Treatment of Aggressive Hemangiomas: Puncture Vertebroplasty and Radiation Therapy

V.A. Manukovsky, G.M. Zharinov, M.N. Kravtsov, S.D. Rud, A.V. Fedorenkov

Military Medical Academy, Saint Petersburg, Russia


To this day vertebral hemangioma is still not completely investigated nosologic unit, its pathogenesis is not determined, optimal medical tactics is not worked out. So today in the majority of hospitals the basic method of treatment of aggressive hemangiomas is radiation therapy, which has very low efficiency.

Performed analysis of treatment results in 91 (100%) patients (137 vertebras) with aggressive hemangiomas, who were treated in neurosurgery dept. of Military medical academy and radiology dept. of Russian Scientific Center of radiological and surgical technologies during period since 1999 till 2007. Percutaneus vertebroplasty performed on 40 vertebras, radiological treatment performed in 87 vertebras with aggressive hemangiomas.

Vertebroplasty is more effective in treatment of aggressive hemangiomas in comparison with radiotherapy. Vertebroplasty more 34.4-77.8% efficient in pain relief in comparison with base-level of 75.7% of patients, when local pain after radiotherapy decreases in 12.9 – 28.1% in 43.1% of patients.

The attributes of hemangiomas aggression, indications for operations, patients’ management tactics were elaborated. Use of percutaneous vertebroplasty for treatment of aggressive hemangiomas results in fast patient recovery. This operation is low invasive, so it reduces the patients hospitalization period and the time necessary for their rehabilitation and social adaptation.

Key words: vertebral hemangioma, vertebroplasty, osteoplasty, radiation therapy, aggressive factors.

Introduction. The first description of vertebral body hemangioma goes back to 1867, when R. Virchow presented information on a vascular tumor, detected at autopsy. He called it hemangioma (Virchow R., 1867). Application of X-rays in medical practice led to lifetime diagnosis of such hemangiomas. In 1927 E. Perman was the first to dwell on rentgenologic semiotics of vertebral body hemangioma. Three years later P. Bailey and P. Bucy described a rentgenologic picture of the disease in detail (Bailey P., Bucy P., 1929).

According to different authors (Konovalov A.N., 2004; Gaidar B.V., 2002), hemangiomas account for 4-13% of all tumors of the spine and spinal cord and 29% of all primary spinal neoplasms (Lantsman Yu.V., Adamyan A.T., 1986).

Modern reports, dealing with statistics and spread of these tumors, often refer to research, carried out by R. Virchow and C. Schmorl at the beginning of the XXth century. The cause of this is as follows: despite broad clinical application of scanning methods of visualization, there are no CT or MRI studies, devoted to prevalence of vertebral hemangiomas.

The majority of researchers stick to the opinion, that hemangiomas are more frequent in women in a postpubertal period. The ratio between males and females with this clinical entity varies from 1:2 up to 1:2.5 (Belonozhko A.G., 1970; Lantsman Yu.V., Adamyan A.T., 1986; Dorfman H.D., Lisovskaya R.A., 1964; Dofman H.D., 1971).However, there are data, demonstrating almost equal distribution of hemangiomas in these different sexual groups (Gray F. et al., 1989).

Vertebral hemangiomas are absent in children. The majority of cases with this pathology are at the age 30-60 (Lantsman Yu.V., Adamyan A.T., 1986; Dofman H.D., 1971; Lisovskaya R.A., 1964; Pedachenko E.G., Kushchaev S.V., 2005). According to some authors’ opinion, a greater number of cases, watched with aging, can be explained by an asymptomatic course of the disease during many years (Lantsman Yu.V., Adamyan A.T., 1986; Lisovskaya R.A., 1964). Pregnancy and trauma can be a provocative factor, capable of activating a neoplastic process (Inamasu J., 2006). According to J. Nguyen et al. (1989), 10% of all complicated hemangiomas of vertebral bodies is revealed in pregnant women.

Hemangiomas can be localized in any area of the spine. Most frequently they are watched in thoracic (60-76%) and lumbar (22-29%) spine; their localization in cervical (2-11) and lumbosacral (up to 1%) spine is rare (Lantsman Yu.V., Adamyan A.T., 1986; Lisovskaya R.A., 1964; Pedachenko E.G., Kushchaev S.V., 2005: Kornienko V.N., 2006; Nicola N., 1987).

Lesions of one vertebra are revealed in 66-90% of cases (Belonozhko A.G., 1970; Kornienko V.N., 2006). Multiple hemangiomas are diagnosed in 10-34% of cases (Junghanns H., 1932: Laredo J. et al., 1986; Picard L. et al., 1989; Kornienko V.N., 2006) and are usually typical of the aged (Klioner I.L., 1962).

A vertebral body is affected most frequently (up to 100%.); simultaneous involvement of vertebral bodies and arches is detected in 52% of cases, while lesions of processes are rare (Lantsman Yu.V., Adamyan A.T., 1986). Isolated lesions of a vertebral arch is a rarity and they are never watched as far as processes are concerned (Belonozhko A.G., 1970).

