Brain MRI findings in relation to clinical characteristics and outcome of tuberculous meningitis.

Neuroradiological abnormalities in tuberculous meningitis (TBM) are common, but the exact relationship with clinical and inflammatory markers has not been well established. We performed magnetic resonance imaging (MRI) at baseline and after two months treatment to characterise neuroradiological patterns in a prospective cohort of adult TBM patients in Indonesia. We included 48 TBM patients (median age 30, 52% female, 8% HIV-infected), most of whom had grade II (90%), bacteriologically confirmed (71%) disease, without antituberculotic resistance. Most patients had more than one brain lesion (83%); baseline MRIs showed meningeal enhancement (89%), tuberculomas (77%), brain infarction (60%) and hydrocephalus (56%). We also performed an exploratory analysis associating MRI findings to clinical parameters, response to treatment, paradoxical reactions and survival. The presence of multiple brain lesion was associated with a lower Glasgow Coma Scale and more pronounced motor, lung, and CSF abnormalities (p-value <0.05). After two months, 33/37 patients (89%) showed worsening of MRI findings, mostly consisting of new or enlarged tuberculomas. Baseline and follow-up MRI findings and paradoxical responses showed no association with six-month mortality. Severe TBM is characterized by extensive MRI abnormalities at baseline, and frequent radiological worsening during treatment.

Introduction

Meningitis is the most severe manifestation of tuberculosis (TB), causing death or disability in up to 50% of those affected [1]. Mycobacterium tuberculosis is hypothesized to spread during an early bacteraemia phase, leading to granuloma development in all organs including the brain [2]. TB meningitis (TBM) occurs as a result of later rupture of meningeal or para-meningeal granulomatous lesions in the subarachnoid space. Entrapment of penetrating arteries, cranial nerves and the ventricular system, can lead to stroke, cranial nerve palsy and hydrocephalus, respectively [2, 3]. As such, localisation, extent and nature of the inflammatory process probably have a major effect on the clinical manifestations and outcome of TBM.

Imaging plays a major role in detection of brain abnormalities related to TBM. Magnetic Resonance Imaging (MRI) offers a non-invasive soft tissue contrast imaging modality that offers high spatial resolution without ionizing radiation [4]. MRI can facilitate the exploration of the morphological features of the human brain, hence advancing our insights into the neurobiology processes of a disease [5]. MRI has a much higher discriminatory value compared to computed tomography (CT), in particular for detecting most pathologic meningeal conditions [6]. Previous studies using MRI have found the common triad of neuroradiological findings in TBM, i.e. basal meningeal enhancement, hydrocephalus, and infarction [3].

Several factors reassure the image quality of MRI, including signal-to-noise ratio (SNR) related to slice thickness and the patient’s ability to remain motionless during examination [7, 8]. The sequence used and reading methods is also important [7]. Gadolinium-enhanced axial T1 two-dimension spin echo (2D-SE), one of the most important sequences, is often used for detecting pathology in the central nervous system (CNS). However, slices thinner than 3 mm are not feasible due to the inferior signal-to-noise ratio, and the subsequent slice gap may result in missing small lesions [9]. Magnetization-prepared rapid gradient-echo (3D MP-RAGE) imaging is a three-dimensional MRI sequence which produces slices with thickness of 1 mm or less, yielding an image with superior tissue contrast and higher spatial resolution compared to 2D-SE [9, 10]. Administration of gadolinium-diethylenetriaminepentaacetic acid (DTPA) allows for visualization of inflamed meninges [11]. We used MRI 3D MP-RAGE to examine the neuroradiological patterns, and explored the possible association of neuroradiological findings to clinical characteristics and outcome in a prospective cohort of clinically well-characterised TBM patients in Indonesia.

Material and methods

Setting, patients and follow-up

We included consecutive patients enrolled in a double-blind randomized clinical trial on high dose rifampicin for tuberculous meningitis between December 2014 and June 2016 in Hasan Sadikin Hospital [12], Bandung, Indonesia—the referral hospital for the province of West Java. Patients above 14 years of age with signs and symptoms of tuberculous meningitis, combined with a CSF/blood glucose ratio <0.5 and CSF leukocyte count ≥5, and with negative CSF India Ink and Gram staining were included. CSF microscopy (Ziehl-Neelsen), liquid M. tuberculosis culture (MODS), and GeneXpert MTB/RIF were done on large (>5 ml) CSF samples in all patients [13]. A detailed medical history and clinical examination was performed, and blood and CSF inflammatory markers as well as chest X-ray abnormalities were assessed before start of anti-tuberculous drugs.

TBM was treated with a combination of rifampicin, isoniazid, ethambutol and pyrazinamide for at least 6 months. All patients were given adjunctive dexamethasone in a tapering regimen [14]. Two thirds of our patients had one month of double or triple rifampicin dosing as a study drug used in this trial. Patients were followed for 6 months. The clinical outcome was death or severe disability 6 months after inclusion. Clinical improvement after anti-tuberculosis administration was assessed using Glasgow Outcome Scale (GOS). Good recovery defined as score GOS 4 or 5 [15]. In case the patient experienced a symptomatic paradoxical response, we increased the dose of dexamethasone back to the starting dose accordingly to their TBM grade.

Written informed consent to participate in the trial was obtained from all patients or from their relatives if the patient could not provide informed consent. Consent was registered in a REDCap clinical research form for all patients, and the same procedure was followed for patients over 14 and under 18 years of age, who are considered adult according to local custom. All study procedures were performed in accordance with relevant guidelines and regulations. This study and the written informed consent were approved by the ethical review of the Medical Faculty of Universitas Padjadjaran, Bandung, Indonesia, No.330/UN6.C1.3.2/KEPK/PN/2016. This trial is registered with ClinicalTrials.gov as trial number NCT02169882.

MRI protocol

Baseline brain MRI was performed within 5 days after diagnosis and start of treatment to capture baseline abnormalities, and again after 2 months (plus or minus 1 week) of anti-tuberculous treatment as a study related protocol using the 1.5 Tesla systems (Magnetom Essensa, Siemens Healthcare, USA. The MRI sequences included T1 spin echo, axial T2 turbo spin echo, axial T2 fluid attenuation inversion recovery (FLAIR) and axial diffusion weighted imaging (DWI)—apparent diffusion coefficient (ADC), and axial T2 gradient echo (GRE) with and without contrast media, and axial T1 3D magnetization-prepared rapid acquisition with gradient-echo (MP-RAGE) (isotropic) with and without intravenous contrast media. Post-contrast study was carried out using Gadolinium 0.2 mmol/kg/body weight intravenously. The 3D MP-RAGE is a high-resolution three-dimensional (3D) MRI sequence of 1 mm thick slices resulting in more precise anatomic localization of lesion morphology [10]. All imaging was performed in the axial plane and with identical geometrical parameters. All images were evaluated using OsiriX MD 8.0 for Mac, Pixmeo, SARL, Switzerland.

