In attempts to improve outcome in pediatric bacterial meningitis (BM), several adjuvant therapies have been tried. In Blantyre, Malawi dexamethasone showed no benefit in children or adults.1,2 In South America, glycerol as adjuvant therapy for BM in children significantly reduced the incidence of unsatisfactory outcomes (severe neurological sequelae or death).3 Acetaminophen has been shown to improve survival in all-cause septicemia in adults.4
Our present study posed 2 questions: (1) Can adjuvant oral glycerol, for the first 48 hours of treatment, improve the outcome of childhood BM in a population with a high prevalence of HIV infection? and (2) Can the outcome be further improved by giving rectal acetaminophen for the first 48 hours of treatment?
Endpoints were death and severe neurological sequelae (including severe sensorineural hearing loss) at hospital discharge and 6 months later. Severity of illness, age, etiological agent, hemoglobin concentration, HIV status and presence of malaria coinfection were covariates. Secondary end points were less severe audiological or neurological sequelae.
In this double-blind, randomized, placebo-controlled study, all consented children aged ≥ 2 months admitted to Queen Elizabeth Central Hospital, Blantyre, Malawi, with possible or confirmed BM were enrolled. All children received intravenous ceftriaxone and were randomized to receive oral glycerol or a placebo (carboxymethycellulose) and either acetaminophen or a dummy (cocoa butter base) suppository.
Exclusion criteria were: trauma, previous neurological disease and previous permanent nonconductive hearing loss. Children who had received an oral antibiotic or only 1 parenteral dose were included.
BM was defined as: (1) a positive CSF bacterial culture, (2) characteristic CSF findings (see below) with a positive blood culture or (3) symptoms and signs compatible with BM with CSF containing ≥100 polymorphonuclear leucocytes/mm3. Cerebrospinal fluid (CSF) specimens were cultured according to standard laboratory methods. Two HIV antibody-based rapid diagnostic tests were done and discordant results were confirmed by a third test. Children <15 months of age with a positive HIV antibody test had HIV infection confirmed by polymerase chain reaction. Randomization was computer generated in permuted blocks of 12. Only cases with BM according to the above criteria were included in the analysis of outcomes.
All children received intravenous ceftriaxone 100 mg/kg/d for 5 days, prolonged at the clinicians’ discretion. Patients with salmonella meningitis were treated with 14 days of ceftriaxone followed by 14 days of oral ciprofloxacin. Randomization was to 1 of 4 groups according to adjunctive therapy: Group A oral and rectal placebos; Group B oral glycerol + rectal placebo; Group C oral placebo + rectal acetaminophen and Group D oral glycerol + rectal acetaminophen. Oral 85% glycerol or placebo was given as 6.0 g/kg/d (6.0 mL) for 2 days divided in 4 equal daily doses (maximum = 25 mL per dose). If the patient was not able to swallow or was vomiting (rare), a nasogastric tube was inserted. Acetaminophen was given as a suppository; the first dose was 35 mg/kg, followed by 20 mg/kg 6 hourly for 42 hours. Placebo was given in the same manner in identical volumes. The first dose of supplementary therapy preceded the first dose of antibiotic.
Supportive treatment was according to departmental protocols. HIV testing with pre- and postcounseling was done for all patients.
Age appropriate behavioral testing (distraction test, performance test, pure tone audiometry) was used to define hearing threshold levels across the speech frequency range (500 Hz to 4 KHz).
Outcome was categorized as: (1) survival to 6 months post discharge with no sequelae, (2) survival with sequelae that affect daily life (eg, hemiplegia, deafness, blindness, seizures, global developmental delay), (3) severe sequelae which incapacitate the child and (4) death.
All children were assessed for hearing, visual, developmental and neurological outcomes at hospital discharge and 30 and 180 days after discharge. Patients who failed to attend were traced and encouraged to return for follow up. Those who were not found were declared lost to follow up.
Sample Size and Data Analysis
For a 15% decrease in the incidence of outcomes 2–4 between the glycerol and placebo groups, with a 5% error (P < 0.05) and a power of 80%, 138 patients were required in each of the 2 main arms. To detect a further 15% difference between the acetaminophen and the placebo subgroups, we aimed to enroll 552 in all. A blinded analysis after 448 enrollments showed that 360 cases fulfilled all enrollment criteria and that further enrolments would not alter the outcomes.
The study was approved by the College of Medicine Research and Ethics Committee and registered with Clinicaltrials.gov NCT 00619203. Independent safety monitors analyzed data after 100, 200 and 300 enrolments.
The enrollment sequence is shown in Figure 1. There were 360 cases allocated to the 4 study arms. In 190 cases (53%), the causative agent was Streptococcus pneumoniae, in 30 (9%) were Haemophilus influenzae, in 25 (6%) were other bacteria and in 115 (32%) no bacteria were cultured. Forty-six percentage of children (n = 165) had received antibiotics before admission; 162 (47%) had seizures before or at the time of admission and 109 (30%) had a Blantyre Coma Score of ≤ 2 (equivalent to a Glasgow coma score of ≤8). The median hematocrit was 31% (range 2–60%). The baseline data for the 4 treatment groups were similar except that more children had received prior antibiotics in Group C; 64% in Group C versus 34%, 38% and 37% in Groups A, B and D, respectively; P = 0.001 and children in Group A had a higher mean hematocrit than those in other groups.
