October 2006, VOLUME 187

Recommend & Share

October 2006, Volume 187, Number 4

Musculoskeletal Imaging

Original Research

Response of Nonmidline Pain to Percutaneous Vertebroplasty

+ Affiliations:
1Department of Internal Medicine, Barnes-Jewish Hospital, St. Louis, MO.

2Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S Kingshighway Blvd., St. Louis, MO 63110.

Citation: American Journal of Roentgenology. 2006;187: 869-872. 10.2214/AJR.05.0084

Next section

OBJECTIVE. The purpose of this study was to determine the prevalence of nonmidline symptoms in patients undergoing percutaneous vertebroplasty and the response of these symptoms to vertebroplasty.

MATERIALS AND METHODS. This is a retrospective study performed through examination of patient records, baseline questionnaires, demographic sheets, follow-up questionnaires, reports from telephone follow-up, and pain location diagrams completed before and after percutaneous vertebroplasty. Data were gathered from 350 patient encounters, 686 vertebroplasty procedures, and 288 patients. After determining the prevalence of nonmidline pain, analysis of acquired data was performed to determine the efficacy of vertebroplasty in relieving nonmidline pain.

RESULTS. Nonmidline pain was present in 240 of 350 patient encounters. Major symptom areas were the ribs; hip, groin, and buttocks; and legs and thighs. Lesser areas of nonmidline symptoms were the abdomen, shoulders, and waistline. Overall there was improvement in nonmidline pain in 83% of the procedures.

CONCLUSION. Most patients presenting for percutaneous vertebroplasty have nonmidline pain. Vertebroplasty reduced or eliminated nonmidline pain in 76-92% of the patients in whom it was reported. Procedures involving lumbar vertebrae tended to be successful less often than others, but even in these patients the procedure was successful in approximately 75%.

Keywords: fracture, nonmidline pain, osteoporosis, percutaneous vertebroplasty, spine, vertebroplasty

Previous sectionNext section

Percutaneous vertebroplasty is an effective treatment for painful vertebral compression fractures arising from a variety of causes, including osteoporosis [1-7], metastases and myeloma [5, 8-10], and hemangiomas [11, 12]. Several authors maintain that the ideal candidate for vertebroplasty has focal nonradiating back pain [11, 13-17]. Others specify that vertebral compression fractures typically lack a radicular quality [6, 18], specifically osteoporotic fractures [19, 20]. According to Stallmeyer and Zoarski [20], radiculopathy is not a contraindication to percutaneous vertebroplasty. The authors caution patients, however, that their radiating or referred symptoms may not be successfully treated and could potentially worsen as a result of the treatment.

Several articles that have made note of nonmidline pain are reviewed in the Discussion. Possibilities for nonmidline pain include radicular pain, which is a well-discriminated neurogenic radiating pain, and referred pain, which is a diffuse nonspecific pain [21].

The main purpose of this study was to determine the prevalence of nonmidline symptoms in patients undergoing percutaneous vertebroplasty and the success of treatment to decrease or eliminate these symptoms. We chose to assess the response of nonmidline symptoms to percutaneous vertebroplasty because we do not believe this has been assessed in the literature.

Materials and Methods
Previous sectionNext section

Vertebroplasty was performed using a standard technique as described in an article published in 2004 [22]. Fluoroscopic control was used with a C-arm unit with patients prone in a radiology suite. Conscious sedation was used in all patients. Most patients were observed for approximately 1 hour after the procedure to be certain they were stable before being discharged. From the 350 patient encounters, 686 levels were treated on 288 patients for vertebral pain thought to be related to fracture or involvement with neoplastic disease. The patients were referred to be evaluated for percutaneous vertebroplasty from spine surgeons, internists, physiatrists, and oncologists. Diagnosis of vertebral compression fractures or neoplastic involvement of bones was made using radiographs supported by physical examination and MRI, bone scanning, or both. Exclusion criteria were patients with pain without fracture or neoplasia involving one or more vertebral segments, unstable fracture, infection, and vertebral body involvement impinging on cord or nerve with continual true radicular pain. Institutional review board approval and patients' informed consents were both obtained, which allowed completion of a baseline questionnaire, follow-up questionnaires, and telephone questionnaires. Follow-up was performed at 2 weeks, 1 month, 3 months, 6 months, 1 year, and 2 years; however, follow-up data were not obtained for all time intervals from all patients. A research assistant not involved with this study performed telephone interviews to avoid response bias. The assistant used an established questionnaire approved by our institutional review board. The paper questionnaires were completed by the patients without input from a medical professional to avoid response bias.

