Table 1.

Animal and human studies investigating the effects of anti-platelet agents in tuberculosis.

Reference	Anti-platelet agent(s) tested	Host species/patients studied	Intervention	Summary of findings
Animal studies
Shroff et al, 1990 (106)	Indomethacin	Swiss white mice	Six intraperitoneal injections of indomethacin, 50µg per mouse, given at 12 hour intervals; immunization with intraperitoneal M. vaccae 12 hours after the last dose	Indomethacin-treated mice showed a delayed type hypersensitivity response, whereas non-treated mice did not.
Hernández-Pando et al, 1995 (107)	Indomethacin	Male 6-8 week old BALB/c mice immunized with culture filtrate proteins (CFP) from H37Rv M.tb delivered via endotracheal route	20mg/kg cyclophosphamide or 5mg/kg indomethacin given by intraperitoneal injection one day prior to endotracheal challenge with CFP	Induction of granulomas with sepharose beads coated with CFP reduced antibody production and delayed-type hypersensitivity responses to mycobacterial antigens; this was reversed with the administration of cyclophosphamide and indomethacin.
Rangel Moreno et al, 2002 (108)	Niflumic acid	Male 6-8 week old BALB/c mice infected with H37Rv M.tb via endotracheal route	500µg/ml niflumic acid given by intragastric cannula twice per day	Mice treated with niflumic acid early in infection produced greater expression of TNF-α, IL-1α, and IFN-γ, with a reduction in inhaled nitric oxide synthase (iNOS) expression and an increased bacterial load. Mice treated after 60 days of infection showed increased pro-inflammatory cytokine concentrations, a striking increment of iNOS expression, and reduced bacillary load.
Dutta et al, 2004 (109)	Diclofenac sodium	45 mycobacterial strains including drug-sensitive and -resistant strains in Kirchner’s liquid culture medium; male Swiss Albino mice infected via intraperitoneal route with M.tb H37Rv102	Mycobacteria were tested with increasing concentrations of diclofenac sodium. Infected mice were injected with 10mg/kg diclofenac sodium daily for 6 weeks	Diclofenac sodium exhibited bactericidal activity in most mycobacterial strains at a range of 15-25µg/ml. Diclofenac sodium treatment of infected mice resulted in reduced macroscopic lesions and a reduction in mycobacterial load.
Hernández-Pando et al, 2006 (110)	Niflumic acid (a COX-2 inhibitor) and soluble betaglycan (an anti-TGF-β agent)	Male 6-8 week old BALB/c mice infected with H37Rv M.tb	30µg soluble betaglycan administered twice a week by intraperitoneal route from 30 days post infection, with/without 500µg niflumic acid administered twice a day by intragastric cannulation, or pan-specific TGF-β antibodies. Mice euthanased at 15, 30, 45, and 60 days	Compared to control animals, treatment produced higher expression of TNF‐α, IFN‐γ, IL‐2, iNOS and lower expression of IL‐4, with decreased lung fibrosis and bacillary load; there was a higher inflammatory response manifested as a greater area of lung affected by pneumonia. The addition of niflumic acid was synergistic, with reduced bacillary load and increased TNF‐α.
Byrne et al, 2007 (111)	Aspirin and ibuprofen	BALB/c mice infected with M.tb H37Rv by aerosol administration	Oral treatment of aspirin (10, 20, 40 mg/kg) and ibuprofen (10, 20, 40 mg/kg) +/- pyrazinamide (150 mg/kg) daily for 1 month	Aspirin and ibuprofen had no effect alone, but increased bactericidal activity of pyrazinamide when given in combination.
Byrne et al, 2007 (112)	Aspirin and ibuprofen	Four week old BALB/c mice infected with M.tb H37Rv by aerosol administration	Aspirin or ibuprofen (10-40mg/kg) with or without isoniazid (25mg/kg) administered by oral gavage five times per week for one month. Mice were euthanized one day after treatment completion	Co-administration of aspirin and isoniazid antagonized the antimycobacterial activity of isoniazid. Ibuprofen had no interaction with isoniazid.
Dutta et al, 2007 (113)	Diclofenac	Male Swiss Albino mice infected via intraperitoneal route with M.tb H37Rv	Infected mice treated with diclofenac 10µg/g or streptomycin 150µg/g, alone or in combination, daily for 4 weeks	Diclofenac and streptomycin both resulted in improved survival, reduced bacillary count, and reduced splenic weight. The two drugs acted synergistically
Peres-Buzalaf et al, 2011 (114)	Celecoxib and MK-886 (5-LO activation protein inhibitor)	Male 5-8 week old BALB/c mice infected intratracheally with M.tb H37Rv strain	Oral treatment with celecoxib 5mg/kg/0.5ml, and/or MK-886 5mg/kg/0.5ml, 1h prior to infection with M.tb and again every 24h for 60 days	Celecoxib enhanced 60-day survival from 86% to 100% and reduced lung bacterial burden; conversely, MK-886 reduced 60-day survival from 86% to 43% and increased lung bacterial load.