Systematic assessment of the imaging studies was undertaken and quantified by consensus. Radiologist (RH, FA, RDS) and a neurologist (SD) evaluated all MRIs for the presence of enhancement of the leptomeninges, and presence of hydrocephalus, tuberculomas, brain infarction and cranial nerve enhancement and did not aware to treatment arm and outcome. The location of the infarction was also specified as: cerebrum, cerebellum or brain stem. Cerebral infarction was further categorized into inside or outside the basal ganglia and thalamus. Infarction was categorized as in the basal ganglia if located in the striatum or caudate nucleus, putamen, and globus pallidum [16].

Paradoxical response

MRIs after 2 months treatment were compared with baseline MRIs. Radiological paradoxical responses were defined as worsening of pre-existing tuberculous lesion or appearance of new lesions in the second brain MRI (, whereas a clinical paradoxical response was defined as a new neurological event, including cranial nerve palsy, motor deficits, seizures, or severe headache, with or without the worsening of brain imaging, in patients whose condition initially improved with anti-tuberculous treatment [17]. Clinical-radiological paradoxical responses were defined as the presence of both radiological and clinical paradoxical responses. Paradoxical response was considered as ‘definite’ if it occurred more than 4 weeks after commencement of anti-tuberculous treatment [17].

General condition

Definite or probable tuberculous meningitis patients with:

Initial improvement during anti-TB therapy

> 28 days of treatment

Compliance to anti-tuberculosis treatment

No anti-tuberculosis drugs resistance

Who present with new or worsening clinical or radiological signs.

Radiological paradoxical response

Development of new or worsening lesions in radiological finding, including:

Hydrocephalus

New event of hydrocephalus either communicating or non-communication or worsening of pre-existing one; or increasing area of trans ependymal leakage (periventricular edema) from the hydrocephalus.

Leptomeningeal enhancement

Formation of new leptomeningeal enhancement in other areas of subarachnoid spaces; or increasing (thickening) of the previous leptomeningeal enhancement.

Tuberculoma

Increasing size of previous tuberculoma; or new or additional formation of tuberculoma at different areas of the brain or subarachnoid spaces; or formation or increasing oedema surrounding the tuberculoma; or increasing mass effect from the tuberculoma such as increasing effacement of sulci and cisterns, increasing midline shift; or new or worsening sign of herniation.

Infarction

New or additional formation of infarction whether it is new area of acute infarction or new area of chronic infarction. Acute infarction is defined as infarct lesion that gives restricted diffusion appearance on DWI-ADC sequence of the MRI. Chronic infarction is defined as infarct lesion that does not give restricted diffusion appearance on DWI-ADC sequence of the MRI. including the formation of new infarction in other areas of the brain or another vasculatory system with or without the sign of vasculitis

Data analysis and statistics

Patient characteristics were presented as medians or proportions as indicated. Clinical characteristics and six-month survival and functional outcome were compared between patients with or without MRI findings at baseline, and with or without a paradoxical worsening of MRI findings after two-months, using Kruskal-Wallis test for continuous variables and Chi-square test for categorical variables. All statistical analysis was performed using IBM SPSS Statistics version 24, p-value of <0.05 were considered significant.

Results

Baseline MRI findings

Sixty patients (52% female, median age 30.5 years, 8% HIV co-infected) were enrolled in a phase IIb dose-finding clinical trial, 48 had baseline MRIs and were eligible for this study. They had an MRI after a median 2.5 days of treatment. Most of them had MRC grade II disease (90%), and most presented with headache (96%), neck stiffness (96%), fever (85%), loss of consciousness (83%) and motor deficits (56%). Thirty three also had pulmonary involvement (69%), including 4 (12%) with miliary disease. All patients showed cerebrospinal fluid abnormalities typical of TB meningitis, and 34 (71%) had bacteriologically proven TBM based on microscopy (13/34), molecular testing (21/34) or culture (31/34). None of the cultured M. tuberculosis isolates were found to be rifampicin-resistant by GeneXpert.

At baseline, 45 out of 48 patients (94%) showed MRI abnormalities (Table 1), 40 (83%) had more than one brain lesion. The most common lesion was meningeal enhancement, mainly in the basal meninges and sylvian fissure (Fig 1A–1C). Tuberculomas were the second most common finding, most of them a miliary type (Fig 1D), while only one patient had a pseudo-abscess (Fig 1E1–1E4). On the contrary, brain infarctions were seen in 29 (60%) patients, mostly acute rather than chronic (Fig 1F1–1F3). Hydrocephalus was also common (56%), and always communicating, characterized by a broader callosal angle (Fig 1G), void signal appearance (Fig 1H), dilated temporal horn (Fig 1I), and larger of Evans’ ratio (Fig 1J). Finally, 19% of our patients had cranial nerve imaging abnormalities (Fig 1K).

Table 1

Neuroradiological abnormalities at baseline.

	Tuberculous meningitis (n = 48)
Meningeal enhancement	39/48 (81%)
Basal meninges	25
Sylvian fissure	22
Convexity	18
Ventricular	3
Hydrocephalus	27/48 (56%)
Communicans	27
Non-communicans	0
Tuberculoma	37/48 (77%)
Miliary	35
Non-miliary	24
Pseudo abscess	1
Brain Infarction	29/48 (60%)
	Acute, n = 26 (90%)	Chronic, n = 4 (14%)
Cerebrum, basal ganglia & thalamus	22	2
Cerebrum, outside basal ganglia & thalamus	22	2
Cerebellum	2	0
Brainstem	5	0
Cranial nerve enhancement	9/48 (19%)

Meningeal enhancement was defined as linear or nodular enhancement of meninges with contrast media [33] at one or more locations: the basal meninges (e.g. basal cistern, ambient cistern, quadrigeminal cistern, prepontine cistern, cerebellopontine cistern, suprasellar cistern, premedullary cistern), sylvian fissure, cerebral or cerebellar convexity/sulci and ventricular system [24]. Hydrocephalus was present if one or more of: dilated temporal horns of lateral ventricles, ballooning of frontal horns of lateral ventricle, ballooning of third ventricle, narrowed callosal angle, presence of flow void in T2W images at the Sylvian aquaducts [34]. Evans’ index [35] of every patients with and without hydrocephalus were measured and compared. Communicating and non-communicating hydrocephalus were defined based on the absence or presence of an obstructing lesion along the intraventricular CSF pathways[34]. Tuberculomas were defined as the presence of nodular or ring enhancement with contrast media [33, 36] and specified as milliary (<2mm) or non-milliary (>2mm) tuberculomas, or pseudo-abscesses [37]. Abscesses were defined as the presence of ring enhancement with restricted diffusion appearance on DWI-ADC [38]. Acute infarctions were defined as lesions with restricted diffusion on DWI-ADC, and increased T2-weighted and fluid-attenuated inversion recovery (FLAIR) signal intensity [39]. Cranial nerve neuropathy was defined as enhancement with or without thickening of the oculomotor nerve, trigeminal nerve, abducens nerve, facial nerve or vestibulocochlear nerve [40].