According to the majority of authors, there are two groups of vertebral hemangiomas:
1. A “stable”, “inert” or “non-aggressive” hemangioma.
2. “Aggressive” or “active” hemangioma.

“Non-aggressive” hemangiomas form the largest group (more than 90%). They are characterized by extremely slow development and absence of tumor cells. Their place in histological classification is unclear. Some authors consider them to be vascular malformations (Krakovsky N.I., 1974; Solovko A.Yu., 1980; Lee J.P. et al, 1990; Djindjian M., 1992; Dagi T.F., 1990; Raco A., 1990; Zozulya Yu.A., 2002). However, it is difficult to imagine, that such developmental pathology can be watched in every tenth human being of the whole population. Absence of vertebral hemangiomas in children and their high rate in people above 50 indicate, that they are unlikely to be vascular malformations. However, it is possible, that a small percentage of hemangiomas, revealed in people, aged 18-40, can be of a congenital character.

G. Schmorl (1927), D. Topfer (1928) and H. Junghanns (1932) supposed the majority of vertebral hemangiomas to be somehow connected with degenerative processes in the spine, which explained their predominance in the aged. Their rate in people above 60 and younger cases was 16% and 3.8% respectively. D. Topfer (1928) revealed hemangiomas in 11.9% of all studied vertebrae. Their distribution was as follows: cases under 50 – 18.3% and patients above 50 – 81.7%. According to more recent studies by home scientists, an anatomical and histological picture of vertebral hemangiomas in the aged was not of a blastomatous character. Thus, it did not allow to regard them as tumors. They considered them to be angioma-like or telangiectatic changes, developing against a background of osteoporosis and adipose involution of the red bone marrow and having no clinical significance (Klioner I.L., 1962; Vinogradova T.P., 1973).

The second group is represented by tumors proper, characterized by rapid growth, involvement of the whole vertebra and spread beyond its limits, compression of nervous structures. From a histological point of view, such a tumor consists of mature vessels, separated by stroma with rare signs of cellular anaplasia and mitoses. It is just the term “hemangioma”, that should be used for description of such neoplasms (Zozulya Yu.A., 2002). Usually vascular neoplasms of vertebral bodies do not tend to manifest their aggressiveness; the result is rare diagnosis of hemaniomas of this localization. They account for 0.9-4% of all vertebral hemangiomas.

According to classifications of bone tissue tumors, hemangiomas are referred to benign neoplasms or a debatable morphologic group, standing between a tumor and developmental vascular defect.

Table 1

Classification of Vascular Spinal Tumors
(Yu.A. Zozulya, E.I. Clynko, 2000)










Neoplastic angioendotheliomatosis

Aneurysmatic osseous cyst


Table 2

Special (Detailed) Classification of Vascular Tumors of Bones
(T.P. Vinogradova, 1973)




Angiosarcoma and its versions:




Malignant hemangiopericytoma


Ewing’s tumor


Adamantinoma of long bones


Other angioplastic malignant tumors

From a histological point of view, hemangiomas are subdivided into capillary, cavernous and mixed.

Thus, pathogenesis and pathomorphology of vertebral hemangiomas are still an object of heated debate. Uncertainty of their place in a histological row is greatly explained by their different behavior, as they vary from inert clinical forms up to active tumors with a destructive growth. Vertebral hemangioma is clinical and rentgenologic diagnosis, characterized by rather distinct criteria and covering neoplasms, which are different in their essence.

Spondylography, CT and MRI are the main methods of vertebral hemangioma diagnosis. Spondylography makes it possible to reveal signs of a vertebral body lesion (changes of bone tissue and a vertebra shape, preservation of terminal laminae and a cortical layer) at a prehospital stage.

CT signs of hemanigioma include:

The most important MR-symptom of vertebral hemangioma is a spotted character of a signal, which is hyperintensive and iso-hypointensive on T2- and T1-weighted images, respectively. When a soft-tissue component of hemangioma is present, MRI permits to estimate its spread.

Today there is no unanimous opinion, concerning hemangioma aggressiveness. What are the signs and their number, which would allow to consider hemangioma to be aggressive?

Comparing clinical symptoms and data of radiodiagnosis, J.D. Loredo et al. (1986) elaborated the following criteria of hemangioma aggressiveness:

  1. Localization at the level of Th3–Th9 (Fig. 1);

    Fig. 1. MRI: Aggressive hemangioma of the ThVIII body.

  2. A total lesion of a vertebral body (Fig. 2);

    A      B
    Fig. 2. CT: Hemangioma, occupying almost the whole volume of the ThVII body (À - axial section; B – reconstruction).

  3. Tumor, spreading to a root and vertebral arches (Fig. 3);

    Fig. 3. CT: Aggressive hemangioma with a total lesion of the ThXII body and spread to its arch.

  4. X-rays indicative of bone expansion and cortex protrusion with poorly defined margins (Fig. 4);

    Fig. 4. X-ray examination: Aggressive hemangioma of the ThIV vertebra. Bone expansion and cortex protrusion with poorly defined margins.

