Classification of vascular anomalies.

Dear Editor:

In 1982, Mulliken and Glowacki described a clinically relevant classification of vascular anomalies in which these lesions were categorized according to their endothelial cell characteristics. (1) This classification, which was adopted by the International Society for the Study of Vascular Anomalies in 1996, differentiates proliferating tumors (most of which are hemangiomas) from vascular malformations, which are structural anomalies involving capillaries, venules, veins, lymphatic channels, and combinations of these structures. Vascular anomalies are classified as follows:

* Tumors
    Juvenile hemangioma
    Rapidly involuting congenital hemangioma
    Noninvoluting congenital hemangioma
    Kaposiform hemangioendothelioma
    Tufted angioma
* Vascular malformations
    High-flow
      Arteriovenous malformation
    Low-flow
      Venous malformation
      Lymphatic malformation
      Lymphatic-venous malformation
      Capillary (or venular) malformation (portwine stain)

Hemangiomas are characterized histologically by high endothelial cell turnover, and their clinical "life cycle" includes proliferative, plateau, and involution phases. Hemangiomas are further characterized by cell markers (GLUT-1, merosin, Lewis Y) that are otherwise found only in human placental tissue. Most other vascular anomalies are properly termed "malformations," which are characterized by normal endothelial cell turnover and abnormal gross vascular anatomy.

With interest, I have read in ENT JOURNAL TWO recent articles that unfortunately perpetuate obsolete terminology and misclassify vascular anomalies in the head and neck:

* In the January 2006 PATHOLOGY CLINIC, Thompson addresses "lymphangioma" and describes cystic, capillary, and cavernous subtypes. (2) However, none of these terms is recognized in the current classification of vascular anomalies. Since the suffix -oma implies a proliferating lesion, such masses are more appropriately termed "lymphatic malformations." Subtypes of lymphatic malformations are macrocystic, microcystic, and mixed, and they are further categorized by their anatomic location.

* Monin et al are to be commended for a skillful approach to resection of a vascular lesion from a difficult-to-reach location, as described in the September 2005 issue. (3) However, the mass they described as a "venous hemangioma" is more appropriately termed a "venous malformation," especially since a hemangioma would likely have involuted completely in a patient aged 28 years.

Referral for treatment of vascular anomalies is dependent on an accurate diagnosis of these lesions by the primary care physician. Because the management of vascular anomalies frequently crosses medical disciplines, specialists must agree on a uniform terminology in order to facilitate communication and develop treatment protocols. Otolaryngologists should strive for precision in describing vascular anomalies.

David H. Darrow, MD, DDS

Associate Professor of Otolaryngology and Pediatrics

Director, Center for Hemangiomas and

Vascular Birthmarks

Eastern Virginia Medical School

Norfolk, Va.

References

(1.) Mulliken JB, Glowacki J. Hemangiomas and vascular malformations in infants and children: A classification based on endothelial characteristics. Plast Reconstr Surg 1982;69:412-22.

(2.) Thompson LDR. Lymphangioma. Ear Nose Throat J 2006;85: 18-19.

(3.) Monin DL, Blumner K, Cohen NA, et al. Endoscopic resection of a venous hemangioma of the optic nerve sheath. Ear Nose Throat J 2005;84:586, 588, 590, 592.

Response from Dr. Thompson:

I used the World Health Organization's classification of lymphangiomas, which categorizes histologic types as cystic, cavernous, and capillary. Even so, the histologic classification does not always relate to the clinical manifestations, to the angiographic findings, or to the behavior of the lesion (clinical malformation). Therefore, while terminology is very important, the histologic separation of these lesions cannot distinguish between high-flow and low-flow lesions. "Juvenile hemangioma" is well recognized as only a clinical term, since a capillary hemangioma can occur in a juvenile setting, as can a cavernous hemangioma.

I do not believe that the CLINIC section is a forum for this type of semantic distinction. It only confuses the issue because a clinical term does not always have a histologic counterpart. For example, there are many separations of Mikulicz's syndrome, Sjogren's syndrome, chronic punctate parotitis, etc., which are all morphologically identical on biopsy. And yet they obviously have vastly different clinical presentations and variations in laboratory investigations.

Lester D.R. Thompson, MD

Department of Pathology

Woodland Hills (Calif.) Medical Center

Response on behalf of Monin et al:

Dr. Darrow refers to a 24-year-old clinical classification system that was based on endothelial proliferation in a small series of lesions and was apparently meant to classify congenital lesions. (1) In that classification system, Mulliken and Glowacki primarily distinguished between hemangiomas (proliferative) and malformations (nonproliferative). They included five types of hemangioma and five types of malformation. The main hemangioma was the juvenile type, also called the "infantile" or "cellular" type. Mulliken and Glowacki determined that the juvenile type may go through a proliferation and involution cycle, and they extrapolated that idea to all hemangiomas. However, the vast majority of hemangiomas do not progress through such a cycle.

