If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Neurology, The Third Xiangya Hospital, Central South University, Changsha, Hunan, 410013, ChinaDepartment of Immunology, School of Basic Medical Science, Central South University, Changsha, Hunan, 410013, China
Corresponding authors at: Department of Neurology, The Third Xiangya Hospital, Central South University, 138 Tongzipo Road, Yuelu District, Changsha, Hunan, 410013, China.
Corresponding authors at: Department of Neurology, The Third Xiangya Hospital, Central South University, 138 Tongzipo Road, Yuelu District, Changsha, Hunan, 410013, China.
Paradoxical tuberculomas are associated with early CSF deterioration.
•
Neutrophils play an important role in tuberculosis-associated paradoxical reactions.
•
Early worsening of CSF parameters predicts subsequent development of tuberculomas.
Abstract
Objectives
We investigated whether early worsening of cerebrospinal fluid (CSF) predicts the later paradoxical tuberculomas and is a potential predictive biomarker.
Methods
Patients of HIV-negative tuberculous meningitis fulfilling the inclusion criteria(n = 98) underwent clinical and CSF evaluation, together with repeated neuroimaging. We compared the baseline clinical data and continuous CSF of patients who did (n = 36) and did not (n = 62) develop paradoxical tuberculomas, and reported the changes associated with symptomatic tuberculomas. A logistic regression analysis was developed to reveal predictors for paradoxical tuberculomas.
Results
The proportion of worsening CSF parameters (WBC count and percent neutrophils) in the paradoxical tuberculomas group (27/36, 75.0%) was significantly higher than the non-paradoxical tuberculomas group (15/62, 24.2%). The logistic regression analysis revealed that worsening CSF parameters was the highest risk predictor for paradoxical tuberculomas. Most worsening CSF parameters (81.0%) occurred within two weeks after treatment (2–24 days, median 7 days), and paradoxical tuberculomas commonly happened two weeks later (12 days to 13 months, median 22 days). The period between worsening CSF parameters and paradoxical tuberculomas ranged from 6 to 383 days (median 21days). There were no significant differences in mortality and prognosis between the two groups.
Conclusions
Early worsening of CSF parameters predicts subsequent development or progression of tuberculomas.
). Intracranial tuberculoma is a common complication of TBM and its formation is related to the interaction of host immune response and Mycobacterium tuberculosis (
). Tuberculomas sometimes present at the time of diagnosis and more often appear or progress during the successful treatment of tuberculosis, which is known as ‘paradoxical reactions’ (
). This is a well-documented phenomenon, which has been described as the worsening of pre-existing tuberculous lesions or the appearance of new TBM lesions whose clinical symptoms initially improved with anti-tuberculosis treatment (ATT) (
). Whether or not the presence of tuberculoma affects clinical manifestation and prognosis, the emergence of tuberculoma as a new lesion during treatment can be anxiety-provoking. The mechanism behind the progression or development of intracranial tuberculomas after treatment remains a subject of speculation. It has been reported that the paradoxical tuberculomas were attributable to several factors such as exaggerated immune responses occurring on ATT, the virulence of the tubercle bacilli, the site of infection, the antigen load, hypersensitivity to tuberculosis, and chemotherapeutic regimen (
Relationship between intracranial granulomas and cerebrospinal fluid levels of gamma interferon and interleukin-10 in patients with tuberculous meningitis.
A therapy-induced transient worsening of cerebrospinal fluid (CSF) parameters is also frequently observed during treatment, such as an initial lymphocytic reaction may change in the direction of neutrophil predominance (
). This phenomenon has been postulated to be the result of an exaggerated immunological reaction related to the release of tubercular proteins during ATT (
) have reported that worsening of CSF parameters in patients generally occurred early in the therapy and resolved within 1–2 weeks, and they were accompanied by tuberculoma development after several weeks. It has been regarded as the pathogenesis of paradoxical tuberculomas of TBM patients (
With the availability of computed tomography and magnetic resonance imaging (MRI), changes in the CNS can be monitored. However, CSF changes earlier and more intuitively than neuroimaging, and may be a predictor of the occurrence of tuberculomas. We aimed to compare the dynamic changes of clinical characteristics, CSF parameters, and imaging findings in TBM patients with and without paradoxical tuberculomas and to assess the link between early worsening CSF parameters and later development of tuberculomas.
