Management of recurrent aggressive vertebral hemangiomas: a comparative study of piecemeal intralesional spondylectomy and modified total en bloc spondylectomy

Abstract

Background

The management of recurrent aggressive vertebral hemangiomas (AVHs) with neurological deficits poses a challenge due to their profuse vascularization and propensity for recurrence. The optimal approach to managing recurrence AVHs remains a topic of debate within the academic community.

Methods

A retrospective analysis was conducted on a cohort of patients who experienced recurrent AVH in the thoracic and lumbar spine at our institution from August 2015 to July 2022. A total of 31 patients, underwent treatment with either piecemeal intralesional spondylectomy (PIS) or modified total en bloc spondylectomy (MTES). Clinical documentation and surgical outcomes, including pre- and postoperative neurological function assessed by the American Spinal Injury Association (ASIA) score, Numeric Rating Scale (NRS), surgical duration, blood loss, recurrence rate, and complications, were retrospectively presented and analyzed.

Results

The study followed a cohort of 16 males and 15 females with recurrent AVHs in the thoracic or lumbar spine for a period ranging from 12 to 106 months. All patients presented with neurological deficits, with 20 patients undergoing PIS and 11 patients undergoing one-stage MTES. Both treatment modalities resulted in significant reductions in pain-NRS and ASIA scores. The MTES group experienced less intraoperative blood loss compared to the PIS group. During the follow-up period, four patients in the PIS group experienced recurrence. No serious complications or evidence of internal fixation failure were observed.

Conclusions

The modified TES technique presents a comprehensive resection, less blood loss, lower incidence of recurrence, and effective tumor control for recurrent AVHs in thoracic and lumbar spine through post-only approach.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12893-025-02942-5.

Background

Vertebral hemangiomas stand as the most prevalent primary spinal tumors, as evidenced by extensive autopsy studies indicating their presence in approximately 10–12% of the population’s spines. Despite their relative rarity, aggressive vertebral hemangiomas (AVHs) have an estimated incidence of 1% among all spinal hemangiomas [1]. Notably, AVHs are distinguished by their invasive nature, often extending into the spinal canal and/or paravertebral space, thereby causing local pain and neurological impairments. The primary treatment options for AVHs encompass radiotherapy, endovascular embolization, percutaneous vertebroplasty, and surgical tumor resection. Recurrence rates have been observed in cases treated with radiotherapy, embolization, and vertebroplasty. Surgical resection for patients with recurrent AVHs and accompanying neurological deficits frequently results in significant bleeding, complicating the surgical procedure and leading to numerous postoperative complications. Consequently, the optimal management strategy for recurrent AVHs remains a subject of ongoing debate within the medical community. Therefore, this study introduces an novel modified total en bloc spondylectomy (MTES) technique tailored for the management of recurrent AVHs in the thoracic and lumbar spine. The primary aim is to elucidate the procedural details of this technique and evaluate its safety and efficacy in treating locally recurrent AVHs.

Methods

Study design

A retrospective analysis was conducted on a cohort of 31 patients with recurrent AVHs in the thoracic and lumbar spine who were hospitalized at our institution from August 2015 to July 2022. The diagnosis of recurrent AVHs was established based on clinical manifestations, radiological evidence, and prior histopathological confirmation.

Inclusion criteria consisted of unendurable pain or neurological deficits, along with symptoms and imaging evidence of recurrent AVHs. Exclusion criteria encompassed primary hemangiomas, spinal tumors other than hemangiomas, and patients in poor health who were unable to tolerate surgery.

Retrospective reviews were conducted on the clinical records and imaging findings of the patients, with additional data collected on demographics, prior surgical history, and pathological diagnoses. All patients underwent X-ray, computed tomography (CT), and magnetic resonance imaging (MRI) scans of the thoracic or lumbar spine affected by the tumor. Embolization of the thoracic and lumbar segmental artery was performed in all cases based on imaging characteristics within 48 h before operation.

