Differentiation of Intraspinal Tuberculosis and Metastatic Cancer Using Magnetic Resonance Imaging
Authors Li Q, Song J, Li X, Luo T, Peng J, Lv F, Li Y, Huang Z, Wang A
Received 24 July 2019
Accepted for publication 8 January 2020
Published 4 February 2020 Volume 2020:13 Pages 341—349
Checked for plagiarism Yes
Review by Single anonymous peer review
Peer reviewer comments 2
Editor who approved publication: Dr Joachim Wink
QuanJiang Li,1,* Juan Song,2,* XinYou Li,1 TianYou Luo,1 Juan Peng,1 FaJin Lv,1 YongMei Li,1 ZhongXin Huang,1 AnRan Wang1
1Department of Radiology, The First Affiliated Hospital of Chongqing Medical University, Chongqing 400016, People’s Republic of China; 2Department of Gastroenterology & Endocrinology, Wuhan No. 9 Hospital, Wuhan, People’s Republic of China
*These authors contributed equally to this work
Correspondence: Juan Peng
Department of Radiology, The First Affiliated Hospital of Chongqing Medical University, 1 Youyi Road, Yuanjiagang, Yuzhong District, Chongqing 400016, People’s Republic of China
Tel +86 1 898 328 0171
Email [email protected]
Purpose: This study aimed to explore the differences in the magnetic resonance imaging (MRI) findings between intraspinal tuberculosis and metastatic cancer, which may aid in making the correct diagnosis.
Patients and Methods: The clinical features and MRI findings of 15 patients with intraspinal tuberculosis and 11 patients with intraspinal metastatic cancers were retrospectively analyzed.
Results: The mean ages of the patients with intraspinal tuberculosis and metastatic cancer were 26.3 (15– 42) and 52.1 (38– 67) years, respectively. All intraspinal tuberculosis cases were secondary to primary extraspinal tuberculosis, including tuberculous meningitis (11/15), as well as pulmonary (9/15), vertebral (5/15), urinary tract (1/15), abdominal (1/15), cervical lymph node (1/15), and multisystem tuberculosis (9/15). The intraspinal metastases originated from the breast (5/11), lung (3/11), kidney (1/11), ovarian (1/11), and nasopharyngeal cancers (1/11). Both intraspinal tuberculosis and metastatic cancers presented with multiple intra- and extramedullary lesions throughout all regional segments of the spinal canal, accompanied by irregularly thickened meninges. Intraspinal tuberculous lesions had indistinct edges that integrated with each other, most of them exhibiting obvious enhancement on MRI. Conversely, intraspinal metastatic lesions were distinctly separated with clear edges and exhibited lesser enhanced MRI than intraspinal tuberculosis.
Conclusion: A combined analysis of clinical features and MRI findings may be helpful in differentiating intraspinal tuberculosis from metastatic cancer.
Keywords: intraspinal tuberculosis, intraspinal metastatic cancer, intramedullary nodule, leptomeningitis
Tuberculosis can involve all human organs. Although the lungs are most commonly involved, extrapulmonary tuberculosis accounted for 15% of all tuberculosis cases according to the WHO in 2016.1,2 As a common type of extrapulmonary tuberculosis, intraspinal tuberculosis occurs increasingly, although its accurate incidence is unknown, which is in the same situation as intraspinal metastasis.3 Both intraspinal tuberculosis and metastatic cancer may be rare, but they are difficult to differentiate because of their similar presentations and the high negative incidence of microbiological test for tuberculosis in cerebrospinal fluid (CSF) and their prognoses are completely different.1,4 Most patients with intraspinal tuberculosis can achieve remarkable recovery after early diagnosis and treatment, whereas intraspinal metastatic cancers have a poor treatment response and poor prognosis.5–7 To the best of our knowledge, although the optimal treatment of both diseases remains controversial, early accurate diagnosis can be helpful to improve neurological deficits and quality of life.8–13
Magnetic resonance imaging (MRI) is the modality of choice for the diagnosis of intraspinal lesions.3,13 According to previous reports, intraspinal tuberculosis and metastatic cancer present differently in several aspects, including average patient age, distribution of the lesions, and clinical histories, but they also share the following features on MRI: significant spinal dural enhancement, intramedullary nodules, edema, space-occupying effect, meningitis, and hydrocephalus.3 In this study, we aimed to explore the differences in the MRI findings between intraspinal tuberculosis and metastatic cancer, thereby facilitating their accurate diagnosis. We hypothesized that combined analysis of clinical and MRI features may be helpful to distinguish intraspinal tuberculosis from metastatic lesions of cancers.