Materials and methods
Patients
This retrospective study was conducted in the Third Xiangya Hospital and Central Hospital of Changsha City in China. HIV negative adult patients with TBM admitted to the Neurology Department from Jan 2018 to Dec 2019 were included respectively. Formal ethical approval was taken from the Third Xiangya Hospital, Central South University, China. We evaluated the records of all in-patients with TBM who had completed at least 6 months of treatment and follow-up.
Inclusion and exclusion criteria
TBM cases comprised of adults (≥14 years old) admitted for definite, probable, or possible TBM according to a published case definition (
), were eligible to enter the trial. Patients with the following were excluded: (1) HIV infection, (2) age <14 years old, (3) intracranial infection with other pathogens or tumor, (4) lumbar puncture less than three times, (5) patients did not have MRI before ATT, (6) data was incomplete.
Clinical evaluation
All enrolled patients were subjected to a detailed clinical evaluation. Demographic information, meningitis symptoms, seizures, focal neurological deficit, cranial nerve palsy, medications, history of tuberculous disease, and the presence of HIV infection and other systemic illnesses were recorded. We also recorded whenever a patient had any deterioration or new findings.
Laboratory investigations
Laboratory investigations were performed at baseline, including a complete hemogram, liver, and renal function tests, fasting blood sugar, electrolytes, erythrocyte sedimentation rate, C-reactive protein level, antibody to HIV, syphilis, and hepatitis virus (HBV and HCV). All consecutive CSF analysis results were collected, including routine examination, biochemistry, cytology, microbiology (including microscopy and culture for fungi and pyogenic bacteria), and modified Ziehl–Neelsen stain. India ink staining and Cryptococcus latex agglutination testing (colloidal gold method, IMMY) were performed to rule out cryptococcal meningitis.
Worsening CSF parameters
We explored the differences between the two groups in changes in intracranial pressure and CSF parameters (including biochemical and cellular components). Both the values and the serial changes in CSF parameters from baseline were analyzed in patients with or without transient worsening. The time to the worsening of CSF parameters is defined as the interval between the initiation of ATT and the onset of the worsening of the CSF parameter. It has been suggested that presumptive worsening CSF may be attributed to the delay of treatment, so the time from onset to initiation of treatment was further compared between patients with and without worsening of CSF parameters.
Definition of neuroimaging findings
Tuberculoma was defined as discrete or coalescing cerebral masses showing nodular or ring-shaped enhancement. Paradoxical tuberculoma was defined as the expansion, increase, or appearance of new tuberculoma after ATT, and non-paradoxical tuberculoma was defined as no cerebral tuberculoma or tuberculoma present before treatment or tuberculoma that decreased or disappeared after treatment. Other neuroimaging definitions are as described previously (
). Patients received 3 months of daily oral isoniazid (5 mg/kg; maximum, 300 mg), rifampicin (10 mg/kg; maximum, 600 mg), pyrazinamide (25 mg/kg; maximum,2 g/day), and ethambutol (20 mg/kg; maximum 1.2 g/day) followed by 6-month administration of isoniazid and rifampicin. Patients also received treatment with a corticosteroid. The corticosteroid regimen consisted of intravenous dexamethasone for 2–4 weeks (0.4 mg/kg/day) and then oral treatment with prednisone for 2–4 weeks. Appropriate symptomatic treatment (mannitol, antiepileptic drugs, analgesics, etc.) was also provided.
Follow-up
After 6 months of treatment, the patients were followed up with Modified Rankin Score (MRS), ranging from 0 (no symptoms) to 5 (totally dependent on others, requiring help day and night). These were classified as a ‘good outcome’ (a score of 0), ‘intermediate outcome’ (scores of 1 or 2), or ‘severe disability’ (scores of 3, 4, or 5). Clinical evaluation and analysis of CSF changes were also performed when clinical deterioration occurred. Cerebral MRI was performed before treatment and repeated every month and at any period of clinical deterioration after treatment, lasting for more than 6 months.
Statistical analysis
All statistical analyses were performed using SPSS version 18.0 software (SPSS Inc., Chicago, IL, USA). For univariate analysis, the Chi-square test was used to analyze categorical variables; the Student’s t-test was used to compare continuous variables if normally distributed and the Mann–Whitney U test if not normally distributed. For multivariate analysis, those variables with a P-value of <0.05 on univariate analysis were included in the binary logistic regression analysis to obtain independent risk factors. A variable was considered significant if two-tailed P-values were <0.05.