Surgical technique

We categorized all patients with recurrent aggressive intraspinal hemangiomas according to the Tomita classification system for spinal tumors [2]. Additionally, we evaluated the Spinal Instability Neoplastic Score (SINS) for each patient [3]. All patients presented with spinal instability accompanied by neurological symptoms attributable to tumor invasion. Treatment was provided to all patients via either piecemeal intralesional spondylectomy (PIS) or modified total en bloc spondylectomy (MTES).

Procedure of PIS

Using piezosurgery, the spinous process, lamina, and pedicles were carefully excised to expose and protect the dural sac. Subsequently, the bilateral thoracic nerve roots were identified, ligated, and transected, whereas the lumbar nerve roots were preserved. Ultimately, piecemeal excision was performed through intralesional spondylectomy.

Procedure of MTES

During the procedure of total spinal resection (Fig. 1), it is crucial to first isolate and ligate the segmental arteries associated with the affected vertebral body (only for thoracic spine) (Step One). Subsequently, the soft tissue surrounding the affected vertebral body should be meticulously separated from both sides using surgical gauze (Step Two). The next steps involve the removal of the posterior structures (Step Three) and the excision of the affected vertebral body (Step Four). This surgical procedure emphasizes the occlusion of blood supply to the affected vertebral body to minimize hemorrhage during the detachment of attachments and the spinal cord. This essential step constitutes the foundational aspect of the surgical technique. The thoracic nerve root can be ligated, while caution to protect the thoracic spinal cord, pleura, and lung tissue. Conversely, the lumbar nerve roots, which are crucial for lower extremity function, must be carefully preserved. The larger size of the lower lumbar vertebrae poses challenges for surgical separation, and the dissection of blood vessels anterior to the lumbar vertebral body is particularly intricate [4]. Consequently, ensuring adequate exposure of the lumbar nerve roots is of paramount importance.

Fig. 1.

Critical procedure of MTES. a Step One: remove the transverse processes and ligate the segmental arteries associated with the affected vertebral body. b Step Two: meticulously separate the soft tissue surrounding the affected vertebral body from both sides by surgical gauze. c Step Three and Four: subsequently remove the posterior structures and the affected vertebral body

After that, the disc and end-plate were cleaned, and a titanium mesh cage or expandable artificial vertebral body filled with allograft bone was inserted between the vertebral bodies for anterior column reconstruction. Titanium rods were then secured to the screws to reconstruct the posterior column structures.

Clinical evaluation

Demographic, disease-specific and surgery-related data were obtained, including age, gender, American Spinal Injury Association (ASIA) score, Numerical Rating Scale (NRS), surgical duration, blood loss, recurrence rate and complications. Data for pain-NRS and ASIA score were collected at pre-, post-operative (one week after surgery) and final follow-up time points. Recurrence rate and complications were collected according to patients visits and image examinations. All data were sourced from doctors, patients and medical records.

Statistical analysis

The data were summarized using descriptive statistics, with continuous variables expressed as mean ± standard deviations and categorical variables expressed as frequencies or percentages. Bilateral t-tests were employed to assess intergroup differences in continuous variables such as age, surgical duration, and follow-up time. Pearson’s chi-squared test or Fisher’s exact test was used to evaluate count data in intergroup differences. Due to the non-homogeneity of variances in blood loss statistics, the Wilcoxon rank-sum test was utilized to compare categorical data related to intraoperative blood loss, NRS scores, and ASIA impairment Scale classifications. Kaplan-Meier curves were plotted to illustrate recurrence time data. All statistical tests were two-sided, with a significance level set at p < 0.05. Statistical analyses were performed using SPSS software version 26.0 for Windows (IBM Corporation, New York, USA).