  5. An irregular trabecular structure of hemangioma (Fig. 5);

    Fig. 5. CT: An irregular trabecular structure of hemangioma of the ThVII body.

  6. Presence of epidural or paravertebral components of tumor (Fig. 6);

    A      B
    Fig. 6. MRI: An epidural component of hemangioma of the LI vertebra with a lesion of its arch pedicle (À);
    hemangioma of ThXI vertebra, spreading paravertebrally and to the right pedicle of its arch (B).

  7. Low and high signal on T1- and T2-weighted images, respectively; CT indicative of contrast accumulation (Fig. 7).

    Fig. 7. MRI: Hemangioma of the LIII vertebra. Hyperintensive and hypointensive signals in T2- and T1-weighted images, respectively.

H. Deramond et al. (1989) proposed to consider hemangioma as an aggressive one in case of revealing any three of the above signs.

Pedachenko E.G. and Kushchaev S.V. (2005) tried to define groups of “absolutely aggressive” and “absolutely non-aggressive” hemangiomas, as well as disputable criteria of aggressiveness.

I. Absolute criteria of aggressiveness

  1. A compression fracture of a vertebral body.
  2. The spinal cord compression.
  3. A tumor component, localized epidurally.
  4. Epidural hematoma.
  5. CT: absence of lipids in intratrabecular space and presence of a soft-tissue solid part of tumor.
  6. MRI: low intensity of a signal from tumor in T1-weighted images against a background of their high intensity in T2-weighted images.
  7. MR-spectroscopy: a low content of lipids.

II. Absolute criteria of non-aggressive hemangioma

  1. CT: adipose tissue in intertrabecular space.
  2. No soft-tissue solid part of tumor.
  3. MRI: high intensity of a signal from tumor in T1- and T2-weighted images.
  4. MR-spectroscopy: a high content of lipids.

III. Disputable signs of hemangioma aggressiveness

  1. Hemangioma localization at a thoracic level.

This classification does not include description of a sign of “large hemangioma”, affecting a greater part of a vertebral body. Many researcher assert, that hemangioma of more than 60% of a vertebral body volume threatens with compression fracture development (Laredo J.D., 1986; Deramond H. et al., 1989). A criterion of hemangioma growth dynamics is still obscure. What hemangiomas demand dynamic control and what size determines necessity of active measures on a surgeon’s part? Is this choice dependent on other signs? All the above criteria cannot be used for making a decision on a choice of indications for treatment, as they contain no explanations, concerning it.

Today tactics of treatment of aggressive hemangiomas is unclear. Some time ago the main method was radiotherapy. However, its expediency is questionable due to a large radiation load necessary for achieving an antalgic effect (D. Rades et al, 2003). Even if radiotherapy does cause improvement in some cases, it results in no change of a rentgenologic picture of hemangioma and is not a guarantee of relapse absence (Alborov G.K., 1970).

According to many authors, puncture vertebroplasty (PVP) is an extremely effective method for treatment of this pathology. P. Galibert and H. Deramond were the first to administer bone cement into an affected vertebral body in 1984. The experience of the first operations demonstrated, that filling of hemangioma with bone cement stopped its growth, ensured restoration of supporting ability of an affected vertebral body and eliminated pain syndrome.

Despite a rather long-term use of vertebroplasty, no large-scale randomized trials of its efficacy in aggressive vertebral hemangiomas have been carried out. The world experience is based on numerous observations, including from 3 up to 50 cases of PVP application in this pathology.

Today there no generally accepted indications and contraindications for PVP. Clinical and rentgenologic signs of vertebral hemangioma aggressiveness are the leading indications. Specific signs of hemangioma “activity” and their relationship with its size are still in question.

The goal of the present study was estimation of PVP and radiation therapy efficacy in treatment of aggressive vertebral hemangiomas.

Materials and Methods.

Examination data and results of PVP were analyzed in 91 cases with aggressive vertebral hemangiomas (100%; 137 vertebrae), treated in the neurosurgical clinic of the Medicomilitary Academy and radiotherapy departments of the Russian Research Center of Radiology and Surgical Technologies in 1999-2002. There were 23 males (25.3%; 25 vertebrae) and 68 females (74.7%; 112 vertebrae). Mean age was 52.7 years and varied from 19 up to 81. The main group consisted of 33 cases (36.3%; 43 vertebrae), who underwent surgical treatment (group I). The control group included 58 cases (63.7%; 94 vertebrae), subjected to radiotherapy (group II). PVP and radiotherapy were used for treatment of 40 and 87 vertebrae, affected by aggressive hemangiomas, respectively. Aggressive hemangioma was diagnosed in outpatient departments in all the cases.

Localization of aggressive hemangiomas in groups I and II was as follows: thoracic spine – 26 (60.4%) and 43 (45.7%), lumbar spine – 14 (32.5%) and 48 (51.1%), cervical spine – 3 (6.9%) and 3 (3.2%), respectively. Multiple hemangiomas were revealed in 7 (21.2%) and 13 (22.4%) of cases of the main and control groups, respectively. Cases with 3 and more hemangiomas were females, whose mean age was 57.4 years.