Results
Demographic and baseline characteristics
The final diagnoses and reasons for the exclusion of patients with TBM are shown in Figure 1. Of the 98 patients included in the final analysis, 66 had definite TBM, 28 highly probable TBM, and 4 possible TBM. Thirty-six (36.7%) manifested paradoxical tuberculomas during treatment, 8 (8.2%) patients developed tuberculomas before treatment and decreased or disappeared after treatment, and 54 (55.1%) were free of tuberculoma. Forty-five (45.9%) were females and the mean age was 42.1 years (range 14–82). The demographic and baseline characteristics of these two groups are shown in Table 1.
Figure 1Flow diagram of patients with features of tuberculous meningitis with or without paradoxical tuberculomas screened for study inclusion.
Parameters of CSF continuous investigations were compared between these two groups (Table2). To facilitate comparison of changes before and after treatment, the CSF parameter values at all time points were divided by baseline value so that the baseline value was treated as 1. Only the worsening of CSF WBC count and percent neutrophil were significantly different between the two groups. Therefore, we continued to explore the characteristics of the worsening of the two CSF parameters. When the CSF WBC count increased after ATT, the majority of the increased WBCs were neutrophils. Since some patients have a higher proportion of neutrophils (more than 50%) at baseline, they did not exhibit the characteristic of dominant cells appearing category-shifted. The serial changes of CSF WBCs and neutrophils are shown in Figure 2 in these two groups. A total of 42 (42.9%) patients developed worsening of CSF parameters after the initiation of ATT. These patients included 15 males and 27 females with a median age of 39 years. Among them, the WBC count increased to two-fold or more of the baseline in 25 patients, the percent neutrophil increased to two-fold or more of the baseline in 19 patients, and lymphocyte dominance (>50%) shifted to neutrophil dominance (>50%) in 17 patients. The interval between ATT and the worsening of CSF parameters was 2–24 days (median 7 days). Most worsening of CSF parameters (81.0%) occurs within two weeks after ATT. In 36 patients with paradoxical tuberculomas, twenty-seven (75.0%) patients observed worsening of CSF parameters before tuberculomas were present; while in 62 paradoxical tuberculomas patients, the other 15 (24.2%) patients with worsening CSF parameters were observed. Additionally, the time interval from onset to initiation of treatment between patients developing worsening CSF parameters (21.2 ± 24.6) and without worsening CSF parameters (20.5 ± 25.8) was not statistically significant (P = 0.902).
Table 2Comparison of the CSF Changes of patients with and without paradoxical tuberculomas.
Figure 2Changes of Cerebrospinal fluid parameters over time in patients who developed paradoxical tuberculomas (left, n = 36) and non-paradoxical tuberculomas (right, n = 62), including white blood cell count (A), percent neutrophils (B). The abscissa represents time point of lumbar punctures, patients without worsening CSF parameters choose the following four time points: at baseline(before treatment), any time within two weeks after treatment, 3 weeks after treatment, 4 weeks or more after treatment. The ordinate represents the ratio of the value of the CSF continuous parameters to that of baseline. The CSF parameter values at all time points were divided by baseline value so that the baseline value was treated as 1. For values with multiple of increase ≥6, they are expressed as 6. Abbreviations: WBC: white blood cell count; CSF: cerebrospinal fluid.
The latent period (time between the start of ATT and the development of tuberculomas) ranged from 12 days to 13 months (median 22 days). Of 36 patients with paradoxical tuberculomas, clinical deterioration correlated with tuberculoma was present in 14 patients; 22 did not reveal any new symptoms. Tuberculomas were classified in terms of their location within the CNS as meningeal, parenchymal, basal cistern, and spinal cord. The clinical and imaging characteristics from two series of patients with the development or progression of intracranial tuberculomas are summarized in Table 3.
Table 3Summary of characteristics from two series of patients with paradoxical development or progression of intracranial tuberculomas.
The outcome of the patients was not related to tuberculomas (Table1). It is worth mentioning that patients with paradoxical tuberculoma have a longer hospital stay (48.7 ± 34.9 vs. 30.1 ± 19.2, P < 0.05)
Predictors of paradoxical tuberculomas
The univariate analysis revealed a statistically significant set of variables: female, symptom duration, history of tuberculosis, hepatitis B, miliary tuberculosis, hydrocephalus, meningeal enhancement (Table 1), and worsening CSF parameters including WBC and neutrophils (Table 2). On binary logistic regression analysis, only worsening CSF parameters, hydrocephalus and meningeal enhancement were significant predictors (Table 4).