Results

Thirty-one patients with recurrent AVH were admitted to our department for treatment between August 2015 to July 2022 (Supplementary Table S1). Of these patients, 16 were male and 15 were female, with a mean age of 51.9 ± 16.9 years. The most commonly reported symptoms were unendurable pain or neurological deficits. All the patients experienced pain exceeding moderate level (NRS score ≥ 4). Among the patients, 24 exhibited myelopathy, 14 presented with radiculopathy, and 13 experienced urinary or defecation disorders. 19 individuals displayed two or more neurological symptoms. Previous surgical procedures included percutaneous vertebroplasty (n = 10), vertebroplasty with laminectomy (n = 15), and intralesional resection (n = 6). The time to first recurrence ranged from 2 to 360 months, with a mean of 77.0 ± 87.9 months. 27 patients had thoracic involvement, while four had lumbar involvement. Two patients had undergone radiotherapy following previous surgery.

Piecemeal intralesional spondylectomy (PIS) was performed on twenty patients (Fig. 2), and modified total en bloc spondylectomy (MTES) was performed on eleven patients (Fig. 3). There was no significant difference between the two groups in the demographic condition of patients. Analysis of the clinical characteristics revealed that the MTES group had significantly lower blood loss (904.6 vs. 1840.0 ml, p < 0.001) compared to the PIS group, with similar durations of surgery (4.95 vs. 5.55 h, p = 0.284) (Table 1).

Fig. 2.

A 67-year-old male presenting with lower extremity weakness and pain owing to a T8 vertebral hemangioma. a-d T1-weighted MRI and CT scan of a hemangioma before surgery. e and f Postoperative frontal and lateral radiographs of the thoracic vertebral hemangioma that has been performed vertebroplasty with laminectomy. g and h T2-weighted MRI show local recurrence after 6 months presenting with lower extremity paralysis and sphincter disturbance. i and j Postoperative frontal and lateral radiographs of the thoracic vertebral hemangioma that has been performed piecemeal intralesional spondylectomy

Fig. 3.

A 28-year-old female presenting with a T12 vertebral hemangioma causing progressive back pain. a and b Frontal and lateral radiographs of the thoracic vertebral hemangioma that has been performed vertebroplasty with laminectomy. c and d T1-weighted MRI show local recurrence three years after first operation. e and f Postoperative frontal and lateral radiographs of the thoracic vertebral show that has been performed modified total en bloc spondylectomy. g and h T2-weighted MRI of thoracic vertebra show no recurrence after four-year follow-up

Table 1.

Characteristics of patients with recurrent AVHs in the thoracic and lumbar spine who received piecemeal intralesional spondyloctomy (PIS) or modified total en-bloc spondylectomy (MTES)

Characteristic	Total	PIS	MTES	P Value
	(n = 31)	(n = 20)	(n = 11)
Age, years	51.9 ± 16.9	54.2 ± 17.3	47.8 ± 16.0	0.321
Sex				0.273
Males	16	12	4
Females	15	8	7
Previous surgical procedure				N/A
Percutaneous vertebroplasty	10	6	4
Vertebroplasty with laminectomy	15	10	5
Intralesional resection	6	4	2
History of radiotherapy	2	2	0	N/A
Location				N/A
Thoracic	27	17	10
Lumbar	4	3	1
Neurological deficits				N/A
Myelopathy	24	15	9
Radiculopathy	14	9	5
Sphincter disturbance	13	8	5
Time to first recurrence, months	77.0 ± 87.9	83.3 ± 100.1	65.7 ± 62.6	0.604
Duration of surgery, hour	5.16 ± 1.46	4.95 ± 1.70	5.55 ± 0.79	0.284
Blood loss, ml	1508.1 ± 972.8	1840.0 ± 1078.2	904.6 ± 96.1	< 0.001
Follow-up time, months	51.1 ± 28.8	50.2 ± 29.2	52.8 ± 29.4	0.810
Local recurrence				0.148†
Yes	4	4	0
No	27	16	11
Complication	5	2	3	0.317

N/A: Not assessed. Bold values denote statistical significance. †: Calculated according to Kaplan-Meier

Recurrence and complication

The mean follow-up time for all patients was 51.1 ± 28.8 months (range from 12 to 106 months). Analysis of recurrence revealed four patients in the PIS group and none in the MTES group experienced local recurrence during the follow-up period. However, these differences were not statistically significant (p = 0.148, Fig. 4). Pleural rupture occurred in two patients in the MTES group, while dural rupture occurred in two patients in the PIS group and one patient in the MTES group. There were no serious complications or evidence of internal fixation failure.