Fig 1

Common baseline MRI findings in adults with tuberculous meningitis.

Meningeal enhancement at basal meninges (A), right Sylvian fissure (B), and ventricular (C); non-miliary and miliary tuberculomas (D); pseudo abscess in axial DWI (E1), axial ADC (E2), axial T2 FLAIR (E3) and axial T1W1 post contras (E4); multiple acute infarctions at the left basal ganglia in axial DWI (F1), axial ADC (F2), and axial T2 FLAIR (F3); Communicating hydrocephalus with narrowed Callosal angle (G), void signal in the aqueduct (H), and dilated temporal horn (I), broader of Evans’ ratio (J); and oculomotor nerve enhancement (K).

We next explored the relation between MRI findings and clinical characteristics. We divided into 5 groups of patient in order to weighted the brain abnormalities: normal, single, two, three, and four abnormalities. Patients with multiple brain abnormalities had a longer duration of illness, lower consciousness, higher numbers of motor deficits and cranial nerve palsy, more CSF abnormalities and higher positivity rate of CSF culture (Table 2). As only 4 (8.3%) of patients were HIV-infected, analysis was not stratified by HIV-status. Median time between onset of neurological symptoms and MRI (time to MRI) was 19 days with an interquartile of range 12–27. In univariate logistic regression analysis, time to MRI was not significantly associated with 6-month mortality (OR 1.02 [0.98–1.05], p = 0.315).

Table 2

Relation between disease characteristics and neuroimaging abnormalities.

	No Abnormality	single abnormality	two abnormalities	three abnormalities	four abnormalities
	n = 3 (6%)	n = 5 (10%)	n = 9 (19%)	n = 15 (31%)	n = 16 (33%)
Sex, Male	2 (67)	1 (20)	5 (56)	7 (47)	8 (50)
Age, years	19 [17–19]	33 [25–47]	36 [30–46]	24 [20–33]	32 [22–40]
Duration, days	30 [7-]	30 [10–90]	8 [7–60]	30 [14–90]	33 [30–90]
GCS	15	14 [13–15]	11 [11–13]	13 [12–14]	12 [12–13]
Grade    I	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)
II	2 (67)	5 (100)	7 (78)	14 (93)	15 (94)
III	1 (33)	0 (0)	2 (22)	1 (7)	1 (6)
Headache	3 (100)	5 (100)	8 (89)	14 (93)	16 (100)
Neck stiffness	3 (100)	5 (100)	9 (100)	13 (87)	16 (100)
Seizures	1 (33)	0	1 (11)	1 (7)	1 (6)
Motor deficits	1 (33)	1 (20)	4 (44)	9 (60)	12 (75)
Cranial nerve palsy	1 (33)	3 (60)	7 (78)	13 (87)	13 (81)
Temperature °C	38 [37-]	37 [36–38]	37 [37–38]	38 [37–39]	38 [37–38]
Fever	1 (100)	3 (60)	6 (67)	15 (100)	14 (88)
Chest x-ray abnormalities	0	1 (20)	8 (89)	10 (67)	14 (88)
Any bacteriological test positive	0	3 (60)	6 (67)	11 (73)	14 (88)
CSF culture positivity	0	3 (60)	5 (56)	10 (67)	13 (81)
HIV-infected	1 (33)	0	0	2 (13)	1 (6)
6-month mortality	1 (33)	1 (20)	3 (33)	3 (20)	3 (19)
6-month GOS	0	1 (20)	1 (11)	0	3 (19)
Leukocytes (cells, μl)	22 [8-]	480 [224–763]	199 [152–456]	176 [84–215]	258 [149–532]
PMN (cells/μl)	326 [19-]	52 [30–252]	34 [1–160]	57 [35–137]	93 [149–532]
% of total CSF leukocytes	14	25 [21–38]	25 [21–71]	31 [7–54]	38 [18–61]
MN (cells/μl)	137 [115-]	155 [95–369]	136 [7–332]	110 [46–176]	124 [67–371]
% of total CSF leukocytes	86 [74-]	75 [62–80]	75 [29–80]	69 [46–93]	62 [39–371]
CSF protein (mg/dl)	158 [32-]	165 [89–182]	209 [141–1434]	188 [71–308]	249 [160–402]
CSF/blood glucose ratio	0.22 [0.21-]	0.25 [0.2–0.40	0.17 [0.13–0.24]	0.25 [0.11–0.36]	0.13 [0.09–0.19]
Haemoglobin (mg/dL)	14 [11-]	13 [11–13]	13 [10–15]	11 [10–13]	12 [10–14]
Leukocyte (109 cells/L)	10 [7-]	10 [6–13]	1 [7–14]	11 [9–14]	11 [9–14]
Neutrophils (cells/ul)	9 [6-]	9 [4–10]	10 [6–12]	10 [7–12]	9 [8–12]
% of total blood leukocytes	84 [79-]	81 [60–88]	84 [79–86]	84 [70–91]	83 [79–88]
Lymphocytes (109 cells/L)	0.9 [0.8-]	0.9 [0.7–1.9]	1.1 [0.8–1.6]	1.2 [0.6–1.4]	0.8 [0.8–1.1]
% of total blood leukocytes	12 [9-]	13 [8–21]	9 [8–15]	11 [5–21]	8 [6–9]
Monocytes (109 cells/ul)	0.5 (0.3-)	0.6 (0.2–1.1)	0.7 (0.5–1)	0.7 (0.5–1)	0.8 [0.4–1.0]
% of total blood leukocytes	4 (3-)	6 (4–13)	6 (5–10)	5 (4–10)	8 [5–10]
Thrombocyte x 109/L	243 [210-]	376 [287–425]	290 [217–338]	312 [239–468]	479 [411–515]
Blood sodium (mEq/dL)	135 [124-]	127 [122–132]	123 [116–136]	124 [119–133]	124 [116–129]

Data are % of patients for categorical data or median value (IQR = interquartile range) for continuous data. Abbreviations: GCS, Glasgow Coma Scale; HIV, Human Immunodeficiency Virus; CSF, Cerebrospinal Fluid; PMN, Polymorphonuclear; MN, Mononuclear.

