Simple transpedicular vertebral biopsy for diagnosis of malignancy in vertebral compression fracture
Correspondence Address: Source of Support: This paper was supported in part by grant B1021027 from the Teh.Tzer Study Group for Human Medical Research Foundation, Taipei, Taiwan, Conflict of Interest: None DOI: 10.4103/0028-3886.125249 Clinical trial registration TSGHIRB 1-101-05-086
Source of Support: This paper was supported in part by grant B1021027 from the Teh.Tzer Study Group for Human Medical Research Foundation, Taipei, Taiwan, Conflict of Interest: None
Clinical trial registration TSGHIRB 1-101-05-086
Background: The transpedicular route in percutaneous vertebroplasty (PVP) is a well-established approach for the treatment of vertebral compression fractures (VCFs). However, the value of simple transpedicular biopsy in VCFs is less addressed. The purpose of this study is to evaluate the value of transpedicular biopsy during PVP for uncovering the malignancy in VCFs in a 10-year retrospective study. Materials and Methods: During the study period of the 1019 patients who underwent PVP for VCFs, 450 patients comprising of 127 male and 323 female underwent transpedicular biopsy during PVP for 705 fractured vertebras. The medical records were analyzed for age, gender, imaging studies, operation notes, pre-operative and post-operative diagnoses, date of vertebroplasty and biopsy, vertebral level and pathological reports. Results: Pathology of the specimens of the 450 patients confirmed non-malignant VCFs in 389 (86.44%) and malignancy in 61 (13.56%). The malignant pathology included: 52 (11.56%) distant metastases to vertebra, in 3 (0.67%) of the spinal metastases was unsuspected and in 49 (10.89%) of them the malignancy was suspected pre-operatively. There were 9 (2%) primary spinal malignancies, 2 (0.44%) unsuspected multiple myeloma and 7 (1.56%) pre-operatively suspected primary malignancies. The frequency of unsuspected malignancy was 1.11% (5/450) in this study. There was no complication associated with transpedicular biopsy during PVP. Conclusions: VCFs harbored 1.11% of unexpected malignancy. During the vertebroplasty, concomitant transpedicular vertebral biopsy is a safe and useful procedure for distinguishing non-malignant from malignant compression fractures, especially in diagnosing unsuspected malignancy.
Keywords: Incidence, malignancy rate, percutaneous vertebroplasty, transpedicular vertebral biopsy, vertebral compression fractures
Vertebral compression fractures (VCFs) often result in intractable back pain and limited mobility resulting in medical comorbidities such as constipation, urinary tract infections and pressure sores. Percutaneous vertebroplasty (PVP) involves injection of polymethylmethacrylate bone cement into the fractured vertebral body and is an accepted procedure effective pain relief. , In some circumstances, VCFs are recognized as the results of reduced vertebral stiffness and strength, which predispose vertebras to fracture, even with relatively minor-energy trauma. These conditions are observed in patients with low bone density, osteomalacia or lytic malignant or infectious bone lesions.  Characteristic magnetic resonance imaging (MRI) features of VCFs differentiate osteoporosis from malignancy.  The reported rate of malignancy in osteoporotic VCFs ranges from 0.4% to 7.3% respectively. ,,,, Malignant VCFs are strongly suspected when imaging reveals: presence of convex contour of the posterior vertebral border, epidural or paravertebral masses and infiltration of the vertebral posterior elements. , Retropulsion of a posterior bone fragment, preservation of normal bone marrow and accumulation of fluid within vertebral bodies are commonly observed in osteoporotic compression fractures.  However, there is some overlap between the imaging characteristics of non-malignant and malignant VCFs using above imaging features, thus making pre-operative diagnosis difficult. , Therefore, histological confirmation is essential for rational post-operative management. Transpedicular route in PVP is a well-established approach for the treatment of VCFs. The purpose of this study is to evaluate the value of transpedicular biopsy during PVP for detecting unsuspected malignancy in VCFs.