All the patients were examined in compliance with a standard scheme. It consisted in general and neurological examination, clinical and laboratory tests, rentgenologic (including CT) and MRI studies. Sometimes examination was supplemented by special methods. Localization of a process, its spread, severity of clinical manifestations, presence of accompanying diseases were estimated on the basis of examination data. As a rule, diagnosis verification was confirmed by revealing local pain syndrome, X-ray and MR syndromes.

Pain severity, motor disorders, connected with musculotonic reflex reactions and analgesics effectiveness were assessed according to the Life Quality Scale, elaborated by J.R. Gaughen et al. (2000) (Table 3), and the Original Scale of Life Quality, worked out in the neurosurgical clinic of Medicomilitary Academy (Table 4).

Table 3

Life Quality Scale (according to J.R. Gaughen et al., 2000)



Severity of Pain Syndrome:
No pain


The most severe pain


Motor Activity:
No limitations


Unassisted walking


Moving in a wheelchair


Limited sitting in bed


Limited movements in bed


Drug Dependence:
No analgesics


Episodic taking of analgesics


Regular taking of analgesics


Episodic oral taking of narcotic drugs


Regular oral taking of narcotic drugs


Regular parenteral administration of narcotic drugs


Table 4

Original Life Quality Scale
(V.A. Manukovsky, M.N. Kravtsov)



Severity of Local Pain Syndrome:
No pain or minor episodic pain


Constant minor pain


Moderate episodic pain


Constant moderate pain


Episodic severe pain


Constant severe pain


Motor Activity:
No limitations


Ability to hold a vertical position during a day and small limitations of a volume of spine movements


Ability to hold a vertical position during a day with periodic rest (in a horizontal position); a limited volume of spine movements


Considerable limitations of physical activity in a vertical position and a volume of spine movements


Inability to hold a vertical position without somebody’s assistance and auxiliary means (a stick, crutches)


A forced position (sitting, lying in bed)


Drug Dependence:
No analgesics


Episodic taking of analgesics


Regular taking of analgesics


Episodic oral taking of narcotic drugs


Regular oral taking of narcotic drugs


Regular parenteral administration of narcotic drugs


CT-examination was carried out in 32 cases (96.7%) with applying the SOMATOM PLUS - 4A device (Siemens, Germany) of the forth generation (a matrix of 1024´1024 pixels). Magnetron 63SP apparatus (Siemens) with a magnetic field induction of 1.5 T was used for MR-examination of 33 patients (100%).

The data of CT- and MR-examinations allowed to estimate all known criteria of vertebral hemangioma aggressiveness, described in medical literature. Indications for treatment were determined on the basis of a score, reflecting severity of clinical and rentgenologic symptoms of aggressiveness and obtained on the basis of the Original Life Quality Scale. The classification by H. Deramond, A. Cotten, C. Depriester (2002) was also taken into account.

Remote irradiation of 58 cases (94 vertebrae) of the control group was carried out with home (LUEV-15 M1) and foreign (SL-20 and SL-75-5, Siemens) linear electron accelerators, whose boundary energy of braking radiation varied from 6 up to 18 MeV (Fig. 8).

Fig. 8. The SL-20 linear electron accelerator (Philips).

Preparation for irradiation was based on using the imitator, manufactured by Philips, CT (or MRI), rentgenologic diagnostic apparatus, adapted for this purpose. Optimum dosimetric planning was carried out with the help of a computer station; tissue heterogeneity and a body surface curvature were taken into consideration.

We used the following stations of dosimetric planning:

  1. KOSPO-2d-planning (Russia), based on Pentium I and Wintime KD digitizer.
  2. ROCS-2.5d-planning (Radiation Oncology Computer Systems, USA; version 5.1.6), based on Pentium I and Numonics digitizer.

Remote irradiation of cases with hemangiomas of the cervical spine (3 vertebrae, 3.4%) was carried out in a static mode via one or two tangential fields. Both mobile and static modes of remote irradiation were used in patients with hemangiomas of the thoracic (42 vertebrae, 48.3%) and lumbar (42 vertebrae, 48.3%) spine. We chose such conditions of irradiation, which ensured isodose distribution of 80-90% over zones of interest (Fig. 9).

Fig. 9. Isodose distribution in local irradiation of the CVI body, carried out in a static mode via two tangential fields
with applying a linear electron accelerator, whose boundary energy of braking radiation was 6 MeV.

Irradiation was carried out in a mode of conventional fractioning and with a single radiation dose of 2 Gy (daily, 5 times a week); a total radiation dose reached 30 Gy. In multiple lesions of the spine (13 cases, 22.4%) static irradiation from two or three fields was preferable. A total radiation dose varied from 30 up to 38 Gy.

Puncture vertebroplasty was performed in 33 cases (40 vertebrae). It was multilevel in 6 patients (3 levels – 1; 2 levels – 5). Multilevel PVP was made during one and the same intervention in 5 cases. One patient with three aggressive hemangiomas was subjected to double-stage operation. PVP was combined with open intervention (removal of an epidural soft-tissue component of hemangioma, spine stabilization) in 2 cases.