View larger version (50K)
Fig. 1 Diagrams showing nonmidline major pain areas on the body.

This is a retrospective study performed through examination of patient records, baseline questionnaires, demographic sheets, follow-up questionnaires, reports from telephone follow-up, and pain location diagrams completed before and after the percutaneous vertebroplasty procedure was performed. These forms were used to determine whether the patient had nonmidline pain before the procedure and where the nonmidline pain occurred. We only evaluated patients with nonmidline pain and did not objectively assess midline pain.

Patients were asked about nonmidline pain on the baseline questionnaire. Before January 19, 2000, the patients from patient encounters 1-99 were asked “Does the pain travel down your arms or legs?” After that date, patients from patient encounters 100-350 were asked “Does the pain travel into your arms, ribs, hips, or legs? If yes, please state where.” For patients who answered yes to this question, or if they indicated on their pain location diagram that their pain traveled beyond the midline of their back, nonmidline pain was recorded. It was also recorded if the patient indicated paravertebral soreness or if their pain only occasionally was nonmidline to another anatomic site; these patients were not included in the group considered to have nonmidline pain and were excluded from analysis. Other exclusions were two patients with a multitude of health problems, making it impossible to clearly state if their pain was related to their vertebral compression fractures. Also, one patient's preprocedure information was missing and could not be obtained.

Seven categories of nonmidline pain were generated from the patient questionnaires and diagrams: abdomen; flanks; hip, groin, and buttocks; legs and thighs; ribs, sides, and chest; shoulders; and waistline. Many patients specified where they had nonmidline pain on the baseline questionnaire. In addition, patients completed a pain location diagram to further assist us in assessing areas of pain involvement. Figure 1 shows representative pain diagrams used by the patients. The areas on the diagrams were compiled and drawn to cover areas indicated by patients. Specific drawn areas were not given to patients to avoid bias. Only those who had nonmidline pain and who specified the area of their nonmidline pain were considered to have sufficient information for use in further analysis. The fracture region treated was recorded using the following definitions: thoracic (T1-T12), lumbar (L1-L5), and both.

Written questionnaires, pain location diagrams, and telephone questionnaires were also used to determine if the patients felt relief from the nonmidline pain after treatment with percutaneous vertebroplasty. Outcomes were reported as pain gone, pain better, pain the same, and pain worse. For analysis, the first two categories were combined into improved and the last two into unimproved. An indication of relief was given only if the patient specifically mentioned the nonmidline pain they reported before the procedure.

Data were analyzed by calculating frequencies and constructing contingency tables. Patterns found in the contingency tables were tested for statistical significance with chi-square tests.

Previous sectionNext section

Of the 288 patients treated with 350 patient encounters, 198 were women and 90 were men (mean age, 71.4 years; median age, 73.8 years; age range, 19.7-102.5 years). The underlying causes of fracture included osteoporosis (211/288, 73%), metastasis or myeloma (69/288, 24%), trauma (5/288, 2%), and hemangioma (3/288, 1%).