The study was approved by the Institutional Review Board of Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan (TSGHIRB approval Number: 1-101-05-086). During December 2002 and February 2011 of the total 1019 patients who underwent PVP for VCFs, 450 patients 127 male, 323 female; range of peak age, 71-80 years) had transpedicular biopsy during PVP for 705 fractured vertebras. The medical records were retrospectively analyzed for age, gender, pre-operative and postoperative diagnoses, date of vertebroplasty and biopsy, vertebral level and pathological reports. Unsuspected malignant VCF was defined when initial diagnosis of malignancy by both neuroradiologists and neurosurgeons as less likely from pre-operative imaging studies or medical records, however the diagnostic possibility of malignancy has not been totally excluded. Malignancy in VCF was established by histopathologists.
All patients with intractable painful VCFs were evaluated pre-operatively by plain X-rays of spine, contrast MRI, computed tomography, or bone scintigraphy. Acute nature of the fracture was determined by increment of radiotracer activity. The inclusion criteria for transpedicular biopsy included patients with the previous history of malignancy, or suspected vertebral malignancy in the pre-operative imagining evaluation. ,,,,
Bone specimens were evaluated by the histopathologists. The bone pathology was divided into: Non-malignant, primary malignant and metastatic lesions. The pathological features, mature bone, degenerative tissue, tissues with acute or chronic inflammation, ﬁbrocartilage and fibrosis and necrosis without any histological malignant features such as nuclear atypia, hypercellularity, or increased mitosis were considered for the diagnosis of non-malignant VCFs. Primary malignant VCFs were defined as a primary spinal malignancy without evidence of systemic metastasis or a tumor originating at another site. The metastatic lesions were identical to the immunophenotypes of the primary tumors.
Patient was placed in a prone position for general anesthesia. Fracture level and puncture points overlying the pedicles were ﬂuoroscopically localized. An 8-gauge biopsy needle was inserted through the pedicles to the bone marrow of the vertebral body under fluoroscopic X-ray guidance. A pair of 2-mm biopsy pliers was then inserted through the cannula of the biopsy needle into the vertebral body and biopsy specimen was obtained by twisting the pliers. The biopsy site depended on suspicious areas identified by pre-operative imaging studies. The biopsy site was also the site for injection of the bone cement (Stryker Simplex P Bone Cement) that was mixed until a liquid consistency suitable for injection was obtained. The mixture was then injected into the selected vertebral body under ﬂuoroscopic monitoring. The needles were then removed and the puncture site was covered with a small bandage.
The Chi-square (χ2 ) was used for statistical analyses, which were evaluated as significant when P < 0.05.
Single-level VCF were observed in 298 patients and multiple-level VCFs in 152 patients. Medical records documented presence of malignancy in 101 patients before the occurrences of VCFs. A total of 1019 patients underwent PVP for VCFs [Figure 1]. Of the 1019 patients who underwent PVP for VCF [Figure 1], 450 patients with 705 VCFs underwent transpedicular biopsy. Pathology was non-malignant in 389 patients and malignancy in 61 patients, 52 patients with distant metastases to vertebra (3 unsuspected spinal metastases, 49 with pre-operative diagnosis of malignancy) and primary spinal malignancy in 9 patients (2 unsuspected multiple myeloma, 7 diagnosed malignancy preoperatively). Thus, the frequency of unsuspected malignancy identified in this study was in 5 (1.11%) patients. None of the patients had any complications associated with transpedicular biopsy during PVP.
Non-malignant vertebral compression fractures
Of the 389 patients, 97 were male and 292 were female with age range between 71 and 80 years with non-malignant VCFs; 61 patients had a history of pre-operative malignancy and 123 had VCFs at multiple levels. The most common involved level was L1, followed by T12 and L2 [Table 1]. Distribution of VCF levels was bimodal with a minor peak at T8 and a major peak at L1 [Figure 2]. In none of them had benign bone tumor.