Fig. 4.

Kaplan–Meier curves of local recurrence time in the PIS and MTES groups. A log-rank test indicated no significant difference in recurrence (p = 0.148)

Change in pain-NRS and ASIA score

The comparison between the PIS and MTES groups showed no statistically significant differences in pain scores assessed by the NRS pain scores at pre-op, post-op and final follow-up (p = 0.730, 0.227, and 0.670). To control postoperative pain, all patients received opioid analgesic pumps within two days after surgery and non-steroidal anti-inflammatory drugs within seven days postoperatively. At both one week after surgery and at the final follow-up, the NRS pain scores were significantly reduced compared to pre-op levels in both the PIS and MTES groups. These findings indicate that both techniques are efficacious in providing rapid pain relief and reveal no significant differences in long-term pain alleviation (Table 2).

Table 2.

Changes over time in scores on the pain-NRS in the PIS and MTES groups

Characteristic	Total	PIS	MTES	P Value*
	(n = 31)	(n = 20)	(n = 11)
Pre-op
0	0	0	0	0.730
1–3	0	0	0
4–6	21	13	8
7–10	10	7	3
Post-op (1 week after surgery)
0	0	0	0	0.227
1–3	20	11	9
4–6	11	9	2
7–10	0	0	0
P value†	< 0.001	< 0.001	0.006
Final follow-up
0	16	11	5	0.670
1–3	15	9	6
4–6	0	0	0
7–10	0	0	0
P value‡	< 0.001	< 0.001	0.003

*: Comparison between the PIS group and MTES group. †: Comparison between pre-op and post-op (1 week after surgery). ‡: Comparison between pre-op and final follow-up. Bold values denote statistical significance

A comparative analysis of the PIS and MTES groups indicated no statistically significant differences in ASIA scores at pre-op, post-op and final follow-up (p = 0.792, 0.451, and 0.730, respectively). However, the ASIA scores at both the post-op and final follow-up stage showed significant improvement compared to the pre-op stage (p = 0.0035 and < 0.001, respectively). In evaluating the outcomes of the two distinct surgical approaches (PIS and MTES), we identified significant differences in patients’ post-op ASIA score improvements. Specifically, patients undergoing the MTES procedure demonstrated more rapid neurological functional recovery in the immediate post-op period compared to those receiving PIS (PIS p = 0.157 and MTES p = 0.046). Nevertheless, both groups exhibited significant improvements in neurological function when assessed for long-term neurological recovery relative to the preoperative period (PIS p = 0.005 and MTES p = 0.02). These findings underscore the superiority of the MTES technique in achieving more comprehensive decompression compared to PIS. However, long-term follow-up results indicate that both approaches yield comparable outcomes in terms of long-term neurological functions (Table 3).

Table 3.

Changes over time in scores on the ASIA in the PIS and MTES groups

Characteristic	Total	PIS	MTES	P Value*
	(n = 31)	(n = 20)	(n = 11)
Pre-op
A	3	2	1	0.792
B	6	5	1
C	10	5	5
D	12	8	4
E	0	0	0
Post-op (1 week after surgery)
A	2	1	1	0.451
B	6	5	1
C	8	5	3
D	12	8	4
E	3	1	2
P value†	0.035	0.157	0.046
Final follow-up
A	0	0	0	0.730
B	5	3	2
C	8	6	2
D	11	7	4
E	7	4	3
P value‡	< 0.001	0.005	0.02

*: Comparison between the PIS group and MTES group. †: Comparison between pre-op and post-op (1 week after surgery). ‡: Comparison between pre-op and final follow-up. Bold values denote statistical significance

Discussion

Vertebral hemangiomas are benign lesions composed of blood vessels intertwined with various tissues [5]. They are histologically classified into capillary, cavernous, arteriovenous, and venous subtypes [6], all capable of causing neurological symptoms without distinct radiographic differences [7]. Diagnosis of AVHs relies on imaging, particularly CT and MRI, which categorize them as typical or atypical [8]. Typical imaging shows a honeycomb pattern (“polka-dot sign”) on CT, with characteristic MRI signals confined to a single vertebral body. Atypical findings, more common in AVHs, show extensive bone destruction, paravertebral mass, and intraspinal involvement.