Materials and Methods
A total of 26 inpatients, admitted in our hospital from November 2012 to May 2018, were retrospectively enrolled in this study, including 15 patients (8 men and 7 women) with intraspinal tuberculosis and 11 (5 men and 6 women) with intraspinal metastases. All these patients received multiple gadolinium contrast MRI examinations of the spine, which showed obvious subdural and intramedullary lesions; patients with epidural lesions were excluded. Their medical records, including clinical features and imaging findings, were reviewed and analyzed. This study was approved by the Ethical Committee of the First Affiliated Hospital of Chongqing Medical University (IRB ID: 2019-357). Waiver of consent was obtained given the observational nature of the project (retrospective archive study). We declare that the patient data confidentiality has been protected in the course of studying and the study has been conducted in accordance with the Helsinki Declaration.
All the MRI examinations were performed on a 1.5 T GE Signa Excite MR image scanner (General Electric, Milwaukee, WI, USA). The MR imaging protocols included sagittal, coronal fast spin-echo (FSE) T2-weighted (repetition time [TR] = 3200 ms, echo time [TE] = 51 ms), and T1-weighted imaging (TR = 340 ms, TE = 10 ms) of the whole spine, with axial T2- and T1-weighted sequences for the interesting regions. The acquisition matrix was 256×256, and the slice thickness was 3.5 mm with a 1.0-mm gap. Contrast-enhanced T1-weighted images were obtained from each patient after injecting the gadopentetate dimeglumine 0.1 mmol/Kg through the median cubital vein with the same scanning parameters.
All the 26 patients experienced progressive neurologic dysfunction of the lower extremities; weakness was the most common initial symptom, followed by pain, hypoesthesia with a sensory level, and eventually astasia. Some patients also experienced sphincter disturbance.
The mean age of the 15 patients with intraspinal tuberculosis was 26.3 (15–42) years. All these patients were negative for acid-fast smear microscopy and had no tuberculosis-specific findings in CSF analysis (including quantitation of lymphocytes, protein, and glucose). They were all secondary diseases, with primary lesions in extraspinal organs: the meninges (n = 11), lungs (n = 9), vertebrates (n = 5), urinary tract (n = 1), abdomen (n = 1), and cervical lymph nodes (n = 1); nine of them had multisystem lesions. Diagnosis of intraspinal tuberculosis was made based on the treatment response and radiographic changes following anti-tuberculosis therapy, as there were no specific clinical symptoms for intraspinal tuberculosis. According to initial manifestations, these intraspinal tuberculosis cases can be classified into two groups: (1) paraparesis (7 patients) and (2) tuberculosis symptoms (8 patients). The seven paraparesis patients were admitted for acute onset of lower back pain and lower extremity weakness (2 patients displaying coincident headache, dysphoria, and disturbance of consciousness); however, their diseases progressed rapidly and showed tuberculosis-related symptoms soon after admission. The eight cases with tuberculosis symptoms as the initial presentations displayed initial tuberculosis symptoms in other organs than the spine; the interval between symptom onset and development of intraspinal disease was <1 year. All the 15 patients with intraspinal tuberculosis achieved partial or complete recovery after anti-tuberculosis therapy; however, three patients eventually died within 2 months because of severe complications.
The mean age of the 11 cancer patients with intraspinal metastasis was 52.1 (38–67) years. All these patients experienced lower extremity weakness and pain within 2 to 19 years after being diagnosed with the primary cancers, which involved the breast (n = 5), lungs (n = 3), ovaries (n = 1), kidneys (n = 1), and nasopharynx (n = 1). All patients had distant metastases in other organs prior to intraspinal metastasis, with brain metastasis being most frequently observed (n = 5), followed by chest wall metastasis (n = 1) and multiple metastases (n = 1). The 11 cancer patients with intraspinal metastasis experienced a rapid worsening of disease after the intraspinal metastasis was diagnosed and died within 4 years after the onset of the initial neurologic deficits due to poor treatment response.