Of the 350 procedures, 240 (69%) occurred in 210 patients who had reported nonmidline pain before the procedure. Follow-up on the pain outcome after percutaneous vertebroplasty occurred in 206 of the 240 (86%) procedures. All 34 procedures missing follow-up were secondary to interval patient death. Two cases (1%) were missing both pain outcome and specific region of pain; for all others complete information was provided for the procedures and the pain outcome. In 90 of 206 (44%) procedures, one fracture level was treated; in 68 of 206 (33%), two fracture levels were treated; in 32 of 206 (16%), three fracture levels were treated; in 13 of 206 (6%), four fracture levels were treated; and in three of 206 (1%), five fracture levels were treated. By spinal region, 76 of 206 (37%) procedures were performed in the lumbar region, 83 of 206 (40%) were performed in the thoracic region, and 47 of 206 (23%) included levels in both the lumbar and thoracic regions. The procedure was the patient's first in 164 of 206 (80%) procedures, their second in 37 of 206 (18%) procedures, and their third in five of 206 (2%) procedures. In 202 of 206 (98%) procedures, patients specified the region of pain. The majority of these patients (135/202, 67%) reported just one region of pain, although a substantial minority (55/202, 27%) reported two, and a few (12/202, 6%) reported three regions of pain. Table 1 shows the distribution of pain by regions.

TABLE 1: Distribution of Nonmidline Pain by Anatomic Region

The average length of follow-up was 237 days (median, 182 days; range, 8-1,141 days). Interval follow-ups occurred during that time, but only the latest follow-up dates referring to nonmidline pain were used for this study. The range of follow-up for 14 patients extended beyond the previously stated 2-year follow-up range. Ten of the 14 patients were in the group for whom institutional review board permission was given for further follow-up. The first follow-up questionnaires for three of 14 patients did not get mailed until 2.5 years after the procedure and were received approximately 2.8 years afterward. The remaining patient had a subsequent treatment so her final follow-up was more than 2 years after her first procedure, but less than 2 years after her final treatment.

Vertebroplasty was effective in treating nonmidline pain. Overall, 172 of 206 (83%) procedures resulted in reduction of nonmidline pain. The success rate was not affected by the number of fracture levels treated (chi-square test, p = 0.67), as shown in Table 2. Treatments of the thoracic spine were somewhat more successful than treatments of the lumbar spine (Table 3). Although the difference was not large, it achieved statistical significance (chi-square test, p = 0.02). The number of treatment sessions had no effect, with improved pain status reported in 135 of 164 (82%) first sessions and 37 of 42 (88%) subsequent sessions (chi-square test, p = 0.35). Pain relief did not vary with region of pain at a statistically significant level, but trends were seen for the hip, groin, and buttocks region to have a somewhat lower rate of success than other regions and for the ribs, side, and chest region to have a somewhat higher rate (Table 4) (chi-square test for hip, groin, buttocks vs combined others, p = 0.050; for ribs, side, chest vs combined others, p = 0.067; and for hip, groin, buttocks vs ribs, side, chest, p = 0.027).

TABLE 2: Reported Improvement in Nonmidline Pain by Number of Treated Fracture Levels

TABLE 3: Reported Improvement in Nonmidline Pain by Region of Treated Fracture

TABLE 4: Reported Improvement in Nonmidline Pain by Reported Anatomic Regions of Pain

Previous sectionNext section

Many of our patients in need of percutaneous vertebroplasty reported nonmidline pain. Nonmidline pain associated with acute vertebral compression fractures is mentioned in several articles. In a 1991 study investigating clinical features of acute vertebral compression fractures in 30 patients, it was documented that “radiation to the flanks and anteriorly was common,” with 16 patients who had pain radiating to the flanks and front. Two patients had lower extremity pain radiation to the lower thigh [23].

A 2002 article by Gaughen et al. [24] describes 10 of 100 consecutive patients with no point tenderness over their fracture site. These patients had “tenderness distant from the fracture (n = 5), tenderness lateral to the fracture (n = 4), or no focal tenderness at all (n = 1).” However, the authors do not indicate the presence of any nonmidline pain in the remaining 90 patients. Their conclusion was that physical examination may lack the sensitivity and specificity that can be gained with MRI and bone scanning when selecting patients.