Bold: p-value <0.05 in comparison Krusskal-wallis test for numerical data and Chi-square test for categorical data. Fever: body temperature >38.5°C.

MRI findings after two months drug treatment

After two months of treatment, one patient was not eligible for a second MRI because of insertion of a metal device for spinal TB, and one had movement artefacts and nine patients died before day 60. Among the remaining 37 patients, 89% (33/37) had new or worsening MRI findings. Only in 39% (13/33) of cases, this was accompanied by worsening clinical symptoms, the most common finding was new cranial nerve abnormalities. Two patients had worsening clinical symptoms without new MRI abnormalities: one patient developed a new central facial and hypoglossal nerve paresis at day 30, and another one had ptosis at day 30.

Among all patients with radiological worsening, 82% (27/33) had new or enlarged tuberculomas and 76% (25/33) had thicker or new location of meningeal enhancement.

(Fig 2). In addition, 24% (8/33) had new cranial nerve enhancement, one had developed a new infarction and two enlargements of hydrocephalus. In contrast to worsening meningeal enhancement and tuberculoma’s, hydrocephalus was improving in 32% (12/37) of patients (Table 3).

Fig 2

Paradoxical response with basal meningeal enhancement after 60 days treatment.

Basal meningeal exudate before (A) and after two months after anti-tuberculosis drugs (B). Both from T1-W1 post contras from one patient with radiological worsening at day 60 days after anti-tuberculosis treatment.

Table 3

Neuroimaging changes after two months of treatment in 37 patients.

Abnormality	Paradoxical worsening	No change	Improving
Meningeal enhancement	25	9	3
Tuberculoma	27	8	2
Hydrocephalus	2	23	12
Infarction	1	36	0
Cranial nerve neuropathy	8	28	1

Paradoxical worsening was evaluated at day 60±1 week based on 37 paired MRI.

Neuroimaging in relation to patient outcome

Next, we examined MRI findings in relation to clinical outcome. As shown in Table 2, baseline MRI abnormalities did not associate to survival. When looking into the patients with paradoxical reactions at 2 months, 1 of 2 patients with clinical paradoxical reactions died, 0 in the group of radiological paradoxical reactions, and 2/13 (15%) in the group with clinical-radiological paradoxical reactions. However, the patients who had no paradoxical reactions or only had radiological changes without clinical worsening had more number of patient with good status recovery after 6 months of treatment (Table 4). These numbers were too small to perform formal statistics.

Table 4

Outcome of paradoxical response.

	Clinical and radiological (n = 13)	Clinical (n = 2)	Radiological (n = 20)	No Paradoxical response (n = 2)
6-month mortality	2 (15)	1 (50)	0	0
6-month good recovery	7 (54)	1 (50)	20 (100)	2 (100)

Good recovery defined as score Glasgow Outcome Scale (GOS) 4 or 5.

Discussion

We studied neuroradiological abnormalities in a prospective and well-characterised cohort of TBM patients in Indonesia. Using 3D MP-RAGE imaging we found abnormalities in virtually all patients, with meningeal enhancement, tuberculomas, infarction, and hydrocephalus being the most frequent. Multiple MRI abnormalities were associated with more severe clinical presentations and more pronounced CSF abnormalities in an exploratory analysis. Many patients showed new or worsening MRI findings during the first two months of treatment, especially miliary tuberculomas and meningeal enhancement. Although caution is warranted because of the size of the study, baseline MRI findings showed no significant association with the occurrence of paradoxical worsening, disability or death during follow-up.

Neuroradiological abnormalities in our setting were more common than in previous reports [18, 19]. This may be partially explained by the fact that TBM patients in this setting usually present late, with advanced disease [20]. The thick basal exudate that developed in the later stage of disease may entrap penetrating arteries and cranial nerves, and may block CSF resorption [2, 3]. Minimal lesions in early disease may be missed on conventional spin-echo images, especially if acquired without contrast [21]. Another possible explanation for the difference with previous reports is that we had a high proportion of microbiologically confirmed TBM, which we have previously found to be associated with inflammation [22]. The three patients without abnormalities on brain MRI in this study also had negative microbiological test results.

Our finding match with earlier observations, which showed that most TBM patients present with multiple baseline MRI abnormalities [23, 24], predominantly meningeal enhancement and tuberculoma [6, 21, 24–26]. Meningeal enhancement is mainly found at the basal subarachnoid cisterns. Dense gelatinous leptomeningeal exudates were the commonest pathology in TBM [20, 27]. In our patients, more extensive brain abnormalities were associated with a lower GCS, motor and cranial nerve palsies, presence of lung TB, microbiological confirmation, and a lower CSF glucose and higher CSF protein, reflecting inflammation or more severe disease. Our result differs from a study in Vietnam that did not find any association between baseline MRI findings on admission with most of clinical or CSF inflammatory markers [24].

The majority of our patients had tuberculomas at baseline, much more than that in previous studies [24-26]. Tuberculomas were predominantly of a miliary type, match those observed in earlier studies [23, 24]. The higher proportion of tuberculomas in our study may be due to the use of the sensitive 3D MP-RAGE. The 3D MP-RAGE sequence used in our protocol produces slices with a thickness of 1 mm without any gap. This allows for robust detection of brain lesions, including small tuberculomas and minimal meningeal enhancement.

Newly developing or enlarging intracranial tuberculomas following initial improvement may be observed despite appropriate anti-tuberculosis therapy [28, 29]. Many of our patients indeed showed worsening of MRI findings (so-called ‘paradoxical reactions’) after two months anti-tuberculosis treatment, with new miliary tuberculomas and meningeal enhancement being the most common. Interestingly, the majority of these radiological paradoxical reactions were not accompanied by new clinical findings. The present finding is consistent with other studies that indicate that the most common radiological worsening was an asymptomatic increase in number or size of tuberculomas [24, 30]; mostly occurring within 3 months of initiating anti-tuberculosis treatment [30].