A patient’s position on an operating table and intraoperative marking did not differ from those, used in PVP for other spinal diseases.

I/V anesthesia, combined with local administration of anesthetics, was usually given in PVP of thoracic and lumbar spine. Puncture vertebroplasty at a cervical level was performed under general anesthesia with trachea intubation.

We used several types of puncture needles (Stryker and Cook companies, USA). Their distal ends differed in their form. They were either tapered (45°, 75°) or conical and tetrahedral (pyramidal). Handles of needles with tapered distal ends had a special mark, showing a cut-off direction (Fig. 10). An outer diameter of needles was 13 G (2.41 mm) and 11 G (3.05 mm). A choice of a needle size and its type was dependent on a vertebra level, its anatomic peculiarities, hemangioma size and localization.

Fig. 10. A needle and mandarin with a tapered distal end for puncturing a vertebral body.

While performing vertebroplasty, we used the most widespread method of navigation, i.e. fluoroscopy (Polystar T.O.P. and Angiostar Plus multifunctional rentgenologic apparatus, Siemens, Germany) (Fig. 11). One operation was made under control of the SOMATOM PLUS-4A spiral computer tomograph (Siemens, Germany).

Fig. 11. Puncture vertebroplasty, performed under fluoroscopic control.

Obtained data were analyzed with applying both general (ACOM.PC program) and special operating stations. The latter permitted to get two and three-dimensional reconstructions. Intraoperative digital images were presented on a monitor screen in a fixed mode and a mode of their electron-optic conversion.

Intraoperative venospondylography was used in 100% of cases. It was done by attaching the Luer-Lock syringe with a contrast substance (Omnipaque, Ultravist) to a puncture needle. A contrast substance dose per one examination was dependent on a volume of vertebral bodies. It was as follows: cervical spine – 3-4 ml, thoracic spine – 6-10 ml and lumbar spine – 6-10 ml (on each side). Dosage of contrast substances was in compliance with recommendations of their manufacturers.

Methods, chosen by us, allowed to get additional information on hemangioma:

  1. To determine a real size of vertebral hemangioma more precisely.
  2. To estimate a degree of hemangioma vascularization and direction of venous outflow (intracranial or extracanal).

The following types of bone cement were used: Simplex P (Stryker, USA), Spineplex, Palacos and CMW (Johnson&Johnson, USA). All of them are based on polymethylmetacrylate and differ from each other by different ratio of monomer and polymer, which conditions cement viscosity.

Mixing of bone cement components was carried out in a piston compartment of the system of bone cement delivery (PCD-Stryker, USA). Sometimes mixers of Stryker (USA) and Somatex (Germany) companies were used (Fig. 12). In case of vertebroplasry of cervical spine it was done by hand.

Fig. 12. The Somatex mixer (Germany) for mixing bone cement components.

Cement was administered into cervical vertebrae via an original device for delivery of viscous masses into vertebral bodies and cerebral vascular malformations (developed and tested at the Faculty of Neurosurgery of the Medicomilitary Academy and Polenov Research Neurosurgical Institute; patent N 72136 of April 10, 2008). It allowed to control pressure of delivered cement (Fig. 13).


Fig. 13. A device for delivery of viscous masses into vertebral bodies and vascular malformations (À – a general view; B – a drawing).

PVP at thoracic and lumbar levels was performed with application of special systems, intended for bone cement delivery. One of them was the Percutaneous Cement Delivery (PCD) System (Stryker, USA) (Fig. 14).

Fig. 14. The PCD system for bone cement delivery (Stryker, USA).

When needles were removed, hemostasis was achieved by pressing; an aseptic bandage was applied for 2-3 days. Control spondylographic examination was carried out at an operation level and an X-ray of lungs was made. If intervention, made under conditions of local and I/V anesthesia, resulted in no complications, a patient was transferred from an operating room into a general ward (10-15 minutes after awakening). In case of intubation narcosis, a patient demanded dynamic observation of 2 hours in an intensive care unit.

The patients became active in 2-3 hours after operation. Non-steroid anti-inflammatory agents were taken on the 2nd-4th days with the purpose of arresting pain, caused by intervention. Control CT-examination was carried out on the next day after operation. If there were no clinically important complications and a patient’s state was satisfactory, he or she was dismissed on the 2nd day after surgery.

Treatment efficacy was estimated on the basis of the Life Quality Scale by J.R. Gaughen et al. (2002) and Original Life Quality Scale, elaborated in the neurosurgical clinic of the Medicomilitary Academy. Radiotherapy efficacy was assessed after its completion and in 6 months. The first results of vertebroplasty were analyzed a week after intervention. Such an approach was conditioned by painful sensations, caused by operation trauma and watched in the majority of cases. They regressed during the first 3-5 days. Assessment of pain syndrome severity, a degree of motor disorders and dependence on anesthetics, as well as control CT and MR-examinations were carried out in 1 and 6 months.