Both intraspinal tuberculosis and metastatic lesions were observed throughout the entire spinal canal and most of them involved the subdural extramedullary region (Figures 1–4). However, their predisposed sites were different. Intraspinal tuberculosis mainly affected the thoracic (5/15), lumbar and sacral (5/15) regions, with the remaining 5 cases involving more than two regions of the spinal cord. For intraspinal metastases, no obvious location predilection was seen, with each of thoracic and lumbar lesions accounting for 1 of 11 cases, respectively, and the remaining 9 cases affecting more than 2 regions (one case involving the entire spinal cord).
Multiple intraspinal nodular lesions together with irregularly thickened meninges on MR imaging were observed in both intraspinal tuberculosis and metastatic cancer patients. All intraspinal tuberculosis cases were additionally accompanied by leptomeningitis, which showed isointense on T1-weighted images and mixed iso- or slightly hyperintense on T2-weighted images, and obvious homogeneous linear, mottled, or patchy enhancement in more than two segments on enhanced T1-weighted images (Figure 1). The lesions had indistinct edges and appeared to integrate with each other, most showing obvious enhancement along with focal or diffuse spinal cord edema and syringomyelia (n = 6) (Figures 1 and 2). Other relevant MRI findings in intraspinal tuberculosis were as follows: adhesions and stenosis of the subarachnoid space (n = 3), enhancing thickened nerve roots (n = 4), and clearly defined intramedullary tuberculomas (n = 2). Tuberculomas were isointense on T1- and iso- or slightly hyperintense on T2-weighted images with homogeneous or ring enhancement (Figure 2).
The MRI findings of intraspinal metastatic cancer were similar to tuberculosis, but metastatic lesions were clearly separated from each other and exhibited lesser enhancement on MRI than intraspinal tuberculosis. Moreover, meningeal thickening caused by metastatic cancer was thinner and most frequently limited to four consecutive vertebral segments (Figures 3 and 4). Intramedullary nodules surrounded by focal edema (isointense on T1- and slightly hyperintense on T2-weighted images) together with a poorly defined border were observed in eight patients with intraspinal metastasis (Figure 3). Two metastatic patients presented multiple patchy areas of enhancing thickened meninges accompanied by the enhancing nodules.
It could be difficult to differentiate intraspinal tuberculosis from metastatic lesions of cancer or other space-occupying lesions by relying solely on clinical symptoms and CSF analysis, as demonstrated in this study. However, these pathologies should be highly suspected when a patient with a current or previous history of tuberculosis or cancer presents with rapidly progressive neurologic deficits. Joint consideration of a patient’s clinical features and MRI findings may be helpful to arrive at a correct early diagnosis and allow expedient appropriate treatment.