In agreement with previous studies in India [25, 31] and Vietnam [24], MRI appearances at baseline and two months were not associated with death or severe disability after six months of the treatment, although this part of our analysis was only exploratory and may be underpowered. Indeed, in larger studies, hydrocephalus [32] and infarction [20] significantly predicted TBM mortality. In accordance with our result, previous studies have demonstrated that paradoxical reactions did not predict 6-month mortality [30]. In our study, brain infarctions and tuberculomas were mostly small without a mass-effect leading to brain oedema. In line with earlier studies [3, 4, 21] we only found one case of tubercular abscess; a large solitary lesion with ring enhancement and surrounding vasogenic oedema and mass effect. However, in contrast to earlier findings [31], we found a better functional recovery among patients without paradoxical reactions, presumably caused by preservation of corticospinal and corticobulbar tract that might be affected by the ischaemic damage of new or enlarged tuberculoma. In this study, patients with tuberculoma’s more often presented with motor and cranial nerve abnormalities compared to those without tuberculomas regardless of other brain lesion (p = 0.032 and 0.001). It is important that the finding of a paradoxical reactions is not interpreted as treatment failure, presence of an alternative diagnosis or drug-resistance as this may lead to hazardous decision of TB-drug cessation.

Strengths of our study are an extensive descriptive analysis facilitated by 3D MP-RAGE in the MRI protocol, in a well-described clinical cohort, with high proportion of microbiologically confirmed disease (71%). However, the study’s sample size limits the possibility of drawing definite conclusions in correlating neuroradiological findings to clinical parameters. Also, our study was not designed to help identify TBM-specific abnormalities as it lacked a control group with patients suffering different brain infections.

In conclusion, MRI abnormalities in this group of TBM patients were common, and paradoxical reactions often occurred despite use of steroids. This result provides further support for using neuroradiological imaging in TBM diagnosis and treatment evaluation, including the evaluation of paradoxical reactions. Also, MRIs could help characterize immunopathology in drug trials or studies focusing on development of new host-directed strategies for TBM.

(XLSX)

Click here for additional data file.

24 Mar 2020

PONE-D-19-33449

Brain MRI findings in relation to clinical characteristics and outcome of tuberculous meningitis

PLOS ONE

Dear dr Dian,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

I would also like to apologize for the time that it has taken to render a decision on your manuscript. I have had a very difficult time securing a second reviewer. However, I believe that the comments that have been received will be valuable in improving the quality of the manuscript. Please carefully address all of the comments that have been raised, or explicitly state why this has not been done. In addition, I have the following points that should be addressed:

1. Line 135: Please change "Swiss" to "Switzerland"

2. Line 143: You state that the basal ganglia was considered as "caudate nucleus, internal capsule, lentiform nucleus and thalamus." This is incorrect. The thalamus is not part of the basal ganglia, and there other brain structures that are part of the basal ganglia, but not listed here (e.g. nucleus accumbens, olfactory tubercle, substantia nigra, and subthalamic nucleus.). There appears to be some confusion here based on the cited reference (Tai, et al. Scientific Reports volume 6, Article number: 38802 (2016)) . Please clarify.

3. Line 251: It is a bit misleading to refer to a "thin gap" in the 3D MP-RAGE sequence. As it is a 3D sequence, there is no gap at all.

We would appreciate receiving your revised manuscript by May 08 2020 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Niels Bergsland

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include the registration number for the clinical trial referenced in the manuscript.

3. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I read carefully the manuscript PONE-D-19-33449

First i will like thanks authors for the quality of the manuscript which is easy to read and well organized

This topic in Tuberculosis disease is rare and recent datas are needed to improve the management of such patient world wide.

The authors through their MRI focus reinforce previously published message on MRI major place as neurological diagnostic tool for TB.

Major concern

- Could the authors give information on time between first symptoms and and first MRI ? It might be more relevant than delay between treatment begining and MRI.

Line 196 "Only in 39% (13/30) of cases..." I do not understand how you find 30 because just before you speak about 33 cases, please fix it.

- Lines 196-1977 the sentence " this was accompanied by worsening clinical symptoms, the most common finding being …" is incomplete, so please complete it ?

- What was the dexamethasone posology and what tappering paln do you use ?

- What have you done with dexamethasone dosis when worsening was proved clinically and/or radiologically ?

- What were the pulmonary findings for these patients (Chest X-Ray and sputum samples) and have you seen association between type of lung involvement and neurological loacalization of TB ?

- You stopped follow up at M6 but neurological tuberculosis need up to 9 months of treatment. Could you argue why a such outcome? Have you datas on M6-M12 periods?

Minor

There is some publication in western country on this topic you could cite :

- A. Bleibtreu et al. / Médecine et maladies infectieuses 48 (2018) 533–539

-Venkatraman N, King T, Bell D, Woltmann G, Wiselka M, AbubakarI, et al. High levels of neurological involvement but low mortalityin miliary tuberculosis: a 6-year case-series from the UK. Eur RespirJ 2016;47(5):1578–81

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Dr Bleibtreu Alexandre MD, PhD. Infectious diseases Specialist

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

4 May 2020

Authors’ Response to the Review Comments

Journal Plos One

Manuscript# PONE-D-19-33449

Title of Paper Brain MRI findings in relation to clinical characteristics and outcome of tuberculous meningitis

Authors Sofiati Dian, Robby Hermawan, Arjan van Laarhoven, Sofia Immaculata, Tri Hanggono Achmad, Rovina Ruslami, Farhan, , Ristaniah D. Soetikno, Ahmad Rizal Ganiem, Reinout van Crevel

We appreciate the time and efforts by the editor and referees in reviewing this manuscript. We have addressed all issues in the review report and hope that our revised manuscript is now suitable for publication in Plos One.

Response to Comments from Academic Editor

Comments:

I would also like to apologize for the time that it has taken to render a decision on your manuscript. I have had a very difficult time securing a second reviewer. However, I believe that the comments that have been received will be valuable in improving the quality of the manuscript.

We would like to thank Dr. Bergsland as the academic editor for the positive appraisal of our manuscript. We greatly appreciate your efforts to carefully review the paper and the valuable suggestions offered.

Suggestions:

1) Line 135: Please change "Swiss" to "Switzerland” (Line 136)

Done

2) Line 143: You state that the basal ganglia were considered as "caudate nucleus, internal capsule, lentiform nucleus and thalamus." This is incorrect. The thalamus is not part of the basal ganglia, and there other brain structures that are part of the basal ganglia, but not listed here (e.g. nucleus accumbens, olfactory tubercle, substantia nigra, and subthalamic nucleus.). There appears to be some confusion here based on the cited reference (Tai, et al. Scientific Reports volume 6, Article number: 38802 (2016)). Please clarify.

We thank the editor for the sharp attention on this matter. Indeed, we made an unclear statement about the categorization and have already made a correction on it.

Original sentence:

Cerebral infarction was further categorized into inside or outside the basal ganglia. Infarction was categorized into basal ganglia if located in caudate nucleus, internal capsule, lentiform nucleus and thalamus.