Vilaplana et al, 2013 (115)	Ibuprofen	C3HeB/FeJ pathogen-free mice infected intravenously with M.tb H37Rv	Ibuprofen 80mg/kg given orally 3 or 4 weeks post infection.	Ibuprofen treated animals had a reduction in size and number of lung lesions, bacillary load, and improved survival.
Marzo et al, 2014 (116)	Aspirin and ibuprofen	C3HeB/FeJ and C3H/HeN pathogen-free mice infected intravenously with M.tb H37Rv	Aspirin 3mg/kg, sodium heparin 20 UI/kg, ibuprofen 80mg/kg orally	Aspirin administration reduced bacterial load, attenuated the severity of histopathological changes, and improved survival. Aspirin and ibuprofen-treated C3HeB/FeJ and C3H/HeN mice had lower levels of pro-inflammatory mediators including TNF-a, IL-6, IL-17 and CXCL5 than untreated C3HeB/FeJ mice.
Kroesen et al, 2018 (104)	Aspirin	C3HeB/FeJ infected with M.tb H37Rv Pasteur strain	Aspirin 3mg/kg given alone or in combination with anti-TB treatment	Low-dose aspirin reduced bacterial load and increased survival, reduced lung pathology, decreased production of pro-inflammatory cytokines and delayed neutrophil recruitment and T cell responses. In combination with anti-TB treatment aspirin enhanced survival and reduced lung pathology.
Hortle et al, 2019 (117)	Aspirin, and tirofiban and eptifibatide (specific inhibitors of glycoprotein IIb/IIIa–fibrinogen binding)	Zebrafish infected with M. marinum	Infected zebrafish treated with 10µg/ml aspirin, 20µg/ml tirofiban, 10uM eptifibatide, or 5uM warfarin, or control	Mycobacterial burden correlated with thrombocyte density. Aspirin, tirofiban, and eptifibatide reduced mycobacterial burden by inhibiting thrombocyte-granuloma interactions.
Mortensen et al, 2019 (118)	Celecoxib and ibuprofen	CB6F1 mice infected with M.tb Erdman strain via aerosol or intravenous (IV) routes	Celecoxib or ibuprofen administered orally in mouse feed 4-12 weeks post infection	Aerosol-infected mice treated with celecoxib or ibuprofen had higher bacillary burdens in lung and spleen and there was no difference in pulmonary immune cell infiltration or cytokine levels measured in plasma or lung homogenates. Celecoxib led to increased bacterial burden and altered function/differentiation of Type 1 helper T cells in mice re-infected following antibiotic treatment. IV-infected mice had reduced inflammation and bacterial burden following ibuprofen treatment.
Human Studies
Misra et al, 2010 (119)	Aspirin	Patients with TB meningitis diagnosed based on clinical, magnetic resonance imaging (MRI), and cerebrospinal fluid (CSF) criteria	118 patients randomized to receive 150mg aspirin per day or placebo. All patients received standard treatment with rifampicin, isoniazid, pyrazinamide and ethambutol with/without corticosteroids	Aspirin resulted in insignificantly fewer strokes (24.2% versus 43.3%, OR 0.42, 95%CI 0.121-1.39) and significantly reduced mortality (21.7% versus 43.4%, p=0.02) compared to placebo. Aspirin was well tolerated.
Schoeman et al, 2011 (120)	Aspirin	Children with TB meningitis	146 children randomized to placebo, low-dose aspirin (75mg/day), and high-dose (100mg/kg/day) aspirin, alongside standard treatment plus prednisolone for the first month of treatment	There was no benefit in morbidity or mortality.
Mai et al, 2018 (121)	Aspirin	HIV-uninfected adults with TB meningitis	120 patients randomized to aspirin (81mg/day or 1000mg/day) or placebo added to the first 60 days of anti-tuberculosis drugs plus dexamethasone	There was no difference in rates of gastro-intestinal or cerebral bleeding, new brain infarction, or death by 60 days. There was a reduction in new infarcts and deaths by day 60 in aspirin-treated participants with microbiologically confirmed TBM compared to placebo.
Lee et al, 2019 (122)	Antiplatelet agents including aspirin	Retrospective cohort study of incident TB cases in the Taiwan National Tuberculosis Registry between 2008 and 2014	9,497 antiplatelet users (including 7,764 aspirin users) compared to 8,864 non-users	After 1:1 propensity score matching, antiplatelet use was associated with longer survival (adjusted hazard ratio (HR): 0.91, 95% confidence interval (CI): 0.88-0.95, p < 0.0001) but no increase in major bleeding risk (p=0.604).
Wang et al, 2020 (123)	Aspirin	Patients with pulmonary TB and type 2 diabetes mellitus	348 patients randomized to aspirin (100mg/day) or placebo. 168 patients completed the trial and were included in analysis	Erythrocyte sedimentation rate and C-reactive protein levels were lower (p=0.000), the sputum-negative conversion rate was higher (86.7% versus 53.8%, p=0.031), and the number and size of cavities was smaller in the aspirin-treated than the placebo-treated group.