The worldwide incidence of intraspinal tuberculosis has been recently increasing.3,6,14,15 Although a male predominance was reported, intraspinal tuberculosis predominantly occurs in younger patients of both genders, with a mortality of 20% in children aged <5 years and 60% in patients aged >50 years.3,7,16–18
Most intraspinal tuberculosis cases arise from a primary extraspinal source, with the lungs and intracranial region being the most common primary sites, as demonstrated in this study; these two primary tuberculous diseases disseminate to the intraspinal region through the blood and CSF, respectively.19 But the primary intraspinal tuberculosis lesion was reported to be the most common original source in one study.17 The thoracic segments were most commonly involved in our series. This involvement pattern can be attributed to the distribution of blood supply to the spinal cord: the thoracic segment receives nearly 45% of the spinal cord blood supply.8,9
Tuberculous leptomeningitis with intraspinal nodules was thought to be the most common feature of patients with intraspinal tuberculosis.15–17 In the present study, enhanced T1-weighted images delineated the thickened meninges better and was more sensitive to detect tiny lesions compared with unenhanced images, while Gupta et al found no enhancement of leptomeningitis in a few patients, inferring that the contrast MRI imaging might have no benefit for differentiating tuberculous leptomeningitis from metastatic lesions of cancer, which was totally contrary to our findings.17 The reason for this discrepancy was not clear and may be due to the different scanning protocols, contrast mediums, and different selection criteria for the enrolled patients. Other findings in patients with tuberculous leptomeningitis included hydrocephalus, intramedullary tuberculoma, spinal arachnoiditis, en plaque appearance of the sacral nerve roots, focal or diffuse edema in the spinal cord, and syringomyelia, which have also been reported by others.3,16,20–23
In this study, intramedullary tuberculous lesions were less common relative to extramedullary ones. As a specific form of intramedullary tuberculous lesions, which may be mistaken for tumor, tuberculoma appeared a regular shape of isointense on T1- and slightly hyperintense on T2-weighted images, with homogeneous nodular enhancement seen in the early stage, while ring enhancement observed in the late stage, presumably due to caseous necrosis in the lesion center. However, tuberculomas were more frequently detected in other studies that had similar imaging presentations to ours.8,17,24 In the present study, we also found a 13% incidence of tuberculoma combined with leptomeningitis, which was similar to the 9.8% incidence, reported in a study of 122 patients.25 All thickened meninges accompanied with nodules had undefined edges and seemed to be fused together, most of them showing enhancement on MRI; when involved, the cauda equina exhibited an en plaque appearance, and the nerve roots exhibited a thickening appearance.16,26
Intraspinal Metastasis of Cancer
The exact incidence of intraspinal metastasis of cancer is unclear, but the incidence of intramedullary spinal cord metastasis has been reported to be 0.9–2.1%.11 Although the lungs are the most common sites of primary cancer, the breast ranked first in the frequency of occurrence locations for primary cancer in this study, which may be due to the small sample size.5,11,12 As for pathologic mechanisms for intraspinal metastasis, three probable mechanisms have been suggested: (1) hematogenous spread via arteries and/or the vertebral venous plexus (Batson’s venous plexus), (2) infiltration along with involved nerve roots, and (3) direct infiltration through the CSF.11 Hematogenous spread is generally believed to be responsible for most cases. With regard to the involved regional areas of intraspinal metastasis, one study found that the cervical cord or the conus medullaris was most commonly affected (62%).12 However, no obvious regional predominance was found in our study, which is consistent with the conception of most investigators that there is no regional segmental predilection for intraspinal metastasis. Furthermore, the involvement of the entire spinal cord was observed in most patients in our study.
Concomitant systemic metastases are often present in patients with intraspinal metastasis at the time of diagnosis, but some may be diagnosed without a definite primary cancer.12 In this study, a known primary cancer had been determined for all patients before the diagnosis of intraspinal metastasis was made; meanwhile, brain metastasis was demonstrated to be the most common concomitant metastasis.
The MRI findings of intraspinal metastasis were similar to those of intraspinal tuberculosis, but some differences were remarkable. For patients with metastatic disease, a greater number of intramedullary nodules (solitary and/or multiple) surrounded by less edema were observed which appeared isointense on T1- and slightly hyperintense on T2-weighted images. When meninges were involved, linear enhancement can be observed, and when the cauda equina involved, thickening of the nerve roots observed. Extramedullary lesions appeared as thickened meninges, similar to those of intraspinal tuberculosis, but had sharp margins, limited scope, and evident distinction between lesions.
The MR spectroscopy was reported to apply in the differentiation of intraspinal nodules by a few studies. Lipid peak was thought to be radiological biomarker of tuberculoma while evaluated choline peak was thought to be specific for the metastasis, but the result was affected by many factors such as strong susceptibility artifacts and small size of intramedullary nodule for single-voxel placement, and it is not straightforward for the application of MR spectroscopy in intraspinal lesions, so the differentiation of intraspinal lesions may mainly rely on the MRI findings.27,28
There were certain limitations to this study. Given its small sample size, it was difficult to extrapolate the results to a large population. In addition, the definite pathological diagnosis of the intraspinal lesions was not obtained; diagnosis of intraspinal tuberculosis was made based on the response to anti-tuberculous therapy and intraspinal metastasis was diagnosed based on patient history of cancer and MRI findings. The validity of the study was also somewhat limited due to the lack of quantitative measurement and analysis regarding the imaging findings. We plan to enlarge the sample size and include quantitative measurements and analysis (such as MR spectroscopy, diffusion-weighted imaging) in a future study.