New sentence:

Cerebral infarction was further categorized into inside or outside the basal ganglia and thalamus. Infarction was categorized as in the basal ganglia if located in the striatum or caudate nucleus, putamen, and globus palidus.(1) (Line 142-144, table 1 below category “Brain Infarction”)

We put thalamus and basal ganglia into one category as due to their vascularization are shared similar anatomical feature and are similarly affected by vasculitis. TB meningitis may induce vasculitis of small and medium-sized cerebral arteries, often the lenticulostriate arteries or the posterior cerebral branches and the thalamoperforating arteries that send their blood supply to basal ganglia and thalamus respectively.(2)

In the definition of basal ganglia, we did not mention the nucleus accumbent and olfactory tubercle because both structures are too small to be detected reliably by MRI.(1) The substantia nigra and subthalamic nucleus were excluded as they are located in the midbrain and are not part of cerebral structure.

2) Line 251: It is a bit misleading to refer to a "thin gap" in the 3D MP-RAGE sequence. As it is a 3D sequence, there is no gap at all.

The editor’s statement is very relevant, indeed we intended to mention “thickness” and not “gap” in this sentence. We have rephrased this sentence.

Original sentence:

The thin gap between slices in our protocol allowed a robust detection of brain lesion, including tuberculomas and meningeal enhancement.

New sentence:

The 3D MP-RAGE sequence used in our protocol produces slices with a thickness of 1 mm without any gap. This allows for robust detection of brain lesions, including small tuberculomas and minimal meningeal enhancement. (Line 263-265)

Response to Comments from Reviewer

Comments:

First, I will like thanks authors for the quality of the manuscript which is easy to read and well organized. This topic in Tuberculosis disease is rare and recent data are needed to improve the management of such patient worldwide. The authors through their MRI focus reinforce previously published message on MRI major place as neurological diagnostic tool for TB.

We thank also this reviewer for his/her compliments and constructive review.

Major concern:

Comments:

1. Could the authors give information on time between first symptoms and first MRI? It might be more relevant than delay between treatment beginning and MRI.

It should be noted that MRI was not used for diagnosis of TB meningitis in our setting. Still, we have added the time between onset of neurological symptoms and MRI (median was 19 days with an interquartile of range 12-27. In univariate logistic regression analysis, time to MRI was not significantly associated with 6-month mortality (OR 1.02 [0.98-1.05], p=0.315). (Line 194-197)

2. Line 196 "Only in 39% (13/30) of cases..." I do not understand how you find 30 because just before you speak about 33 cases, please fix it.

Thank you for pointing out this mistake, it was a typo and we have corrected the denominator in the newer version of the manuscript. (Line 204)

3. Lines 196-1977 the sentence " this was accompanied by worsening clinical symptoms, the most common finding being …" is incomplete, so please complete it?

The reviewer is correct, we have completed the sentence in the newer version of the manuscript:

Only in 39% (13/33) of cases, this was accompanied by worsening clinical symptoms, the most common finding was new cranial nerve abnormalities. Two patients had worsening clinical symptoms without new MRI abnormalities: one patient developed a new central facial and hypoglossal nerve palsy at day 30, and another one had ptosis at day 30. (Line 204-207)

4. What was the dexamethasone posology and what tapering plan do you use?

We adopted the doses of dexamethasone for tuberculous meningitis accordingly to the randomized controlled trial by Thwaites, 2004.(3) (We added the reference in Line 106)

Patients with grade I of BMRC:

- Week 1: 0.3 mg per kilogram per day intravenously

- Week 2: 0.2 mg per kilogram per day intravenously

- Week 3: oral treatment at a total of 0.1 mg per kilogram per day

- Week 4: oral treatment at a total of 3 mg per day

- Week 5: oral treatment at a total of 2 mg per day

- Week 6: oral treatment at a total of 1 mg per day

Patients with grade II or III:

- Week 1: 0.4 mg per kilogram per day intravenously

- Week 2: 0.3 mg per kilogram per day intravenously

- Week 3: 0.2 mg per kilogram per day intravenously

- Week 4: 0.1 mg per kilogram per day intravenously

- Week 5: Oral treatment at a total of 4 mg per day

- Week 6: Oral treatment at a total of 3 mg per day

- Week 7: Oral treatment at a total of 2 mg per day

- Week 8: Oral treatment at a total of 1 mg per day

In the case of early discharge, intra venous dexamethasone will be substituted with oral preparation, given in similar dose to continue dexamethasone administration until its completion, or switch to equivalent dose of methylprednisolone.

Equivalent dose: Dexamethasone 2mg ~ Methylprednisolone 12mg

5. What have you done with dexamethasone doses when worsening was proved clinically and/or radiologically?

Thank you for this question, we have rephrased the sentence about this to make it clearer.

Original sentence:

Corticosteroids were given to patients with symptomatic paradoxical response. (Line 114)

New sentence:

In case the patient experienced a paradoxical response, we increased the dose of dexamethasone back to the starting dose accordingly to their TBM grade. (Line 110-112)

6. What were the pulmonary findings for these patients (Chest X-Ray and sputum samples) and have you seen association between type of lung involvement and neurological localization of TB?

In this study, we found 33 (69%) patients with pulmonary TB, including 4 (12%) with miliary disease (lines 176-177). Among those 4, 2 had brain infarction in the basal ganglia and thalamus, 3 had brain infarction outside basal ganglia and thalamus, and 2 had brain infarction in the cerebellum and brain stem. Among 33 patients with pulmonary TB, 19 (58%) had brain infarction in the basal ganglia and thalamus, and only 2 (6%) and 3 (9%) had cerebellum or brainstem infarction. The patients who had normal chest x-ray (n=15) were mostly had no brain infarction 10/15 (67%) but had quite high proportion of tuberculoma 10/15 (67%). Of note, none of the patients had single location lesion (infarction or tuberculoma). We failed to find the association between type of this lung involvement with the neurological localization on TB.

6. You stopped follow up at M6 but neurological tuberculosis needs up to 9 months of treatment. Could you argue why a such outcome? Have you data on M6-M12 periods?

Minor

At the time of this randomized clinical trial, the national TB guideline in Indonesia recommended 6 months treatment.(4) We did not specifically collect data beyond 6 months of follow-up.