Another group performed percutaneous vertebroplasty on seven patients with severe radiating leg pain. The radiating pain was a result of coexisting intervertebral foramen stenosis with root compression. These patients had effective relief of their radicular pain with vertebroplasty alone, which the authors postulated was from indirect decompression by the vertebroplasty [25]. Hirsch et al. [26] reported performing percutaneous vertebroplasty with benefit in a patient with pseudo Kümmell's disease who suffered a retropulsed fragment and radiating pain around her chest and underneath her breast. A 2003 article noted percutaneous vertebroplasty candidates may occasionally have pain radiating to the ribs with thoracic spine fractures and that fractures at the level of the conus medullaris may cause pain radiating to the hips without evidence of cord compression [27].

One may ask why radiating pain can occur with a spine fracture when no definite impingement on exiting nerves can be seen on cross-sectional imaging. In actual fact, the apparent radiating pain may be referred pain [21]. In the article by Jinkins et al. [21], the authors explain in detail known and proposed mechanisms of vertebrogenic-induced referred pain. They show that referred pain can be simulated by central and anterior disk extrusions, which are common with spine fractures. The referred pain is elicited through the sympathetic nerve pathways and may cause symptoms perceived as radiating pain; however, nerve impingement is not seen on imaging. Radiating pain is a well-defined neurogenic pain classically caused by nerve impingement, whereas distant referred “pain is referred to the region corresponding roughly to the somatic distribution of the afferent fibers of the spinal nerve with which the afferent sympathetic fibers enter the spinal canal.” Recognition of the referred pain patterns that can be associated with vertebral fractures allows clear understanding of the pain patterns seen with these patients.

O'Connor et al. [28] comment on thoracic radiculopathy in a bandlike distribution to the anterior thorax, chest, or abdomen like that seen in our patients. The explanation given for the referred pain is that the ventral rami of the thoracic spine run anterolaterally between the ribs to innervate the chest and abdominal wall. This is unlike the cervical and lumbar ventral rami, which enter their corresponding plexus. This difference in anatomy can contribute to the bandlike distribution of thoracic radiculopathy to the anterior thorax, chest, or abdominal area.

Although not prospectively evaluated in the patients in this study, it was empirically noted that some patients described vague very distant pain that, in retrospect, would be more characteristic of referred rather than radiating pain. Knowledge that visceral and somatic pain systems overlap helps explain why, in some patients, pain was perceived in the abdomen and sternum and other sites uncharacteristic for radiating pain.

From the literature it is unclear what effect percutaneous vertebroplasty has on nonmidline pain occurring without nerve impingement. All the patients in this study had percutaneous vertebroplasty performed because of pain felt to be caused by fracture pain, and none had neurologic symptoms other than local or nonmidline pain.

The results of our study indicate that nonmidline pain is a frequent symptom in patients presenting for treatment without imaging findings for specific nerve impingement. The number of patients with nonmidline symptoms may be understated, as 13 patients who answered “No” to the question regarding nonmidline symptoms lacked a pain location diagram. Some of these patients may have had nonmidline pain to the side or elsewhere not addressed in the questionnaire.

Pain to the legs and thighs was improved in 81% of cases with vertebroplasty alone. Before the procedure, each patient underwent MRI, bone scanning, or both and a physical examination that supported the fact that pain was coming from the fracture and not from nerve impingement.

The greatest weakness of this study is common in retrospective studies: somewhat incomplete and inconsistent data. Pain outcome was missing in 14% of the cases. This probably does not affect the results, given the strength and consistency of the patterns of nonmidline pain relief. The inconsistency of nonmidline pain location reporting is a more pervasive problem. It is difficult to be sure that even the broad regions used in this study are correct representations of what the patients felt. This is a secondary issue, however, because the patterns of outcomes were consistent with those of the fracture level treated.

An additional problem arose from the wording of the baseline questionnaire regarding the pattern of nonmidline pain. A question asked if the pain was felt in the patients' “legs”; it would have been preferable to inquire about the thigh and leg (knee to ankle) individually. Frequently patients would answer “Yes” to this question without providing specific information about which part of their leg hurt. Ideally we would have liked to have examined thighs separately from legs, but it was not possible to discern that information in retrospect from a questionnaire that was established prospectively.