Results. Usually aggressive vertebral hemangiomas manifested themselves by constant pain syndrome of moderate severity (the score of 6 and 3.2 according to Gaughen’s Scale and the Original Scale, respectively). There was no considerable difference between the compared groups. Some cases (8, 24.2%) characterized their pain as violent (the score of 7-8 and 4 according to Gaugen’s Scale and the Original Scale, respectively). Pain syndrome was predominant in cases with osteoporosis and marked degenerative-dystrophic changes at the level of vertebral hemangioma (the score of 2.7 versus 3.2 according to the Original Scale and the score of 5.1 versus 5.6 according to Gaughen’s Life Quality Scale).

Local pain syndrome, reflex muscular-tonic responses, radicular pain limited movements in the spine and general motor activity of cases during a day. However, in the majority of cases motor disorders did not reach a degree, which led to inability to move without somebody’s assistance (the score of 2.4-2.4 in groups I and II according to the Original Scale). Thus, the Life Quality Scale by J.R. Gaughen et al. (2000) did not reflect the whole spectrum of motor disorders in cases with vertebral hemangiomas. The score of impaired motor activity, estimated on the basis of Gaughen’s Life Quality Scale, never exceeded 1 in both groups. It meant, that the sample did not include cases, being in a forced horizontal position or moving in a wheelchair.

A choice of a pharmacological group of anesthetics and frequency of their taking were directly dependent on severity of local pain syndrome, radicular syndrome and individual pain tolerance. Narcotic analgesics in tablets were taken from time to time by 3 cases of each group (9.1% and 5.1%, respectively). The rest patients took non-steroid anti-inflammatory drug either episodically or constantly.

The Original Life Quality Scale of the Medicomilitary Academy was used for subdivision of all the patients into three categories:

  1. Cases with good life quality (the score of 0-3).
  2. Cases with satisfactory life quality (the score of 4-9).
  3. Cases with bad life quality (the score of 10-15).

The categories ratio within the groups is demonstrated in Fig. 15.

Fig. 15. Life quality of cases with aggressive hemangiomas (groups I and II), estimated on the basis of the Original Scale of the Medicomilitary Academy.

Distribution of hemangiomas according to their localization, size and spread was as follows (Table 5):

Table 5

Localization, Size and Spread of Vertebral Hemangiomas


Group I

Group II

Hemangioma Localization:

Hemangioma localization within the limits of a vertebral body

25 (58.1%)

57 (60.6%)

Simultaneous lesion of a vertebral body and a vertebral arch pedicle

15 (34.9%)

36 (38.3%)

Involvement of only posterior vertebral structures



Total lesion of a vertebra

3 (7%)


A Degree of a Vertebral Body Lesion:

Up to 1/3 of a vertebral body

11 (25.6%)

32 (35.1%)

Up to 2/3 of a vertebral body

9 (20.9%)

28 (29.8%)

Up to 3/3 of a vertebral body

23 (53.5%)

30 (35.1%)

Hemangioma Spread:

Within the limits of a vertebra

37 (86.1%)

91 (96.7%)

Epidural spread

4 (9.3%)

1 (1.1%)

Paravertebral spread

2 (4.6%)

2 (2.2%)

Indications for treatment of hemangiomas were dependent on such a notion, as aggressiveness. Diagnosis of aggressive hemangioma was based on a complex of clinical and rentgenologic symptoms. Final diagnosis was confirmed by results of histological study.

Considerable weakening of mechanical strength of a vertebral body was regarded to be rentgenologic (CT) and MR-criteria of hemangioma aggressiveness. We also took into account signs, characterizing hemangioma as primary neoplasms, i.e. bone tissue destruction, growth and spread beyond vertebra limits. Clinical criteria of aggressiveness included local pain syndrome, radicular and conductive disorders, caused by hemangioma and corresponding to a level of its localization. The rate of the above criteria is presented in Table 6.

Table 6

Rate of Occurence of CT- and MR-criteria of Aggressiveness


Aggressiveness Signs

Group I

Group II


Localization at the level of Th3-Th9

13 (32.5%)

27 (31%)


Total lesion of a vertebral body

23 (57.5%)

30 (34.5%)


Spread to a vertebral arch root

18 (45%)

37 (42.5%)


An irregular trabecular structure of hemangioma

30 (75%)

33 (37.9%)


Bone expansion with cortex protrusion and indistinct margins

2 (5%)

7 (8%)


Presence of neoplastic epidural or paravertebral components

6 (15%)

3 (3.4%)


Low and high signals on T1- and T2-weighted images, respectively; accumulation of a contrast substance in CT examination

21 (52.5%)

40 (45.9%)


A pathologic compression fracture

6 (15%)

10 (11.5%)


Epidural hematoma

0 (0%)

0 (0%)


Lesion of a cortical layer (thinning or destruction)

11 (27.5%)

23 (26.4%)


Increased accumulation of a radioactive drug in skeleton scintigraphy

0 (0%)


Increased accumulation of a radioactive drug in scintigraphy with applying labeled erythrocytes

0 (0%)

The term “absolute criteria of hemangioma aggressiveness”, generally adopted by researchers and clinicians and implying considerable decrease of vertebra strength, and our personal clinical experience were used for scoring each of the aggressiveness signs (Table 7).