Primary spinal malignancy
Multiple myeloma was the pathological diagnosis in all the 9 patients (6 males and 3 females with peak age of 61-70 years. 4/9) with primary spinal malignancy. One patient had the pre-operative history of multiple myeloma and in 8 patients the diagnosis was confirmed by serum and urine screen and all the 9 patients were treated by hematologists. Only one patient underwent a first PVP for a VCF at a single level. Majority of the VCFs involved levels T12, L2, and L3 [Figure 2].
Metastases were the pathological diagnosis in A 52 patients (24 male, 28 female, common age group 61-70 years, 23/53) with 79 fractures. Thirty-two patients were treated for a VCF at a single level, (mainly at T11-L4) [Figure 1]. Among 49 patients, the diagnosis of malignancy was suspected or diagnosed pre-operatively. Progression of the original tumor stage with distant metastases was found after biopsy in 21 of these patients. Primary site malignancy was adenocarcinoma of lung (n = 13), adenocarcinoma of breast (n = 11), adenocarcinoma of colon (n = 6), oral squamous cell carcinoma (n = 3), squamous cell carcinoma of the lung (n = 2), hepatocellular carcinoma (n = 2), prostate cancer (n = 2), cardiac angiosarcoma (n = 1), T-cell lymphoma (n = 1), non-small cell lung cancer (n = 1), cholangiosarcoma (n = 1), adenocarcinoma of pancreas (n = 1), adenocarcinoma of stomach (n = 1), cervical squamous cell carcinoma (n = 1), papillary thyroid cancer (n = 1), renal cell carcinoma (n = 1), pre-B acute lymphoblastic leukemia (n = 1), urothelial cell carcinoma (n = 1), retroperitoneal fibrous histiocytoma (n = 1) and endometrioid adenocarcinoma (n = 1) [Table 2]. There were three unsuspected spinal metastases: Adenocarcinoma of the breast (n = 1), colon cancer (n = 1) and T-cell lymphoma (n = 1), respectively.
The distribution of spinal fracture segments (thoracic vs. thoracolumbar and lumbar) in non-malignant, primary malignant and metastatic vertebral compression fractures
The distributions of non-malignant VCFs were predominately in the thoracolumbar and lumbar spine (n = 520) and above the thoracic segments (n = 84, P < 0.001). The distribution of pathological VCFs (including primary spinal malignancy and metastases) was similar to the non-malignant VCFs, being most common at the thoracolumbar and lumbar spine [Table 3]. The distributions of vertebral metastases-associated VCFs were significantly higher in the thoracolumbar and lumbar spine (n = 63) than thoracic segments (n = 11, P < 0.001).
The relationship of multiple-level involvement in non-malignant, primary malignant and metastatic vertebral compression fractures
Multiple-level involvement was found in 80 (20.6%) patients of the 389 patients with non-malignant VCFs. It was in 8 (88.9%) patients with primary malignancy, in and 20 (38.5%) patients with vertebral metastases. Statistically, the multiple-level involvement was more common in primary malignancy, and vertebral metastases when compared with non-malignant VCFs [P < 0.001, [Table 4]].
The overlap of imaging characteristics of malignant and non-malignant VCFs adds to the difficulty of pre-operative distinction between non-malignant and malignant VCFs. , Knowledge of the pathology of VCFs is important in planning the subsequent therapeutic strategy. Routine vertebral body biopsy during PVP for VCFs has been recommended to avoid another invasive procedure for biopsy and it also helps to establish the diagnosis of any unsuspected pathology. The reported rates of unsuspected malignancy range from 0.4% to 7.3% in patients with osteoporotic VCFs treated with PVP, including primary malignancy and spinal metastases. ,,,, An 1.11% (5/450) rate of unsuspected malignancy was identified in our study, including three patients with spinal metastases and two patients with multiple myeloma. Vertebral biopsy may also help in differential diagnosis.  Doing routine vertebral biopsy in patient with VCFs is worth considering as it helps in establishing the pathology of VCF and may also increase the reported rates of unsuspected pathology.