7. There is some publication in western country on this topic you could cite:

- A. Bleibtreu et al. / Médecine et maladies infectieuses 48 (2018) 533–539

-Venkatraman N, King T, Bell D, Woltmann G, Wiselka M, AbubakarI, et al. High levels of neurological involvement but low mortality in miliary tuberculosis: a 6-year case-series from the UK. Eur Respir J 2016;47(5):1578–81

Thank you for the references. We have read the publication with a great interest. Both articles recommend the neuroimaging assessment in patients with lung miliary tuberculosis (TB), since they found high incidence of brain lesions among patients with miliary TB, despite no neurological symptoms. We personally agree with this recommendation. However, since our study included patients with clinical and CSF findings compatible to tuberculous meningitis, we could not relate this finding with our findings.

Reference

1. Lanciego JL, Luquin N, Obeso JA. Functional neuroanatomy of the basal ganglia. Cold Spring Harb Perspect Med. 2012;2(12):a009621.

2. Abdel Razek AA, Alvarez H, Bagg S, Refaat S, Castillo M. Imaging spectrum of CNS vasculitis. Radiographics. 2014;34(4):873-94.

3. Thwaites GE, Nguyen DB, Nguyen HD, Hoang TQ, Do TT, Nguyen TC, et al. Dexamethasone for the treatment of tuberculous meningitis in adolescents and adults. N Engl J Med.2004;351(17):1741-51.

4. WHO. Treatment of Tuberculosis guidelines. Geneva, Switzerland.2010.

When submitting your revision, we need you to address these additional requirements.

1. We have revised this manuscript as requested accordingly to the PLOS ONE's style requirements, including those for file naming.

2. We have included the registration number for the clinical trial referenced in the manuscript.

3. Upon the repository information, all relevant data are within the manuscript and additional data will provide when asked.

19 Jun 2020

PONE-D-19-33449R1

Brain MRI findings in relation to clinical characteristics and outcome of tuberculous meningitis

PLOS ONE

Dear Dr. Dian,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we have decided that your manuscript does not meet our criteria for publication and must therefore be rejected.

I recognize that you have waited for a long time to have a decision on the manuscript. The manuscript was sent to an additional reviewer who specifically assessed the statistical analysis. The reviewer had major concerns about the statistical approach, resulting in severe doubts about the validity of the results.

I am sorry that we cannot be more positive on this occasion, but hope that you appreciate the reasons for this decision.

Yours sincerely,

Niels Bergsland

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: No

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: No

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: All concerns were adressed by the authors. The response to reviewers concerns is point by point argue and clear.

I am agree with their decision to not improved the references.

For me the manuscript is suitable for publication

Reviewer #2: In this study, the authors used MRI data collected at baseline in 48 patients with tuberculous meningitis (TBM). They then attempted to test the association between MRI findings (which included meningeal enhancement, tuberculomas, brain infarctions, and hydrocephalus).and clinical characteristics (motor deficits, cranial nerve palsy, microbiological confirmation, and CSF abnormalities). This analysis, however is flawed because none of these comparisons are “clean.” This is due to the fact that so many of the patients had multiple MRI abnormalities. It doesn’t make sense to test each abnormality separately against the others and make conclusions about an individual abnormality given that there are a large number of possible permutations of abnormalities that could occur in any given patient. There is no way to be certain that any one abnormality is associated with any clinical characteristic, since the abnormalities so frequently occur in combination.

The way to do this correctly would be to create a model in which the presence or absence of the clinical characteristic is the outcome, and a binary variable is created for each of the potential MRI abnormalities (allowing for one subject to have multiple abnormalities). It is doubtful that there are enough data to give sufficient power to such an analysis. Furthermore, it would also be advisable to consider interactions between the MRI findings (e.g., do patients with both meningeal enhancement AND tuberculomas have a different relationship with a clinical finding than patients who have meningeal enhancement WITHOUT tuberculomas). The small sample would clearly lack sufficient power to see any sort of interactions. I don’t think there is any way this can be fixed.

The same problem is inherent in the analysis of the effect of brain abnormalities on six month mortality and functional outcome. It isn’t correct to examine each one separately, for the above reasons. A model would need to contain a variable (yes/no) for EACH abnormality. One possible option would be to count the number of brain abnormalities any patient had, and check whether the number of abnormalities was associated with the different clinical outcomes. I’m not sure this would yield useful information, however.

It is not surprising that the authors found none of the MRI abnormalities (again, examined separately) were associated with mortality, disability, or worsening. There is too much overlap between any “one” abnormality and “all others,” since very few subjects HAD just one abnormality. Again, it requires a much more complex analysis than is possible given the small size of the data.

I feel that the only way these data could be presented would be descriptively, without an attempt to derive conclusions from very oversimplified (and incorrect) analyses.

Minor points:

The authors state that 60 patients were enrolled in the “parent” study (a Phase 2b dose-finding study). That is true, but if 12 of these did not have a baseline MRI, then they would never have been considered for the current study. A better way to present this would be to say that of 60 patients enrolled in a Phase 2b dose-finding clinical trial, 48 had baseline MRIs and were therefore eligible for the current study.

l.162 Multivariable, not multivariate

Survival analysis, not logistic regression, should be used to look at six-month mortality, but again, the fact that subjects could have multiple MRI abnormalities would have to be accounted for.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Bleibtreu Alexandre

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

- - - - -

For journal use only: PONEDEC3

28 Sep 2020

Authors’ Response to the Review Comments

Journal Plos One

Manuscript# PONE-D-19-33449R1

Title of Paper Brain MRI findings in relation to clinical characteristics and outcome of tuberculous meningitis

Authors Sofiati Dian, Robby Hermawan, Arjan van Laarhoven, Sofia Immaculata, Tri Hanggono Achmad, Rovina Ruslami, Farhan Anwary, Ristaniah D. Soetikno, Ahmad Rizal Ganiem, Reinout van Crevel

Response to Comments from Reviewer #1

Comments:

All concerns were addressed by the authors. The response to reviewers’ concerns is point by point argue and clear. I am agree with their decision to not improved the references.

For me the manuscript is suitable for publication.

We would like to thank Dr. Bleibtreu for the positive appraisal of our manuscript.

Response to Comments from Reviewer #2

Comment 1:

In this study, the authors used MRI data collected at baseline in 48 patients with tuberculous meningitis (TBM). They then attempted to test the association between MRI findings (which included meningeal enhancement, tuberculomas, brain infarctions, and hydrocephalus).and clinical characteristics (motor deficits, cranial nerve palsy, microbiological confirmation, and CSF abnormalities). This analysis, however is flawed because none of these comparisons are “clean.” This is due to the fact that so many of the patients had multiple MRI abnormalities. It doesn’t make sense to test each abnormality separately against the others and make conclusions about an individual abnormality given that there are a large number of possible permutations of abnormalities that could occur in any given patient. There is no way to be certain that any one abnormality is associated with any clinical characteristic, since the abnormalities so frequently occur in combination.