It is imperative to differentiate intraspinal tuberculosis from intraspinal metastasis of cancer due to their completely different prognosis, and combined consideration of a patient’s clinical features and MRI findings can be helpful for their antidiastole. Intraspinal tuberculosis lesions may exist more obvious enhancement on MRI compared with intraspinal metastatic lesions and the meninges will become thicker when involved. Intraspinal metastatic lesions may have sharp margins, with evident distinction between each other, while tuberculosis lesions may appear to be fused and integrated with each other.
This study was supported by the Science and Technology Bureau of YuZhong District, Chongqing, China (no 20190111) and the health bureau of Chongqing, China (no 2016MSXM005).
The author reports no conflicts of interest in this work.
1. Kumari P, Lavania S, Tyagi S, et al. A novel aptamer-based test for the rapid and accurate diagnosis of pleural tuberculosis. Anal Biochem. 2019;564–565:80–87. doi:10.1016/j.ab.2018.10.019
2. Lavania S, Das R, Dhiman A, et al. Aptamer-based TB antigen tests for the rapid diagnosis of pulmonary tuberculosis: potential utility in screening for tuberculosis. ACS Infect Dis. 2018;4(12):1718–1726. doi:10.1021/acsinfecdis.8b00201
3. Sonmez G, Ozturk E, Sildiroglu HO, et al. MRI findings of intracranial tuberculomas. Clin Imaging. 2008;32(2):88–92. doi:10.1016/j.clinimag.2007.08.024
4. Dhiman A, Haldar S, Mishra SK, et al. Generation and application of DNA aptamers against HspX for accurate diagnosis of tuberculous meningitis. Tuberculosis. 2018;112:27–36. doi:10.1016/j.tube.2018.07.004
5. Watanabe M, Nomura T, Toh E, et al. Intramedullary spinal cord metastasis: a clinical and imaging study of seven patients. J Spinal Disord Tech. 2006;19(1):43–47. doi:10.1097/01.bsd.0000188661.08342.2a
6. Roca B. Intradural extramedullary tuberculoma of the spinal cord: a review of reported cases. J Infect. 2005;50(5):425–431. doi:10.1016/j.jinf.2004.07.012
7. Sharma MC, Arora R, Deol PS, et al. Intramedullary tuberculoma of the spinal cord: a series of 10 cases. Clin Neurol Neurosurg. 2002;104(4):279–284. doi:10.1016/S0303-8467(01)00196-2
8. Lim YS, Kim SB, Kim MK, et al. Disseminated tuberculosis of central nervous system: spinal intramedullary and intracranial tuberculomas. J Korean Neurosurg Soc. 2013;54(1):61–64. doi:10.3340/jkns.2013.54.1.61
9. Macdonell AH, Baird RW, Bronze MS. Intramedullary tuberculomas of the spinal cord: case report and review. Rev Infect Dis. 1990;12(3):432–439. doi:10.1093/clinids/12.3.432
10. Muthukumar N, Venkatesh G, Senthilbabu S, et al. Surgery for intramedullary tuberculoma of the spinal cord: report of 2 cases. Surg Neurol. 2006;66(1):69–74. doi:10.1016/j.surneu.2005.10.024
11. Kalayci M, Cagavi F, Gul S, et al. Intramedullary spinal cord metastases: diagnosis and treatment-an illustrated review. Acta Neurochir. 2004;146(12):1347–1354. doi:10.1007/s00701-004-0386-1
12. Dam-Hieu P, Seizeur R, Mineo JF, et al. Retrospective study of 19 patients with intramedullary spinal cord metastasis. Clin Neurol Neurosurg. 2009;111(1):10–17. doi:10.1016/j.clineuro.2008.06.019
13. Nguyen NC, Sayed MM, Taalab K, et al. Spinal cord metastases from lung cancer: detection with F-18 FDG PET/CT. Clin Nucl Med. 2008;33(5):356–358. doi:10.1097/RLU.0b013e31816a784c