Òàáëèöà N7

Scoring of Aggressiveness Criteria of Hemangioma Vertebrae


Aggressiveness Signs



Presence of epidural or paravertebral components of hemangioma



A compression fracture and compression deformity of a vertebral body, affected by hemangioma



Bone expansion with protrusion of a cortical layer



A hemangioma size of more than 2/3 of a vertebral body volume



Lesion of a cortical layer (thinning and/or destruction)



A hemangioma size, varying from 1/3 up to 2/3 of a vertebral body volume



An irregular trabecular structure of hemangioma



Spread to a vertebral arch root



Low and high signals from hemangioma on T1- and T2-weighted images, respectively



Local pain syndrome, corresponding to a level of an affected vertebra; topical radicular and myelopathic syndromes; atypical pain, characteristic of hemangiomas of Th3-Th9 vertebrae


The threshold score, indicative of hemangioma aggressiveness, is 5.

The classification by H. Deramond, A. Cotton, C. Depriester (2002) divides patients with vertebral hemangiomas into 4 groups according to clinical and rentgenologic signs. Using this classification and the above scores, we grouped cases with hemangiomas from the point of view of treatment necessity:

  1. Non-aggressive hemangiomas, i.e. tumors, localized in a vertebral body and having a size of less than 1/3 of its volume; the score of no more than 2.
  2. Relatively aggressive hemangiomas; the score of less than 5.
  3. Absolutely aggressive hemangiomas; the score of 5 and more:
    à) Tumors with epidural spread and signs of compression of the vertebral canal structures or a radicular nerve canal.
    á) Tumors with no signs, mentioned in item 3-a.

Cases of the first and second groups require dynamic observation with control CT-examination.

Treatment of patients, belonging to 3-b group, should be directed at restoration of mechanical strength of a vertebral body (puncture vertebroplasty).

Cases of group 3-a demand open intervention (decompression of nervous structures with intraoperative puncture vertebroplasty).

Results of PVP and radiotherapy were estimated with taking into account the following factors:

  1. Changes of a rentgenologic picture and MR-images of hemangioma.
  2. PVP efficacy.
  3. Dynamics of clinical manifestations of the disease.
  4. Presence or absence of complications.

PVP results demonstrated, that cement within a vertebral body (CT examination) produced a high-density signal, which considerably exceeded bone tissue density. As for MRI, it was characterized by a hypointensive signal in T1- and T2-weighted images. Filling of hemangioma cavity by 80% and more was considered to be total. It was watched in 92.5% of cases (37 levels). Control CT- and MR-examinations, carried out 6 months and a year after operation, revealed no further growth of hemangioma in all the cases, as well as no signs of local responses of bone tissue to cement.

Radiotherapy results were indicative of no change of a rentgenologic picture in 1 and 6 months (55 cases, 94.8%). Control examination, carried out in 3-4 months, revealed pathologic compression fractures of vertebral bodies, affected by hemangioma, in 3 cases (5.2%). A particular MR-sign of vertebrae, subjected to ionizing radiation, was adipose involution of bone marrow (25 cases, 43.1%), which developed during the first months. It resulted in a hyperintensive signal from irradiated vertebral bodies and hemangioma in T1- and T2-weighted images (Fig. 16).

A      B
Fig. 16. Radiotherapy of hemangioma of the LIII vertebra (A – before treatment; B – after treatment).

By the end of the first postoperative week (PVP) local pain syndrome intensity subsided by 30% (J.R. Gaughen’s Scale) or 33.9% (the Original Scale of the Medicomilitary Academy) in 23 cases (69.7%). Though it was the same in 7 out of 10 cases (30.3%), they noted its different character (regression of burning and bursting pain in projection of an affected vertebra). Pain syndrome intensification and minor aggravation of neurological deficit were watched in 1 case (3%) after PVP. Pain regression was accompanied by improved motor activity and better tolerance of static physical loads in 19 cases (57.6%); 20 patients (60.6%) continued to take non-steroid anti-inflammatory drugs episodically; 7 cases (21.2%) gave up taking analgesics and 5 patients (15.1%) needed them constantly.

By the end of the first month a positive effect of PVP (pain regression) was watched in 18 cases (54.5%). A mean score decreased up to 2.8 (J.R. Gaughen’s Scale) and 1.5 (the Original Scale of the Medicomilitary Academy), i.e. by 47.2% and 50% in comparison with a preoperative level, respectively. It was indicative of insignificant pain severity. There was a tendency to motor activity improvement (14 cases, 42.4%) and lesser dependence on analgesics (17 cases (51.5%) did not need them anymore). Radicular syndrome, observed in 8 cases (24.2%) of group I, regressed completely or partially in 3 and 1 of them, respectively, a month after PVP.