Response 1:

Thanks for including a separate statistical review, bringing up relevant point. Indeed many of the associations in these clinical studies are ‘unclean’, as pointed out, i.e. in previous report by Thwaites et.al, Kalita et.al, and Wasay et.al, many events occur together. Moreover, due to its relative infrequent nature, series of tuberculous meningitis patients are often small. Among all typical brain MRI findings, none has consistently been linked with a bad outcome. Therefore, to make it sure, we weighted the abnormalities based on the total number of it: ‘no abnormalities’, a single abnormality, or 2, 3 or 4 abnormalities (table 2), showing that the presence of 4 abnormalities is associated with more severe clinical presentation, bacteriological confirmation and CSF inflammation.

Comment 2:

The way to do this correctly would be to create a model in which the presence or absence of the clinical characteristic is the outcome, and a binary variable is created for each of the potential MRI abnormalities (allowing for one subject to have multiple abnormalities). It is doubtful that there are enough data to give sufficient power to such an analysis. Furthermore, it would also be advisable to consider interactions between the MRI findings (e.g., do patients with both meningeal enhancement AND tuberculomas have a different relationship with a clinical finding than patients who have meningeal enhancement WITHOUT tuberculomas). The small sample would clearly lack sufficient power to see any sort of interactions. I don’t think there is any way this can be fixed.

Response 2:

We agree with the reviewer that sample size is a concern (line 286). Still, this study represents serial MRI analysis of the 2nd largest cohort of well characterized and very high proportion of bacteriologically proven disease of TBM patients, a disease with a 1% incidence among human tuberculosis patients. In response to the reviewer’s comment, we now stress more clearly that the associative analysis of neuroradiology findings with the clinical and laboratory features is of an explorative nature (line 50, 89, 186, 227, 269). We also incorporated the suggestion of weighting the brain abnormality by the number of the finding (table 2).

Comment 3:

The same problem is inherent in the analysis of the effect of brain abnormalities on six-month mortality and functional outcome. It isn’t correct to examine each one separately, for the above reasons. A model would need to contain a variable (yes/no) for EACH abnormality. One possible option would be to count the number of brain abnormalities any patient had, and check whether the number of abnormalities was associated with the different clinical outcomes. I’m not sure this would yield useful information, however.

Response 3:

Regarding this study more as descriptive and exploratory, a complete statistical model – also given the study size – might not be appropriate. However, following the reviewer’s suggestion we have examined a possible association between the number of MRI abnormalities, paradoxical response and patient outcome (table 2 and table 4). This shows a comparable number of patients who died or had a good recovery after 6-month of treatment among groups with different number of brain abnormalities. However, good recovery after 6-month of treatment seems more common in group of patients with only radiological or had no paradoxical response.

Comment 4:

It is not surprising that the authors found none of the MRI abnormalities (again, examined separately) were associated with mortality, disability, or worsening. There is too much overlap between any “one” abnormality and “all others,” since very few subjects HAD just one abnormality. Again, it requires a much more complex analysis than is possible given the small size of the data.

I feel that the only way these data could be presented would be descriptively, without an attempt to derive conclusions from very oversimplified (and incorrect) analyses.

Response 4:

Following the reviewer’s suggestion, we now present the data in a more descriptive way. We have rephrased several statements to make it less strong in the abstract (lines 50), introduction (line 86), results (lines 186) and discussion (lines 227, 269):

Minor points:

The authors state that 60 patients were enrolled in the “parent” study (a Phase 2b dose-finding study). That is true, but if 12 of these did not have a baseline MRI, then they would never have been considered for the current study. A better way to present this would be to say that of 60 patients enrolled in a Phase 2b dose-finding clinical trial, 48 had baseline MRIs and were therefore eligible for the current study.

Response minor point:

Thanks for this suggestion, we have rephrased this sentence (line 167-171).

Comment 5

l.162 Multivariable, not multivariate

Response 5:

Thank you for pointing out this mistake; we deleted this sentence altogether because we changed the analysis.

Comment 6:

Survival analysis, not logistic regression, should be used to look at six-month mortality, but again, the fact that subjects could have multiple MRI abnormalities would have to be accounted for.

Response 6:

We thank the reviewer for pointing this out. Indeed, it will be less accurate to predict the mortality by cox or logistic model using this single brain lesion due to the ‘unclean’ variable of each brain abnormality as pointed out by the reviewer. As suggested and a previously applied (Kalita, et.al) we adopted to a descriptive approach (table 2). in a larger study, this approach might have been used (for instance as done by Wasay et.al(3) in their study), but we have now used a more descriptive approach as Kalita, et.al(2) have done (table 2).

We again thank the editor and reviewers for their time and efforts to help improve our manuscript which we hope is now suitable for publication.

Reference

1. Thwaites GE, Macmullen-Price J, Tran TH, Pham PM, Nguyen TD, Simmons CP, et al. Serial MRI to determine the effect of dexamethasone on the cerebral pathology of tuberculous meningitis: an observational study. Lancet Neurol. 2007;6(3):230-6.

2. Kalita J, Prasad S, Misra UK. Predictors of paradoxical tuberculoma in tuberculous meningitis. Int J Tuberc Lung Dis. 2014;18(4):486-91.

3. Wasay M, Farooq S, Khowaja ZA, Bawa ZA, Ali SM, Awan S, et al. Cerebral infarction and tuberculoma in central nervous system tuberculosis: frequency and prognostic implications. J Neurol Neurosurg Psychiatry. 2014;85(11):1260-4.

4. Anuradha HK, Garg RK, Agarwal A, Sinha MK, Verma R, Singh MK, et al. Predictors of stroke in patients of tuberculous meningitis and its effect on the outcome. QJM. 2010;103(9):671-8.

5. Singh AK, Malhotra HS, Garg RK, Jain A, Kumar N, Kohli N, et al. Paradoxical reaction in tuberculous meningitis: presentation, predictors and impact on prognosis. BMC Infect Dis. 2016;16(1):306.

Submitted filename: Response to the Reviewers_Appeal.docx

Click here for additional data file.

26 Oct 2020

Brain MRI findings in relation to clinical characteristics and outcome of tuberculous meningitis

PONE-D-19-33449R2

Dear Dr. Dian,

First, I would like to thank you for your enormous patience during this entire process!

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Niels Bergsland

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

4 Nov 2020

PONE-D-19-33449R2

Brain MRI findings in relation to clinical characteristics and outcome of tuberculous meningitis

Dear Dr. Dian:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Niels Bergsland

Academic Editor

PLOS ONE