Our study showed consistent fracture foci peaking in the mid-thoracic spine at T8 level with a second focus at the thoracolumbar junction L1 level. Importantly, the distribution of malignant VCFs was similar to the incidence of non-malignant VCFs, further supporting the importance of transpedicular vertebral biopsy for diagnosis of malignancy in VCFs. Consistently, previous studies of architectural strength and biomechanics of vertebra have shown that these regions bear the major axial compressive loading. ,, Non-malignant VCFs predominately occurred over the thoracolumbar junction and lumbar spine, especially around the thoracolumbar junction (T11-L2), the transition region between the rigid thoracic kyphosis and the flexible lumbar lordosis; the curvature predisposes this region to fracture. The vertebral bodies at this junction are smaller than those at the lower lumbar spine, which is correlated with their ability to resist deformity. The rib cage provides spinal support due to which fewer VCFs occur in the thoracic spine (T1-T10). In addition to the thoracolumbar transition and the weight-bearing lumbar segment, axial compressive forces concentrate on the apex of the kyphotic curves of the spine at the mid-thoracic segment centered at the T8 level.
Multiple myeloma is the most common primary vertebral tumor, accounting for 26% of primary vertebral tumors.  Spinal lesions have a tendency to occur in the vertebral body and lower thoracic and upper lumbar spine rather than in other spinal regions.  In our study, multiple myeloma was confirmed after transpedicular biopsy during PVP in nine patients with 17 VCFs. Distribution of these pathological VCFs was predominately below the T12 level and rarely occurred above the upper thoracic spine. The thoracic spine is the most common site of spinal metastases due to its greater vertebral volume relative to the cervical and lumbar segments.  However, in our study, most vertebral metastases with VCFs occurred in the thoracolumbar and lumbar segments. In our population, non-malignant VCFs were predominately occurred in the thoracolumbar and lumbar spine because these are the major weight-bearing segments and they are not supported by the rib cage. However, distribution of pathological VCFs based on primary spinal malignancy and metastases was similar to that of the non-malignant VCFs, being most common at the thoracolumbar and lumbar spine. This finding is in conflict with previous reported findings. The increased volume and venous drainage of a lumbar vertebra may contribute the condition. The clinical presentation of multiple fracture levels was reported not to be helpful in differentiating between osteoporosis and malignant VCFs. , However, in this study multiple fracture levels were significantly high in malignant VCFs.
There were several limitations to the current study. Firstly, it was a retrospective study performed at a single medical center. The medical records analyzed also did not include bone mineral density measurements; thus we were not able to quantitatively study the relationship between osteoporosis and VCFs. Moreover, VCFs that had occurred within the preceding 2-5 months of surgery and might have altered signal intensity on MRI  this has limitation to exactly determine the age of the fracture in patients with multiple level VCFs. Transpedicular biopsy was not a routine practice and were done mostly in patients with image findings suggestive of malignancy or in those with pre-operatively diagnosed malignancy. In addition, there was no long-term follow-up of patients with VCFs with negative histological findings.
Transpedicular vertebral biopsy is a simple, safe and useful procedure for distinguishing non-malignant from malignant compression fractures, especially in discovering unexpected malignancy during vertebroplasty. We showed that the incidence of unexpected malignancy was 1.11%. In this study, we found that distribution of non-malignant VCFs in our population was similar to that in previously published studies with one major peak at the thoracolumbar segment and one minor peak at the mid-thoracic segment. However, distribution of malignant VCFs reported in previous studies differed from that observed among our patients. The greatest incidence of malignant and metastatic VCFs was on the thoracolumbar and lumbar spine in our study. Similar distribution of the non-malignant and malignant VCFs identified in our study indicates the necessary for further investigations to determine the reasons for the similarity to formulate a suitable therapeutic strategy for VCFs. Further prospectively perform a biopsy systematically would pick up more unexpected malignancy.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4]