ReportInfectious Disease

A meta-analysis of clinical studies conducted during the West Africa Ebola virus disease outbreak confirms the need for randomized control groups

See allHide authors and affiliations

Science Translational Medicine  27 Nov 2019:
Vol. 11, Issue 520, eaaw1049
DOI: 10.1126/scitranslmed.aaw1049
  • Fig. 1 Evaluation of mortality rates and cycle thresholds by enrollment date for control groups from the six clinical studies.

    These six control groups were included as comparator arms for the following six clinical studies: treatment with the antiviral drug favipiravir (Favi-Bai, Favi-JIKI), treatment with the antimalarial drug amodiaquine, administration of convalescent whole blood from Ebola virus disease survivors (ConvBlood), administration of convalescent plasma from Ebola virus disease survivors (ConvPlasma), and treatment with the triple monoclonal antibody therapy ZMapp (Prevail-ZMapp). (A) Mortality rates by dates during which enrollment was open for each study. (B) Mean cycle thresholds (a proxy for viral load) for each study according to enrollment dates.

  • Fig. 2 Comparison of the control groups from the six clinical studies.

    (A) Raw mortality rates from the control groups of the six clinical studies. These studies included treatment with the antiviral drug favipiravir (Favi-Bai, Favi-JIKI), treatment with the antimalarial drug amodiaquine, administration of convalescent whole blood from Ebola virus disease survivors (ConvBlood), administration of convalescent plasma from Ebola virus disease survivors (ConvPlasma), and treatment with the triple monoclonal antibody therapy ZMapp (Prevail-ZMapp). (B) Expected mortality as a function of cycle threshold (a measure of viral load) for a woman aged 34 across the six studies. (C) A Galbraith plot confirmed heterogeneity between studies. (D) Predicted mortality rates for covariate values corresponding to 53% females, with an average age of 30.1 years, and an average cycle threshold of 25.6.

  • Fig. 3 Standardized odds ratios for treatment effects for four investigational drugs.

    Standardized odds ratios for treatment effects are shown for the experimental arms relative to each control group of the clinical studies investigating (A) the antimalarial drug amodiaquine, (B) the siRNA drug TKM-130803, (C) the triple monoclonal antibody therapy ZMapp, and (D) IFN-β1a. For each standardized odds ratio, a logistic regression model was fit using data from the control group listed in the left column and the experimental arm indicated above the plot. The odds ratios for the models are reported in table S3.

  • Fig. 4 Standardized odds ratios for treatment effects for four investigational therapies.

    Standardized odds ratios for treatment effects are shown for the experimental arms relative to each control group of the studies investigating: (A) the antiviral drug favipiravir (Favi-JIKI), (B) administration of convalescent plasma from Ebola virus disease survivors (ConvPlasma), (C) the antiviral drug favipiravir (Favi-Bai), and (D) administration of convalescent whole blood from Ebola virus disease survivors (ConvBlood). For each standardized odds ratio, a logistic regression model was fit using data from the control group listed in the left column and the experimental arm indicated above the plot. The odds ratios for the models are reported in table S4.

  • Table 1 Demographic summary of the six clinical studies included in the meta-analysis by treatment group.

    Trt, treatment; IVF, intravenous fluids; NA, not applicable.

    Meta-analysis dataset*Published study results
    StudyArm (N)Enrollment
    dates
    Age Mean (SD)FemalesCT Mean (SD)MortalityNMortalityIVF use
    Prevail II–ZMapp
    (16)
    Control (29)3/1/2015–
    11/1/2015
    33 (13.1)41.4%22.9 (3.7)31.0%3537%63%
    Trt (31)3/1/2015–
    11/1/2015
    27.6 (17.5)58.1%24.1 (5.2)19.4%3622%61%
    Amodiaquine
    (17)
    Control (169)6/5/2014–
    10/24/2014
    31.1 (14.8)48.5%20.2 (4.1)62.1%19464.4%32%
    Trt (58)8/18/2014–
    8/30/2014
    29.5 (18.3)46.6%19.9 (4.2)51.7%7150.7%35.2%
    ConvBlood
    (18)
    Control (11)12/1/2014–
    4/30/2015
    35.8 (13.0)36.4%27.7 (7.2)54.5%2544%NA
    Trt (20)12/1/2014–
    4/30/2015
    26.2 (11.5)60%25.4 (5.0)20%4328%100%
    ConvPlasma
    (19)
    Control (382)9/1/2014–
    1/1/2015
    33.2 (15.8)49.5%26.4 (4.1)36.9%41838%NA
    Trt (78)1/1/2015–
    7/7/2015
    31.6 (14.8)56.4%27.9 (4.1)32.1%8431%NA
    Favi-Bai (20)Control (78)10/10/2014–
    10/30/2014
    29.6 (15.0)53.8%26.0 (4.6)65.4%8564.7%0%
    Trt (38)11/1/2014–
    11/10/2014
    31.7 (17.8)47.4%27.3 (5.4)42.1%3943.6%0%
    Favi-JIKI (21)Control (478)9/15/2014–
    12/15/2014
    36.0 (17.2)50.6%20.8 (4.2)56.9%54058%NA
    Trt (99)12/17/2014–
    4/8/2015
    37.6 (16.7)63.6%21.0 (4.3)51.5%9951.5%92%
    IFN-β1a (22)Control (NA)3/26/2015–
    6/12/2015
    NANANANA2181.0%NA
    Trt (9)3/26/2015–
    6/12/2015
    33.9 (14.6)55.6%23.5 (4.4)33.3%933.3%100%
    TKM-130803§
    (23)
    Control (NA)NANANANANA182055%NA
    Trt (13)3/11/2015–
    6/15/2015
    41.7 (18.1)38.5%22.8 (3.9)76.9%1275%100%