Control examination and estimation of PVP efficacy, carried out 6 months after operation, demonstrated a positive result. There was regression of pain syndrome intensity by 33.4-77.8% (56.7% on the average; the Original Scale of the Medicomilitary Academy) and 52.8% (J.R. Gaughen’s Scale) from the initial level (25 cases, 75.7%). Besides, one could watch improvement of motor activity and better resistance to heavier static loads on the spine (24 cases, 72.7%). Neurological disorders (radicular syndrome) regressed partially or completely in 62.5% of cases. Several patients (16, 48.5%) gave up taking analgesics (Fig. 17).

Fig. 17. Dynamics of mean scores, characterizing pain syndrome, motor activity limitations and dependence
on analgesics in cases of group I (the Original Scale of the Medicomilitary Academy).

Fig. 18. A complication of puncture vertebroplasty in aggressive hemangioma. Epidural spread of bone cement.

One case (3%) developed a significant complication, conditioned by bone cement penetration into paravertebral and epidural spaces (Fig. 18). Conservative therapy resulted in complete regression of radicular syndrome.

According to the Original Life Quality Scale of the Medicomilitary Academy and J.R. Gaughen’s Scale, radiotherapy (group II) led to decrease of local pain syndrome intensity by 18.7-20% in 30 (57.1%) and 33 patients (56.9%), respectively. Radiotherapy did not cause aggravation in any case. Pain intensity remained the same in 25 cases (43.1-48.3$% according to different scales); 60.3% of patients characterized it as moderate and sometimes severe. Improvement of general motor activity and a volume of movements in the spine was observed in 20 cases (34.6%). Manifestations of muscular-tonic syndrome and low tolerance of static loads were present in 38 cases (62.3%). Dependence on non-steroid anti-inflammatory drugs reduced in 23 patients (39.6%).

There were no considerable changes of life quality indices 6 months after radiotherapy. Local pain syndrome intensity decreased by 12.9-28.1% in 25 cases (43.1%). Motor activity, limited by pain and muscular-tonic syndromes, improved by 11.5% in 20 cases (34.5%). Constant taking of analgesics was given up by 35 cases (60.3%).

Radicular syndrome regression, caused by radiotherapy, was watched by the beginning of the second and sixth months in 4 (30%) and 2 more cases, respectively. Episodic or constant radicular pain and sensitivity disorders, corresponding to a level of an affected vertebra, were watched in 6 cases (50%) (Fig. 19).

Fig. 19. Dynamics of mean scores, characterizing pain syndrome, motor activity limitations
and dependence on analgesics in cases of group II (the Original Scale of the Medicomilitary Academy).

Radiotherapy caused gastrointestinal disorders in 2 cases (3/4%). They became an indication for its temporary discontinuation. We did not estimate possible long-term complications of radiotherapy. The results of PVP and radiotherapy, used for treatment of aggressive vertebral hemangiomas, were compared. This comparison demonstrated intra-group redistribution within three levels of life quality. It was watched after the first and sixth months of treatment (Fig. 20-21).

Fig. 20. Life quality of cases with aggressive hemangiomas (Groups I and II)
a month after treatment (the Original Scale of the Medicomilitary Academy).

Fig. 21. Life quality of cases with aggressive hemangiomas (Groups I and II)
six months after treatment (the Original Scale of the Medicomilitary Academy).

According to the obtained data, there was considerable numerical superiority of cases with good life quality, being the result of PVP (51.5%), in comparison with patients, subjected to radiotherapy (20.7%). It should be emphasized, that cases without marked positive dynamics in a clinical picture, watched during the first month of treatment (especially after PVP), would never demonstrate subsequent considerable improvement. Thus, efficacy of PVP and radiotherapy can be estimated a month after treatment. There is no sense to cherish hopes for big future changes of pain intensity and reflex muscular-tonic responses, connected with it.


  1. The optimum diagnostic algorithm in suspected vertebral hemangioma is to include neurosurgical examination, CT, spondylography (an outpatient stage) and MRI of an affected spinal segment.
  2. PVP is a more effective method of treatment of aggressive vertebral hemangiomas in comparison with radiotherapy. It promotes decrease of pain syndrome intensity by 34.4-77.8% from the initial level in 75.5% of cases, while a course of radiotherapy results in regression of pain syndrome by 12.9-28.1% in 43.1% of patients.
  3. Radiotherapy does not change a rentgenologic picture of aggressive vertebral hemangioma and, hence, cannot avert the danger of compression fractures. Its antalgic effect is connected with non-specific anti-inflammatory effect of ionizing radiation, as well as with sclerosing of vessels and changes of hemodynamics at the vertebral plexus level. However, it demands use of smaller radiation doses.
  4. PVP is a little-invasive method, characterized by a minimum rate of complications and reducing a period of stay in hospital. Increase of a vertebral body strength and normalization of venous hemodynamics after PVP promote pain regression, better motor activity and resistance to severe static loads on the spine during the first week after operation.
  5. The threshold score of 5 and more (the Original Life Quality Scale) is an indication for puncture vertebroplasty.