Radiation therapy, especially stereotactic, is key for symptomatic spinal hemangiomas, with a 65.0-79.2% improvement rate in neurological symptoms [9–12]. This efficacy is due to radiation sensitivity of vascular endothelial cells, leading to vessel sclerosis, narrowing, and hemangioma atrophy. Radiotherapy is indicated for painful or symptomatic hemangiomas, with a recommended dose of 30–40 Gy [13]. However, it has risks such as radiation myelitis, sarcomatous transformation, nerve damage exacerbation, local edema, and increased surgical complications [14]. Compared to surgery, radiotherapy offers slower and less complete symptom improvement. Currently, with advancements in surgical techniques and the enhancement of surgical auxiliary facilities, coupled with the numerous complications associated with radiation therapy and its delayed onset, radiation therapy is no longer the primary treatment option for symptomatic hemangiomas accompanied by spinal cord compression. However, for patients experiencing symptom relief but incomplete tumor resection, radiation therapy may be considered as an adjunctive therapy, effectively reducing the recurrence rate [14].

Percutaneous vertebroplasty, recognized for its ease of operation and minimal invasiveness, is widely utilized in the treatment of spinal hemangiomas [15]. This treatment has shown effectiveness in alleviating pain associated with hemangiomas and stabilizing the spine. A multicenter study conducted by Goldstein [16] reported a 4-year recurrence rate of approximately 5.4% following vertebroplasty for invasive hemangioma, with 10% of patients also receiving local radiotherapy. While vertebroplasty surgery for AVHs offers benefits such as reduced bleeding and minimal trauma, it is important to note that there are significant limitations associated with this procedure, including the risk of bone cement leakage causing spinal cord injury and its unsuitability for cases with incomplete vertebral posterior walls or soft tissue masses causing spinal cord compression. To address these challenges, some surgeons have adopted intraoperative vertebroplasty during surgical decompression and instrumentation for AVH patients [17], aiming to reduce bleeding, surgical duration, and recurrence risk.

Tumor resection surgery is regarded as a vital approach for the treatment of AVHs with neurological deficits. Numerous studies have demonstrated that, various surgical techniques, including separation surgery, intralesional resection, and total en bloc spondylectomy, are both effective and safe for managing AVHs. However, patients with AVHs undergo surgical treatment followed by adjuvant radiotherapy continue to face a variable risk of recurrence. A study of 11 AVH cases reported a 10-year recurrence rate of 27.3% after subtotal tumor resection and postoperative radiation therapy [18]. Pastushyn [19] observed a 25% local recurrence rate in 64 spinal invasive hemangioma cases, with 61% receiving radiotherapy. Jiang [20] reported a recurrence rate of 14.3% in 21 patients with AVH who underwent surgical treatment, despite 66.7% of these patients receiving adjunctive radiotherapy. In our study, the PIS group experienced 4 recurrences, corresponding to a 20% recurrence rate.

Total en bloc spondylectomy (TES), a well-established and intricated surgical technique, is employed in the treatment of primary spinal tumors [21]. The principal advantage of TES lies in its superior oncologic outcomes. Unlike intralesional resection, TES entails the complete excision of the vertebral body, thereby achieving a high rate of negative surgical margins. Consequently, TES can substantially reduce the local recurrence of primary spinal tumors, which has been repeatedly demonstrated in previous studies [22–25]. In patients with isolated spinal metastases, total en bloc resection of solitary metastases is associated with significantly lower recurrence rates and prolonged survival compared to intralesional surgery [26–28]. To date, no studies have yet compared en bloc resection with intralesional resection for the treatment of AVH. Nevertheless, relevant studies have supported that TES offers advantages in reducing recurrence rates and improving patient prognosis. Inoue [29] was the first to describe total spondylectomy in a patient with a symptomatic hemangioma of the lumbar spine, who remained completely asymptomatic with no evidence of tumor recurrence at a four-year follow-up. A study examining 23 cases of AVH treated with total spinal resection, reported no recurrences during the follow-up period [30]. Similarly, no recurrences were noted in the MTES group in our study, although statistical analysis revealed no significant difference between the MTES and the PIS groups.

