Intracranial tuberculomas are rare in pregnant patients, yet they remain an important differential diagnosis for intracranial space-occupying lesions, particularly in areas with a higher prevalence of tuberculosis 25,26. The absence of active TB can lead to misinterpretation of tuberculoma symptoms as indicative of tumors. While some authors have established a staging system for tuberculous meningitis (TBM), no such correlation currently exists for tuberculoma, complicating diagnosis and treatment further 27. This challenge is amplified in pregnant patients, where both diagnosis and treatment options are limited by the physiological changes during pregnancy and the potential impact on the fetus. Literature on central nervous system tuberculosis in pregnancy is scarce, making evidence-based management of such cases particularly difficult. This systematic review provides critical insights into the clinical characteristics, diagnostic methods, treatments, maternal outcomes and newborn complications of cerebral tuberculoma in pregnant women.
Our systematic review found 26 cases of tuberculoma in pregnant women worldwide, mainly presenting with headaches, fever, seizures, and signs of increased intracranial pressure. Brain scans often revealed round lesions, mainly located in the supratentorial regions but also notably in the infratentorial areas. These lesions typically appeared isointense on T1-weighted images and hypointense on T2/FLAIR sequences, with contrast enhancement. Indirect tests, especially positive interferon-gamma release assays, indicated a strong immune response, while direct mycobacterial cultures detection methods and PCR were less frequently positive. Cerebrospinal fluid analysis in about half of the cases revealed proteinorrachia and a mononuclear predominance. Seventeen patients underwent surgery, with microbiological proof in 4 and histopathological evidence of granulomatous inflammation, caseous necrosis, and Langhans giant cells in 6. Treatment included prolonged standard anti-TB medication courses. Follow-ups mostly showed lesion improvement, with one death. Notable complications for newborns were neonatal tuberculosis, high preterm delivery rates, and 2 abortions.
Patients diagnosed with this condition were all below 30 years of age, immunocompetent women, typically in their third trimester of pregnancy. Although most lacked a personal history of tuberculosis (TB) or constitutional symptoms, respiratory symptoms appeared in about a quarter and nearly half demonstrated pulmonary compromise on imaging. This is consistent with the natural history of tuberculosis, where Mycobacteria primarily infect the lungs and then spread via the bloodstream or lymphatic system, with extension to the cerebrospinal fluid (CSF) occurring through bacterial migration along cortical veins and small penetrating arteries in the brain 28,29.
The emergence of central nervous system tuberculosis, particularly tuberculomas, is likely a result of the reactivation of latent tuberculosis in a post-primary form 30. This reactivation could be attributed to immune changes during pregnancy, such as the progressive depression of lymphocyte function in response to a specific antigen 31, predisposing patients to active disease 32. In this cohort of pregnant patients, approximately 30% had a history of TB contact, yet indirect tests indicating exposure were positive in about 80% of the patients. This contact history is similar to rates reported in other immunocompetent, non-pregnant cohorts 33, suggesting that the condition is not due to increased exposure, but rather an altered immune response during pregnancy, leading to the manifestation of the disease.
Seizures were a frequent symptom, presenting in nearly half of the cases we reviewed. This finding is particularly significant considering that they are reported in less than 1% of all pregnancies 34. The occurrence of seizures during pregnancy is a cause for significant concern, as it relates to a broad spectrum of potential diagnoses, which include vascular disorders, neurological conditions, and neoplasms, among others. For instance, eclampsia and tuberculomas share symptoms such as headaches (66.67%), visual changes (29.17%), and altered mental status (27.78%) 35. However, tuberculomas are more frequently present with fever (66.67%) and other neurological abnormalities, including cranial nerve palsies (61.11%), paresis (55.56%), meningismus (33.33%) and papilledema (33.33%). These latter symptoms have also been similarly reported in cohorts of non-pregnant patients 36. When signs of a lateralizing lesion, bleeding, infarction, or infection are present, complementary diagnostic tests need to be performed 37.
MRI findings in these patients predominantly revealed round and multiple lesions, distributed mainly across supratentorial regions. Furthermore, there was a consistent isointense appearance on T1 and hypointense on T2/FLAIR, with all lesions showing contrast enhancement. Lesion localization is consistent with reports from other cohorts of healthy adults, which have also primarily identified lesions in supratentorial locations 38,39. However, infratentorial locations were observed in nearly 30% of the patients, mirroring the behavior seen in the pediatric population 40,41. Such findings prompt further exploration into how pregnancy might alter the typical presentation of tuberculoma in patients older than 20 years.