The outcomes of the 4 groups were similar. Percentages for outcomes 1, 2, 3 and 4 were in Group A 41%, 29%, 7% and 23%; Group B 34%, 28%, 13% and 24%; Group C 38%, 25%, 10% and 26%; Group D 41%, 15%, 14% and 29%. One-third (36%, n = 123) of the children were HIV infected, and adjuvant therapy did not affect their outcome although HIV patients may have been more susceptible to sequelae (65% of HIV-positive vs. 56% of HIV-negative patients P = 0.054). The results did not change when the 2 groups receiving glycerol were compared with the 2 groups receiving acetaminophen (P = 0.303 vs. P = 0.311).
Nearly one-third (29%) of survivors had hearing sequelae of whom 14% (n = 51) had bilateral hearing loss that was unaffected by adjuvant therapy. Overall, 15% (n = 53) of children were left with neurological sequelae. The study was not powered to look at outcome by causative agent but S. pneumoniae was the most common pathogen isolated (n = 190 of 243; 78%). In this subgroup, mortality was unaffected by adjuvant therapy. S. pneumoniae meningitis had a 5-fold higher case fatality rate (55%) than Hib meningitis (12%). Compared with all other meningitides combined (Hib 12%, meningococcus 0%, no growth 24%, Salmonella spp 5% and other 2%), the mortality from S. pneumoniae was significantly worse (P = 0.034).
The results of our study differ from those of a large, multicenter, multicountry study in South America in which adjuvant glycerol significantly reduced the number of severe neurological sequelae [OR: 0.39 (95% confidence interval: 0.17–0.93); P = 0.033].4 There are differences between the studies; the present study was smaller (360 vs. 654 cases) and the most common causative agent identified in Blantyre was S. pneumoniae (53% of all cases, 78% of isolates; n = 190), whereas in South America it was H. influenzae type b (33% of all cases; n=221). In South America, glycerol was most beneficial in patients with Hib meningitis [OR: 0.11 (95% confidence interval: 0.01–0.95); P = 0.045].
Adjuvant glycerol in adult Malawians with BM was associated with a higher case fatality [by day 40 (OR: 2.4, 95% confidence interval: 1.3–4.2); P = 0.003),5 in a recent Cochrane review no benefit or harm was reported on review of 4 trials with a total of 1091 patients with BM.6 Acetaminophen had no effect on outcome both with and without glycerol.
We did not enroll the expected number of patients because the incidence of BM has decreased since the introduction of Hib conjugate vaccine in 20027,8 and pneumococcal infections have been fewer over the last decade even before the introduction of the conjugate vaccine in 2011.9
Glycerol and acetaminophen as adjuvant therapies were of no benefit to children with bacterial meningitis in a setting with high HIV prevalence and in whom pneumococcus was the most common causative agent.
The authors thank the staff of the pediatric department and the research ward for their help in caring for the children. All the laboratory investigations were undertaken in The Malawi Liverpool Wellcome Trust Research Unit. Dr. M. Tidd prepared all the study medications for which we are very grateful. We especially thank the children and parents in our care for participating in the study.
1. Molyneux EM, Walsh AL, Forsyth H, et al. Dexamethasone treatment in childhood bacterial meningitis in Malawi: a randomised controlled trial. Lancet. 2002;360:211–218
2. Scarborough M, Gordon SB, Whitty CJ, et al. Corticosteroids for bacterial meningitis in adults in sub-Saharan Africa. N Engl J Med. 2007;357:2441–2450
3. Peltola H, Roine I, Fernández J, et al. Adjuvant glycerol and/or dexamethasone to improve the outcomes of childhood bacterial meningitis: a prospective, randomized, double-blind, placebo-controlled trial. Clin Infect Dis. 2007;45:1277–1286
4. Kuikka A. Factors associated with the outcome of bacteraemia during twenty years (1976–1996) with special reference to bacteraemias due to Streptococcus pneumoniae, Staphylococcus aureus, Escherichia coli
, and Pseudomonas aeruginosa
. 1999 Academic dissertation, University of Helsinki, Helsinki:83
5. Ajdukiewicz KM, Cartwright KE, Scarborough M, et al. Glycerol adjuvant therapy in adults with bacterial meningitis in a high HIV seroprevalence setting in Malawi: a double-blind, randomised controlled trial. Lancet Infect Dis. 2011;11:293–300
6. Wall ECB, Ajdukeiwics KMB, Heyderman R, et al. Osmotic therapies added to antibiotics for acute bacterial meningitis. Cochrane review. Cochrane Database Syst Rev. 2013;3:CD008806
7. Daza P, Banda R, Misoya K, et al. The impact of routine infant immunization with Haemophilus influenzae
type b conjugate vaccine in Malawi, a country with high human immunodeficiency virus prevalence. Vaccine. 2006;24:6232–6239
8. McCormick DW, Molyneux EM. Bacterial meningitis and Haemophilus influenzae
type b conjugate vaccine, Malawi. Emerg Infect Dis. 2011;17:688–690
9. Everett DB, Mukaka M, Denis B, et al. Ten years of surveillance for invasive Streptococcus pneumoniae
during the era of antiretroviral scale-up and cotrimoxazole prophylaxis in Malawi. PLoS One. 2011;6:e17765