14. Golden MP, Vikram HR. Extrapulmonary tuberculosis: an overview. Am Fam Physician. 2005;72(9):1761–1768.
15. Garg RK, Malhotra HS, Gupta R. Spinal cord involvement in tuberculous meningitis. Spinal Cord. 2015;53(9):649–657. doi:10.1038/sc.2015.58
16. Gupta R, Garg RK, Jain A, et al. Spinal cord and spinal nerve root involvement (myeloradiculopathy) in tuberculous meningitis. Medicine. 2015;94(3):e404. doi:10.1097/MD.0000000000000404
17. Gupta RK, Gupta S, Kumar S, et al. MRI in intraspinal tuberculosis. Neuroradiology. 1994;36(1):39–43. doi:10.1007/BF00599194
18. Oncul O, Baylan O, Mutlu H, et al. Tuberculous meningitis with multiple intracranial tuberculomas mimicking neurocysticercosis clinical and radiological findings. Jpn J Infect Dis. 2005;58(6):387–389.
19. Tez S, Tunca A, Karaarslan Y, et al. Acute spinal cord dysfunction in bacterial meningitis: MRI findings. Clin Radiol Extra. 2004;59(2):11–14. doi:10.1016/S1477-6804(03)00031-1
20. Muthukumar N, Sureshkumar V. Concurrent syringomyelia and intradural extramedullary tuberculoma as late complications of tuberculous meningitis. J Clin Neurosci. 2007;14(12):1225–1230. doi:10.1016/j.jocn.2006.05.013
21. Celik ÜS, Alabaz D, Yildizdas D, et al. Cerebral salt wasting in tuberculous meningitis: treatment with fludrocortisone. Ann Trop Paediatr. 2005;25(4):297–302. doi:10.1179/146532805X72458
22. Husain M, Jha DK, Rastogi M, et al. Role of neuroendoscopy in the management of patients with tuberculous meningitis hydrocephalus. Neurosurg Rev. 2005;28(4):278–283. doi:10.1007/s10143-005-0397-2
23. Ira S. Paradoxical appearance of intracranial tuberculoma in a child with tuberculous meningitis. J Trop Pediatr. 2005;51(3):191–193. doi:10.1093/tropej/fmi019
24. Luthra G, Parihar A, Nath K, et al. Comparative evaluation of fungal, tubercular, and pyogenic brain abscesses with conventional and diffusion MR imaging and proton MR spectroscopy. Am J Neuroradiol. 2007;28(7):1332–1338. doi:10.3174/ajnr.A0548
25. Kilani B, Ammari L, Tiouiri H, et al. Neuroradiologic manifestations of central nervous system tuberculosis in 122 adults. Revue Méd Int. 2003;24(2):86–96. doi:10.1016/S0248-8663(02)00716-6
26. Juan P, Tian-You L, Fa-Jin L, et al. MRI in diagnosis and differential diagnosis of intraspinal tuberculoma, metastasis and neurofibromatosis. Chin J Med Imaging Technol. 2011;27(4):702–705.
27. Sathyanathan BP, Raju BP, Natarajan K, et al. 3T proton MR spectroscopy evaluation of spinal cord lesions. Indian J Radiol Imaging. 2018;28(3):285–295. doi:10.4103/ijri.IJRI_122_17
28. Hock A, Henning A, Boesiger P, et al. 1H-MR spectroscopy in the human spinal cord. Am J Neuroradiol. 2013;34(9):1682–1689. doi:10.3174/ajnr.A3342
This work is published and licensed by Dove Medical Press Limited. The full terms of this license are available at https://www.dovepress.com/terms.php and incorporate the Creative Commons Attribution - Non Commercial (unported, v3.0) License. By accessing the work you hereby accept the Terms. Non-commercial uses of the work are permitted without any further permission from Dove Medical Press Limited, provided the work is properly attributed. For permission for commercial use of this work, please see paragraphs 4.2 and 5 of our Terms.Download Article [PDF]