The correlation between the histopathological evolution of tuberculomas and their imaging characteristics on MRI can provide valuable insights, potentially guiding the interpretation of common radiological features to approximate the lesion's evolutionary state 42. Initially, non-caseating tuberculomas present with a distinct pattern on MRI, appearing hypointense on T1-weighted imaging and hyperintense on T2 and FLAIR sequences, accompanied by homogeneous enhancement with contrast. As the disease progresses to caseation, the imaging evolves to display an iso- to hypointense signal across T1, T2, and FLAIR modalities, and in some patients, it is characterized by an isointense to hyperintense rim, peripheral edema, and ring-like enhancement. In advanced stages, where the caseating granuloma has central liquefaction, the imaging signature becomes notably iso- to hypointense on T1 and hyperintense on T2, consistently featuring ring-like enhancement 9,43–45
In our review, a correlation between histopathological findings and MRI features could only be established for around 12% of the pregnant patients. Each of these cases was characterized by caseous granulomas, conforming to imaging characteristics previously documented in the literature. Across T1, T2, and FLAIR sequences, imaging for these patients consistently appeared isointense and/or hypointense. Notably, in the case reported by Malhotra, instances where T1 and T2 sequences displayed a hyperintense component were observed, which is in line with what other authors have reported 9,44. Significantly, every patient showed evidence of perilesional edema, consistent with the literature that indicates such edema manifests as hyperintense on T2-weighted and FLAIR sequences, and may at times be more prominent than the lesion itself 45. Interestingly, in one of the cases we treated in our institution, the MRI demonstrated a pronounced hyperintensity on T2 and hypointensity on FLAIR, a radiological pattern known as T2/FLAIR mismatch sign. This unusual presentation, which is not typically associated with tuberculoma, posed diagnostic challenges.
While the imaging findings described in the literature, such as isointensity and/or hypointensity in T1, T2, and FLAIR sequences and the presence of perilesional edema, align with our observations in pregnant patients, there exist more advanced imaging modalities that enhance our diagnostic capabilities. One notable instance in our case review involved the use of Magnetic Resonance Spectroscopy (MRS) for a pregnant patient, where the analysis revealed an increased choline-creatinine ratio, decreased N-acetyl aspartate, and a lipid peak, as per the findings reported by Vempati et al 23. This instance underscores the utility of MRS and Gd-DTPA enhanced MR images in providing a more nuanced differential diagnosis. MRS offers a non-invasive glimpse into the body's biochemical processes, with a characteristic signature indicating high sensitivity and specificity for differentiating tuberculomas from other lesions46. Similarly, Gd-DTPA enhanced MR images (T1WI) have demonstrated that tuberculomas typically exhibit abnormal meningeal enhancement, signaling active inflammation, and often present as conglomerated ring-enhancing nodules 47,48. These advanced imaging techniques not only provide critical insights into the pathology of tuberculomas but also play a pivotal role in the diagnosis and management of this condition, particularly in the context of pregnancy.
Cerebrospinal fluid (CSF) analysis from 13 patients showed that a significant portion of the patients presented with proteinorrachia and pleocytosis, accompanied by a predominance of mononuclear cells. Such findings align with existing literature on central nervous system tuberculosis, where similar inflammatory profiles have been documented as characteristic responses to mycobacterial infection49. Contrastingly, glucose and lactate levels in our study did not demonstrate any significant trend, and the adenosine deaminase (ADA) test was conducted in only two patients, yielding one positive result. Reports from other studies indicate that patients with CNS tuberculosis may present with low glucose levels and normal protein levels in the CSF, alongside low ADA activity and negative results from cartridge-based nucleic acid amplification 17. Despite its poor specificity, when interpreted alongside other laboratory findings, ADA levels ranging from 1 to 4 U/L can aid in the diagnosis
Indirect diagnostic tests, including antigenic testing, the tuberculin test, and interferon-gamma release assays, were commonly used in pregnant patients and yielded a higher proportion of positive results compared to direct methods. Despite the high percentage of false negatives associated with these diagnostic tests, the tuberculin skin test is often the preferred method for diagnosing tuberculosis due to its low cost, higher positivity rates, and reproducibility 50,51.