Percutaneous vertebroplasty reduced or eliminated nonmidline pain in 76-92% of the cases in which it was reported. Procedures involving lumbar vertebrae tended to be successful less often than others, but even in these cases the procedure was successful in approximately 75% of the cases.

Address correspondence to L. A. Gilula ().

Previous sectionNext section
1. Centenera LV, Choi S, Hirsch JA. Percutaneous vertebroplasty treats compression fractures. Diagn Imaging (San Franc) 2000; 22:147-148, 153 [Google Scholar]
2. Cortet B, Cotten A, Boutry N, et al. Percutaneous vertebroplasty in the treatment of osteoporotic vertebral compression fractures: an open prospective study. J Rheumatol 1999; 26:2222-2228 [Google Scholar]
3. Cyteval C, Sarrabere MP, Roux JO, et al. Acute osteoporotic vertebral collapse: open study on percutaneous injection of acrylic surgical cement in 20 patients. AJR 1999; 173:1685-1690 [Abstract] [Google Scholar]
4. Debussche-Depriester C, Deramond H, Fardellone P, et al. Percutaneous vertebroplasty with acrylic cement in the treatment of osteoporotic vertebral crush fracture syndrome. Neuroradiology 1991; 33S:149-152 [Google Scholar]
5. Deramond H, Depriester C, Galibert P, Le Gars D. Percutaneous vertebroplasty with polymethylmethacrylate: technique, indications, and results. Radiol Clin North Am 1998; 36:533-546 [Google Scholar]
6. Jensen ME, Dion JE. Percutaneous vertebroplasty in the treatment of osteoporotic compression fractures. Neuroimaging Clin N Am 2000; 10:547-568 [Google Scholar]
7. Zoarski GH, Snow P, Olan WJ, et al. Percutaneous vertebroplasty for osteoporotic compression fractures: quantitative prospective evaluation of long-term outcomes. J Vasc Interv Radiol 2002; 13:139-148 [Google Scholar]
8. Cotten A, Dewatre F, Cortet B, et al. Percutaneous vertebroplasty for osteolytic metastases and myeloma: effects of the percentage of lesion filling and the leakage of methyl methacrylate at clinical follow-up. Radiology 1996; 200:525-530 [Google Scholar]
9. Murphy KJ, Deramond H. Percutaneous vertebroplasty in benign and malignant disease. Neuroimaging Clin N Am 2000; 10:535-545 [Google Scholar]
10. Weill A, Chiras J, Simon JM, Rose M, Sola-Martinez T, Enkaoua E. Spinal metastases: indications for and results of percutaneous injection of acrylic surgical cement. Radiology 1996; 199:241-247 [Google Scholar]
11. Dousset V, Mousselard H, de Monck d'User L, et al. Asymptomatic cervical haemangioma treated by percutaneous vertebroplasty. Neuroradiology 1996; 38:392-394 [Google Scholar]
12. Feydy A, Cognard C, Miaux Y, et al. Acrylic vertebroplasty in symptomatic cervical vertebral haemangiomas: report of 2 cases. Neuroradiology 1996; 38:389-391 [Google Scholar]
13. Watts NB, Harris ST, Genant HK. Treatment of painful osteoporotic vertebral fractures with percutaneous vertebroplasty or kyphoplasty. Osteoporos Int 2001; 12:429-437 [Google Scholar]
14. Maynard AS, Jensen ME, Schweickert PA, Marx WF, Short JG, Kallmes DF. Value of bone scan imaging in predicting pain relief from percutaneous vertebroplasty in osteoporotic vertebral fractures. Am J Neuroradiol 2000; 21:1807-1812 [Google Scholar]
15. Ng PP, Caragine LP Jr, Dowd CF. Percutaneous vertebroplasty: an emerging therapy for vertebral compression fractures. Semin Neurol 2002; 22:149-156 [Google Scholar]
16. Grados F, Depriester C, Cayrolle G, Hardy N, Deramond H, Fardellone P. Long-term observations of vertebral osteoporotic fractures treated by percutaneous vertebroplasty. Rheumatology (Oxford) 2000; 39:1410-1414 [Google Scholar]
17. Do HM. Magnetic resonance imaging in the evaluation of patients for percutaneous vertebroplasty. Top Magn Reson Imaging 2000; 11:235-244 [Google Scholar]
18. Mathis JM, Barr JD, Belkoff SM, Barr MS, Jensen ME, Deramond H. Percutaneous vertebroplasty: a developing standard of care for vertebral compression fractures. Am J Neuroradiol 2001; 22:373-381 [Google Scholar]
19. Eck JC, Hodges SD, Humphreys SC. Vertebroplasty: a new treatment strategy for osteoporotic compression fractures. Am J Orthop 2002; 31:123-127; discussion 128 [Google Scholar]
20. Stallmeyer MJ, Zoarski GH. Patient evaluation and selection. In: Mathis J, Deramond H, Belkoff SM, eds. Percutaneous vertebroplasty. New York, NY: Springer-Verlag, 2002:41-60 [Google Scholar]
21. Jinkins JR, Whittemore AR, Bradley WG. The anatomic basis of vertebrogenic pain and the autonomic syndrome associated with lumbar disk extrusion. AJR 1989; 152:1277-1289 [Abstract] [Google Scholar]
22. Appel NB, Gilula LA. Percutaneous vertebroplasty in patients with spinal canal compromise. AJR 2004; 182:947-951 [Abstract] [Google Scholar]
23. Patel U, Skingle S, Campbell GA, Crisp AJ, Boyle IT. Clinical profile of acute vertebral compression fractures in osteoporosis. Br J Rheumatol 1991; 30:418-421 [Google Scholar]
24. Gaughen JR Jr, Jensen ME, Schweickert PA, Kaufmann TJ, Marx WF, Kallmes DF. Lack of preoperative spinous process tenderness does not affect clinical success of percutaneous vertebroplasty. J Vasc Interv Radiol 2002; 13:1135-1138 [Google Scholar]
25. Chung SK, Lee SH, Kim DY, Lee HY. Treatment of lower lumbar radiculopathy caused by osteoporotic compression fracture: the role of vertebroplasty. J Spinal Disord Tech 2002; 15:461-468 [Google Scholar]
26. Hirsch J, Reddy A, Linfante I, Rachlin J. Pseudo-Kummel's disease: a unique application for vertebroplasty. Pain Physician 2003; 6:207-211 [Google Scholar]
27. Peh WC, Gilula LA. Percutaneous vertebroplasty: indications, contraindications, and technique. Br J Radiol 2003; 76:69-75 [Google Scholar]
28. O'Connor RC, Andary MT, Russo RB, DeLano M. Thoracic radiculopathy. Phys Med Rehabil Clin N Am 2002; 13:623-644, viii [Google Scholar]