Mulliken and Glowacki were very focused on certain lesions of the head and neck in the congenital setting. They offered a simplified approach, which Dr. Thompson notes that he does not prefer. It is true that some pathologists likely did call port-wine stains "hemangiomas," even though they are malformations. Mulliken and Glowacki called attention to that misclassification (actually misdiagnosis). By no means was it then, nor is it now, applicable to the complete and accurate classification of bodily vascular lesions.

The classification system used today takes into account advances that have occurred over the past several decades. There are now well over 30 different types of true hemangioma, some of which have proven by cytogenetic studies to be neoplastic. The latest pathology classifications (2-5) of vascular entities, combining blood vascular and lymphatic lesions, includes six categories: hamartomas, malformations, dilations of preexisting vessels (e.g., telangiectasias and angiokeratomas), hyperplasias (e.g., bacillary angiomatosis and papillary endothelial hyperplasia), benign neoplasms (e.g., many hemangioma subtypes such as hobnail, microvenular, sinusoidal), and malignancies (four hemangioendotheliomas, several types of angiosarcoma, and Kaposi's sarcoma). A number of these entities, particularly the benign lesions, are associated with various syndromes. For the record, while many benign hemangiomas may have a proliferative element (now identified by Ki-67/MIB-1 staining), there are also hemangiomas with a very low rate of proliferation and malformations with obvious proliferative foci secondary to induction by high blood flow. Likewise, while many lymphatic lesions ("lymphangiomas") are malformations, they may have proliferative areas, and while other lesions are acquired, still others are neoplastic (e.g., acquired proliferative lymphangiomas). It is not correct to assume that all forms of lymphangioma are malformations.

In the case we described in our article, the lesion was a neoplasm with a capsule and dilated vascular spaces, which happened to be surrounded by smooth muscle. (6) In fact, if the smooth muscle had not been noticed and confirmed with smooth-muscle actin immunostaining, the lesion would have been called a "cavernous hemangioma" (itself distinct from a malformation). Some pathologists might have diagnosed our lesion as an angiomyoma because it also had this combination; however, it was more akin to a leiomyoma with abundant vessels and lacked the arrangement seen in venous hemangiomas. Nevertheless, our lesion was not a malformation. Malformations, which are commonly seen by one of us (J.S.J.B., a soft-tissue pathologist), are unencapsulated groups of vessels, often with both arteries and veins. The structures are frequently dysmorphic, meaning that each vascular channel has variably thick and thin muscle walls. Furthermore, the walls of the veins have a characteristic bundled structure, unlike the neoplasm in our patient. We agree that malformations are not neoplastic, but we emphasize that the lesion in our patient was not a malformation, nor was it misclassified or misdiagnosed.

Classifications change with time and understanding. What seems clear is that classifications used in different specialties may differ. Confusion may stem from the fact that two types of classifications are employed: clinically useful classifications for patient management and the more inclusive pathologic classifications, which specify the six categories of vascular lesions. Typically, the former are not designed to be all-inclusive. Perhaps it is time for subspecialty groups to convene and arrive at a mutual consensus. Until then, we recommend that proliferation not be used as the sole or clearest feature to distinguish between a malformation and a neoplasm, and we call attention to the expanding array of entities within vascular lesions.

John S.J. Brooks, MD

Department of Pathology and Laboratory Medicine

Noam A. Cohen, MD, PhD

David W. Kennedy, MD

Department of Otolaryngology-Head and Neck Surgery

University of Pennsylvania School of Medicine

Philadelphia

References

(1.) Mulliken JB, Glowacki J. Hemangiomas and vascular malformations in infants and children: A classification based on endothelial characteristics. Plast Reconstr Surg 1982;69:412-22.

(2.) Kempson RL, Fletcher CD, Evans HL, et al. Tumors of the Soft Tissues. 3rd series, fascicle 30. Washington, D.C.: Armed Forces Institute of Pathology; 1998:307-86.

(3.) Weiss SW, Goldblum JR, Enzinger FM. Enzinger and Weiss's Soft Tissue Tumors. 4th ed. St. Louis: Mosby; 2001:837-1036.

(4.) Weiss SW, Brooks JJ, eds. Soft Tissue Pathology. United States and Canadian Academy of Pathology Long Course. Philadelphia: Williams & Wilkins, 1996.

(5.) Sangueza OP, Requena L. Pathology of Vascular Skin Lesions: Clinicopathological Correlations. Totowa N.J.: Humana Press; 2003.

(6.) Monin DL, Blumner K, Cohen NA, et al. Endoscopic resection of a venous hemangioma of the optic nerve sheath. Ear Nose Throat J 2005;84:586, 588, 590, 592.