The interferon-gamma release assay (IGRA), also known as the T-spot test, has been employed with notable success in diagnosing tuberculoma in pregnant patients. Despite the immunosuppression commonly observed during pregnancy, this condition does not seem to adversely affect the test's outcome, distinguishing it from results observed in other immunocompromised groups, such as individuals with HIV or drug addiction issues 52. Our review found that all patients tested positive for the CSF IGRA assay, underscoring its superior diagnostic efficacy. In comparison, a previous meta-analysis on CSF IGRAs in non-pregnant patients reported a pooled sensitivity of 79% (95% CI: 0.71–0.85) and a specificity of 95% (95% CI: 0.88–0.98), demonstrating its high diagnostic performance 53. However, it is critical to note that blood samples must be processed within 8–30 hours post-collection for viability, and a positive result does not necessarily indicate active disease. Despite its advantages, IGRA's application is limited by its cost and the paucity of data on its use in recently exposed or immunocompromised individuals 54.
Isolating M. tuberculosis through direct visualization, culture, and polymerase chain reaction (PCR) poses significant challenges, a fact corroborated by our study which found that a substantial portion of cases did not exhibit the presence of the microorganism. Specifically, in our systematic review of pregnant patients with tuberculoma, only three patients had positive culture results, and three had positive MTB-stains. Additionally, a biopsy in one patient yielded a positive PCR result for tuberculosis. This difficulty in diagnosing tuberculosis reflects broader diagnostic challenges noted in the literature 55,56.
The most common treatment regimen in this study predominantly included first-line medications such as isoniazid, rifampicin, pyrazinamide, and ethambutol, with the majority of patients receiving treatment for over 12 months. Intravenous (IV) steroids were widely used. Most patients who experienced favorable clinical outcomes were treated with anti-TB therapy for more than 9 months. Insights into the optimal duration of anti-tuberculous therapy could be derived from a randomized controlled trial that compared the length of TB therapy based on brain imaging 57,58.
Recently, a meta-analysis that included 408 studies showed that ant tuberculosis drugs did not increase the incidence of serious maternal adverse events, drug hepatotoxicity, peripheral nerve disease, or maternal mortality, and could significantly reduce adverse pregnancy outcomes 59. First-line anti-TB drugs are recommended for management 60.
The treatment of tuberculoma during pregnancy resulted in significant radiological improvement in the majority of cases. However, about 31% of the patients experienced neurological sequelae, such as nerve palsy, seizures, and gait difficulties, accompanied by a mortality rate of 6%. This contrasts with a UK cohort study among non-pregnant patients, which included 14 patients with TB mass lesions (TBML) without meningitis. In this cohort, there were no mortality cases, and 29% of patients with TBML experienced persistent neurological disabilities at 12 months. Comparatively, our cases exhibited a higher mortality rate of 6.67% and a slightly higher incidence of neurological sequelae at 31% 61.
On the other hand, the outcomes for the neonates indicate significant concerns, with 12 out of 20 experiencing neonatal complications. The incidence of small for gestational age infants, spontaneous abortions, neonatal deaths, and one confirmed case of tuberculosis with a fatal outcome, reflect the considerable risks these neonates face. These findings necessitate a thorough discussion on preventative measures, prenatal and postnatal management of pregnant women with tuberculoma, and strategies to mitigate risks to neonates. We report a high frequency of neonatal complications, typical of this entity. Neonatal tuberculosis and preterm delivery occurred more frequently; however, we did not observe cases of hydrocephalus as a complication, which have been reported in other studies 62–64. The data underscores a high disease burden for neonates, highlighting the critical need for targeted interventions and further research in this vulnerable population.
Limitations
This study, while providing valuable insights into the clinical characteristics, diagnostic methods, treatments, maternal outcomes and newborn complications of cerebral tuberculoma in pregnant women, is subject to several limitations. First, the reliance on case reports inherently carries the risk of publication bias, as these reports may preferentially publish more unusual or successfully managed cases. Additionally, the descriptive nature of our analysis, necessitated by the varied and often incomplete data available from these reports, limits our ability to perform quantitative assessments or draw firm conclusions about treatment efficacy and resistance patterns. The heterogeneity in reporting standards across the case reports led to missing data on key variables, which could have provided a more comprehensive understanding of cerebral tuberculoma management in this specific population. Furthermore, the lack of control groups in the studies included prevents direct comparison of treatment outcomes, making it challenging to ascertain the relative effectiveness of different therapeutic approaches. Finally, our findings are constrained by the scope of the literature search and the quality of the reported cases, highlighting the need for more standardized reporting and rigorous research in this area.