The concordance of clinical and radiological diagnoses showed a remarkable increase over the years, although the number of concordant diagnoses observed in this study could be biased by the increased use of MRI over the years and assumptions that had to be made due to the retrospective origin of this study. Adherence to the ISSVA terminology by clinicians and radiologists correlated to concordant diagnoses and the period of diagnostic imaging. This implies that both factors are of importance for the achievement of concordant diagnoses. Horbach et al investigated the level of agreement between clinical and histopathologic diagnoses in patients with vascular malformations.6 They also found differences and suggested that this was partially caused by differing terminology used by clinicians and pathologists. Inconsistent terminology for vascular anomalies in and between clinical disciplines is well known and can lead to miscommunication and incorrect diagnosis.9–11

On the other hand, we can envision that there were several other factors that have positively influenced the increase in the level of concordance of diagnoses. Our Vascular Anomalies Center Rotterdam (WEVAR) started in the early nineties and has received increased recognition over the years resulting in increased patient referrals. Throughout the time of the study cohort, there were no structural changes in the outpatient clinic or the team of involved physicians. The WEVAR implemented formal multidisciplinary meetings with all specialists from 2015 onward. The efforts to speak the same vascular language (ISSVA) is of importance. However, we also observed improvements in our interdisciplinary communication, general knowledge and expertise considering vascular anomalies patient care. These developments have likely contributed to an improved level of concordance between the clinicians and radiologists over time.

Currently, the WEVAR is a multidisciplinary collaboration of pediatric and adult medical disciplines with a high level of expertise in treating patients of all ages with vascular anomalies. Multidisciplinary approach, shared decision making and transition from child to adult care are hallmarks of WEVAR patient care. Close collaboration with other national vascular centers, patient advocacy organizations and international centers (ISSVA) is important to preserve and increase our expertise in the field of vascular anomalies.


In this study, we found that the concordance of initial diagnosis of vascular malformations between clinicians and radiologists increased over the study period. We believe that the simultaneous increase in adherence to the same vascular terminology (ISSVA classification) with the implementation of regular multidisciplinary meetings has been essential. We conclude that close collaboration and clear communication with adherent use of ISSVA terminology between involved disciplines in the care for patients with vascular anomalies increased the level of expertise within our vascular anomalies team.


We gratefully acknowledge Mrs. A.M. O’Byrne for her valuable suggestions.


1. Mattila KA, Kervinen K, Kalajoki-Helmiö T, et al. An interdisciplinary specialist team leads to improved diagnostics and treatment for paediatric patients with vascular anomalies. Acta Paediatr Int J Paediatr. 2015;104:1109–1116.
2. Ernemann U, Kramer U, Miller S, et al. Current concepts in the classification, diagnosis and treatment of vascular anomalies. Eur J Radiol. 2010;75:2–11.
3. Wassef M, Blei F, Adams D, et al.; ISSVA Board and Scientific Committee. Vascular anomalies classification: Recommendations from the international society for the study of vascular anomalies. Pediatrics. 2015;136:e203–e214.
4. Mulliken JB, Glowacki J. Hemangiomas and vascular malfromations in infants and children_ a classification based on endothelial characteristics. Plast Reconstr Surg. 1982;69:412–422.
5. ISSVA Classification of Vascular Anomalies ©2018 International Society for the Study of Vascular Anomalies. issva.org/classification. Accessed December 16, 2021.
6. Horbach SER, Utami AM, Meijer-Jorna LB, et al. Discrepancy between the clinical and histopathologic diagnosis of soft tissue vascular malformations. J Am Acad Dermatol. 2017;77:920–929.e1.
7. Ahlawat S, Fayad LM, Durand DJ, Puttgen K, Tekes A. International society for the study of vascular anomalies classification of soft tissue vascular anomalies: survey-based assessment of musculoskeletal radiologists’ use in clinical practice. Curr Probl Diagn Radiol. 2019;48:10–16.
8. ISSVA Classification of Vascular Anomalies ©2014 International Society for the Study of Vascular Anomalies. Available at: issva.org/classification. Accessed December 16, 2021.
9. Hassanein AH, Mulliken JB, Fishman SJ, Greene AK. Evaluation of terminology for vascular anomalies in current literature. Plast Reconstr Surg. 2011;127:347–351.
10. Liberale C, Rozell-Shannon L, Moneghini L, Stavros T, Giacomo C, Nocini R. Stop calling me cavernous hemangioma! A literature review on misdiagnosed bony vascular anomalies. J Investig Surg. 2022;35:141–150.
11. Steele L, Zbeidy S, Thomson J, Flohr C. How is the term haemangioma used in the literature? An evaluation against the revised ISSVA classification. Pediatr Dermatol. 2019;36:628–633.

diagnostics; ISSVA; vascular anomalies; vascular malformations

Copyright © 2022 The Authors. Published by Wolters Kluwer Health, Inc. on behalf of The International Society for the Study of Vascular Anomalies.