Recommended Articles

Response of Nonmidline Pain to Percutaneous Vertebroplasty

Full Access, ,
American Journal of Roentgenology. 2006;187:859-868. 10.2214/AJR.05.0145
Abstract | Full Text | PDF (5188 KB) | PDF Plus (5272 KB) 
Full Access, , , ,
American Journal of Roentgenology. 2006;187:W341-W349. 10.2214/AJR.04.1932
Abstract | Full Text | PDF (4740 KB) | PDF Plus (4820 KB) 
Full Access, , , , , ,
American Journal of Roentgenology. 2010;194:1323-1329. 10.2214/AJR.09.3287
Abstract | Full Text | PDF (720 KB) | PDF Plus (791 KB) 
Full Access, , , , , ,
American Journal of Roentgenology. 2006;187:873-880. 10.2214/AJR.05.1088
Abstract | Full Text | PDF (2712 KB) | PDF Plus (2799 KB) 
Full Access, ,
American Journal of Roentgenology. 1989;152:1277-1289. 10.2214/ajr.152.6.1277
Abstract | PDF (2265 KB) | PDF Plus (872 KB) 
Full Access,
American Journal of Roentgenology. 2006;187:881-886. 10.2214/AJR.05.1047
Abstract | Full Text | PDF (1798 KB) | PDF Plus (1883 KB)