Table 4Summary of logistic regression analysis (significant predictors of paradoxical tuberculoma).
Patients with any of the following CSF changes were classified as having worsening CSF parameters: an increase in WBC counts/percent neutrophils to two-fold or more of the baseline after treatment, lymphocyte predominance (>50%) before treatment shifted to neutrophil predominance (>50%) after treatment.
2.016
7.511 (2.223–25.379)
0.001
Hydrocephalus
1.881
6.561 (1.331−32.338)
0.021
Basal meningeal enhancement
1.702
5.486 (1.350−22.288)
0.017
Abbreviations: CI = confidence interval; CSF = cerebrospinal fluid.
a Patients with any of the following CSF changes were classified as having worsening CSF parameters: an increase in WBC counts/percent neutrophils to two-fold or more of the baseline after treatment, lymphocyte predominance (>50%) before treatment shifted to neutrophil predominance (>50%) after treatment.
In a retrospectively followed cohort of 98 TBM patients, the development or progression of tuberculomas after effective treatment was noted in 36.7% of patients. Paradoxical tuberculomas were linked with previously reported risk factors (hydrocephalus and meningeal enhancement) (
) and a newly identified factor (worsening CSF parameters including WBC count and neutrophils). Furthermore, we found that there was an obvious time interval between the early worsening CSF parameters and the late paradoxical tuberculomas.
We confirmed that the early worsening of CSF parameters was associated with later tuberculomas, and there have been several previous reports with similar findings. Coulter et al. describe a girl with TBM (given ATT at week 1) who presented a progressive rise in CSF WBC count with a marked increase in the neutrophil count at week 2–4, and complicated by multiple tuberculomas at week 15 (
). Pauranik et al reported a patient receiving ATT after three months; her CSF WBC increased to more than two-fold on admission to hospital (200/mm3 rise to 575/mm3), while CSF dominated cells convert from lymphocytes to neutrophils (70% lymphocytes transformation for 60% neutrophils), and CT re-examination at the following 3 months revealed intracranial tuberculoma (
). Additionally, Thwaites et al. reported that the presence of paradoxical tuberculomas on MRI after 60 days of treatment was associated with increased numbers of white cells in the CSF, a higher proportion of which were neutrophils (
). A study of a TBM case series reported paradoxical CSF reaction observed in 32.8% (20/61) of TBM patients, half of which subsequently developed symptomatic tuberculomas (
). Thus, a possible association between higher CSF WBC (neutrophils) during treatment and subsequent development of tuberculomas may provide important and novel insight into the pathogenesis of tuberculomas.
In this study, not only did the majority of worsening CSF parameters (27/42) occur in the paradoxical tuberculoma group, the magnitude of CSF WBC increase was also significantly greater than in the non-paradoxical tuberculoma group (Table2, Figure 2). The triggering of inflammatory reactions by a therapeutic intervention can be responsible for the underlying pathology of worsening CSF parameters and paradoxical tuberculomas, as their shared feature is the initiation of effective ATT. Various hypotheses have been put forward to explain the treatment-associated CSF changes. The most likely explanation is the interplay between the host’s immune response and mycobacterial products, i.e., the rapid killing of mycobacteria with adequate anti-tuberculous drugs may result in the release of large amounts of microbial components that stimulate an exuberant inflammatory response, leading to worsening of CSF parameters (
The observed link between neutrophil-dependent inflammation and tuberculomas is intriguing. Neutrophils are the first defensive cells recruited to tissue following infection, where their role has been thought to involve eliminating invading pathogens via mechanisms such as the generation of reactive oxygen species (
). This may lead to the subsequent appearance of tuberculomas. Additionally, neutrophils have immunomodulatory functions. When stimulated by the persistent mycobacterial antigen after receiving ATT, neutrophils release an array of cytokines and chemokines contribute to the upregulation of immunity that attracts other inflammatory cells (
). Several studies argue that if such immune enhancement occurs at microscopic intracranial foci, the result could be the development of tuberculomas at those sites (
). Consistently, Singh et al. conducted pathological analysis of a large number of tuberculomas and found that polymorphs predominated in tuberculous abscesses (
). Therefore, neutrophils are essential for granuloma formation during chronic Mycobacterium tuberculosis infection; the presence of worsening CSF parameters could be more common in the tuberculoma group.