A significant challenge in the surgical treatment of AVHs is the tumors’ extensive vascularization, which can lead to considerable intraoperative blood loss. Acosta [31] reported an average blood loss of 2100 ml (range, 0.8–5 L) in 10 spondylectomy cases using intralesional resection. While en bloc resection entails the complete removal of the tumor along with surrounding healthy tissue, it is also associated with significant blood loss in the treatment of AVH. Kato [32] reported an average blood loss of 2424 ml in five AVH cases treated with total spinal bloc resection, with one requiring revision surgery. Ji [30] presented a treatment of AVH by total spinal resection, with an average blood loss of 1883.3 ml, including two of them without preoperative embolization experienced blood losses of 2800 ml and 4000 ml, respectively.

It is important to note that, in various TES techniques, the initial phase typically involves laminar decompression, which can lead to considerable blood loss due to the venous plexus and tumor hemorrhage. We have enhanced the surgical procedure by implementing early ligation of the vertebral feeding vessels. This refined surgical step is crucial in minimizing blood loss and reducing postoperative recurrence rates in the treatment of AVHs. Considering the significant blood supply involved, we have further developed the surgical technique to mitigate these risks. Optimizing surgical technique to minimize both bleeding and recurrence rates represents a crucial component in the management of AVHs. Currently, there is a notable lack of literature focusing on the surgical treatment of recurrent AVHs. The recurrence of AVHs presents substantial therapeutic challenges, primarily due to previous surgical interventions and the resultant alterations in local anatomical structures. This study proposes a modified total en bloc spondylectomy technique aimed at improving neurological outcomes and reducing postoperative recurrence rates in patients with manageable blood loss. This refined technique enables more effective management of the therapeutic process for AVH patients, reducing complications, enhancing surgical success rates, and improving patients’ quality of life.

Conclusions

The use of the MTES technique in managing recurrent AVHs in the thoracic and lumbar spine has been demonstrated to be a safe and efficacious surgical strategy. This approach enables complete decompression of the spinal cord, improve preoperative nerve compression symptoms, minimizes intraoperative hemorrhage, and diminishes the likelihood of postoperative recurrence.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

AVHs

Aggressive vertebral hemangiomas

PIS

Piecemeal intralesional spondylectomy

MTES

Modified total en bloc spondylectomy

ASIA

American Spinal Injury Association

NRS

Numeric Rating Scale

CT

Computed tomography

MRI

Magnetic resonance imaging

SINS

Spinal Instability Neoplastic Score

VL

Vertebroplasty with laminectomy

PV

Percutaneous vertebroplasty

IR

Intralesional resection

SCST

Surgical classification of spinal tumor

Author contributions

All authors contributed to the study conception and design. DJJ, ZHZ and QZ collected the data and wrote the main manuscript. DJJ, SHH and HTS prepared the figures. QZ, ZHZ and XPC prepared the tables. JRX, HFW, TLL and XHY performed the surgeries. JRX and HFW supervised the project and made a critical revision. All authors read and approved the final manuscript.

Funding

The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Data availability

The datasets utilized and analyzed in the present study can be obtained from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the principles outlined in the Declaration of Helsinki. The study was approved by the Medical Ethics Committee of Shanghai Changzheng Hospital of Naval Medical University. Due to its retrospective design, the study was granted an exemption from the requirement for informed consent by the medical ethics committee.

Consent for publication

The manuscript has been approved for publication by all the authors. Written informed consent has been obtained from the patients and/or their legal guardians for the publication, as well as for any accompanying images, sex, and age of these patients.

Competing interests

The authors declare no competing interests.