    *Patients below 6 years of age were excluded from this meta-analysis. Three studies had subjects with missing cycle threshold (CT) values: amodiaquine (n = 9), Favi-Bai (n = 1), and TKM-130803 (n = 1).

    †In the Favi-JIKI study, the difference in the proportion of females was statistically significant at P < 0.05. In the ConvPlasma study, the mean cycle thresholds were different at P < 0.05.

    ‡In the published manuscript for the Favi-Bai study, primary analyses were reported excluding patients who were transferred out of the Ebola treatment unit, leading to 18 controls and 17 cases, with a 72 and 35% mortality rate (P = 0.044), respectively. The data provided for the meta-analysis included all patients, including those transferred out of the Ebola treatment unit.

    §In the published paper for the TKM-130803 trial, the primary analyses reported mortality eliminated deaths within the first 48 hours. However, in the meta-analysis, all deaths were included to parallel the other studies.

    • Table 2 Expected study-specific mortality risk for control patients according to cycle threshold (CT) values, adjusting for sex and age based on a logistic regression model for a 34-year-old female.

      Prevail II–ZMapp (16)Amodiaquine (17)ConvWBlood (18)ConvPlasma (19)Favi-Bai (20)Favi-JIKI (21)
      CT 200.47 (0.23, 0.72)0.55 (0.37, 0.71)0.83 (0.31, 0.98)0.61 (0.49, 0.71)0.74 (0.54, 0.87)0.57 (0.5, 0.64)
      CT 250.06 (0.01, 0.28)0.15 (0.04, 0.45)0.55 (0.22, 0.84)0.35 (0.29, 0.42)0.66 (0.54, 0.76)0.18 (0.13, 0.24)
      CT 300.01 (0, 0.17)0.04 (0, 0.29)0.29 (0.06, 0.7)0.19 (0.14, 0.25)0.59 (0.43, 0.73)0.05 (0.03, 0.08)

    Supplementary Materials

    • stm.sciencemag.org/cgi/content/full/11/520/eaaw1049/DC1

      Materials and Methods

      Fig. S1. PRISMA 2009 flow diagram.

      Fig. S2. Adjusted odds ratios for amodiaquine, TKM-130803, Prevail II–ZMapp, and IFN-β1a study treatment arms for patients with a cycle threshold of ≥20 relative to each control group.

      Fig. S3. Adjusted odds ratios for Favi-JIKI, convalescent plasma, Favi-Bai, and convalescent whole-blood study treatment arms for patients with a cycle threshold of ≥20 relative to each control group.

      Fig. S4. Adjusted odds ratios for treatment relative to each of six control groups for patients of the TKM-130803 study, from the per-protocol and intent-to-treat samples.

      Table S1. Weighted least squares regression for models in Fig. 1, evaluating association of midpoint of study enrollment dates and mortality rate/cycle threshold.

      Table S2. Model estimates from logistic regression models for Fig. 2D.

      Table S3. Model estimates from logistic regression models for Fig. 3 (A to D).

      Table S4. Model estimates from logistic regression models for Fig. 4 (A to D).

      Table S5. PRISMA 2009 checklist.

    • This PDF file includes:

      • Materials and Methods
      • Fig. S1. PRISMA 2009 flow diagram.
      • Fig. S2. Adjusted odds ratios for amodiaquine, TKM-130803, Prevail II–ZMapp, and IFN-β1a study treatment arms for patients with a cycle threshold of ≥20 relative to each control group.
      • Fig. S3. Adjusted odds ratios for Favi-JIKI, convalescent plasma, Favi-Bai, and convalescent whole-blood study treatment arms for patients with a cycle threshold of ≥20 relative to each control group.
      • Fig. S4. Adjusted odds ratios for treatment relative to each of six control groups for patients of the TKM-130803 study, from the per-protocol and intent-to-treat samples.
      • Table S1. Weighted least squares regression for models in Fig. 1, evaluating association of midpoint of study enrollment dates and mortality rate/cycle threshold.
      • Table S2. Model estimates from logistic regression models for Fig. 2D.
      • Table S3. Model estimates from logistic regression models for Fig. 3 (A to D).
      • Table S4. Model estimates from logistic regression models for Fig. 4 (A to D).
      • Table S5. PRISMA 2009 checklist.

      [Download PDF]

    Stay Connected to Science Translational Medicine

    Navigate This Article