We observed that 34 of 42 (80.95%) worsening CSF parameters of patients developed cellular changes within two weeks after starting ATT, and the worsening CSF disappeared in the following week. In contrast, tuberculomas almost always occurred two weeks after ATT, with one case delayed to 13 months. Most (94.4%) tuberculomas developed or progressed within 3 months of starting therapy. The period between worsening CSF parameters and tuberculoma development ranged from 6 to 383 days (median 21days). This period may be the formation of tuberculoma from microscopic to macroscopic. The absence of an association of increased CSF neutrophil levels at baseline with detection of later tuberculomas (Table1) suggests that a delayed decrease in inflammation, rather than the height of the early inflammatory response, determines the local histological outcome in the brain. These findings support the observation of continuous dynamic changes of CSF, which is of great significance for predicting the progression of later tuberculomas.
Paradoxical tuberculomas and worsening CSF parameters are the consequence of an exaggerated immune reaction that should be considered, thus the treatment is based on the use of immunomodulators and not in the change of anti-tuberculous drugs. Here seems to be some agreement as to the use of steroids in tuberculomas. In several case reports, enlargement of tuberculomas has been noted when steroids were reduced, concluding that steroids play a role in suppressing the immunological response which is thought to be responsible for the tuberculomas (
). The use of steroids was unable to prevent the development of tuberculomas in our series in which all of the patients were treated with steroids. We could not analyze an effect of adjuvant steroid therapy because all patients received steroids and death before the required follow-up (6 months) excluded patients from entry. Although clinical deterioration was seen in patients when symptomatic tuberculomas developed, we could not find any statistical difference between MRS at month 6 after ATT of TBM patients with and without tuberculomas, as has also been reported in other studies (
). We also found tuberculoma development is not a poor prognostic factor in TBM patients and does not require alteration in the treatment regimen, but it causes longer hospital stays and therefore needs a prolonged course of ATT.
There were certain limitations to our study. First, the persistent and high bacterial load (due to drug resistance, etc.) in the CSF has been associated with the development of complications, including tuberculomas (
) for the etiological diagnosis of TBM patients instead of culture and nucleic acid amplification technique, assessment of bacterial load and drug resistance testing was not performed. However, almost all of the patients enrolled improved without changing the anti-tuberculosis treatment, and the possibility of drug resistance was low. Second, we could not identify all tuberculomas, and worsening CSF parameter cases were included in this study due to insufficient data in some cases. Third, it is difficult to determine the exact time of occurrence of tuberculoma and worsening CSF parameters in clinical practice. Finally, since our data are observational, there was no in-depth immunological experimental study, and the effect of CSF cell components and cytokines on the formation of tuberculoma remains to be further investigated.
Conclusions
In conclusion, this study found that paradoxical tuberculomas occur in approximately one-third of patients with TBM despite the use of ATT and adjunctive corticosteroid therapy. Three-quarters of patients with paradoxical tuberculoma had worsening CSF parameters at an early stage. We emphasized that patients with worsening CSF parameters have the highest risk of developing paradoxical tuberculomas during subsequent treatment. Neutrophils play an important role in the formation of subsequent tuberculomas. In this way, we can better understand the pathophysiological process and the dynamic nature of the disease. Our findings suggest that more studies are needed to further understand the effect of worsening CSF parameters on paradoxical tuberculomas. Steroid use did not prevent the development of tuberculoma. Tuberculoma development was not a bad prognostic factor in TBM patients and does not require alteration in the treatment regimen, it causes longer hospital stays and we believe this requires a prolonged course of anti-tuberculosis therapy.
Conflict of interest
The authors state that they have no conflicts of interest.
Funding
This work was funded by the National Natural Science Foundation of China, grant number: 81801295.
Ethical approval
Ethical approval was obtained from the Institutional Ethical Committee of the Third Xiangya Hospital, Central South University.
Acknowledgments
We acknowledge the patients and their families, and the staff at the Third Xiangya Hospital, Central South University and Changsha Central Hospital, as well as members of the laboratory of the Department of Neurology, the Third Xiangya Hospital, Central South University.
References
Afghani B.
Lieberman J.M.
Paradoxical enlargement or development of intracranial tuberculomas during therapy: case report and review.
Relationship between intracranial granulomas and cerebrospinal fluid levels of gamma interferon and interleukin-10 in patients with tuberculous meningitis.
32173-1&id=gr1.jpg){kind=link}
32173-1&id=gr2.jpg){kind=link}