The burden of malaria parasites in an infected human is an important measure of significant interest for biologic investigation. It is typically characterised by Area Under the malaria parasite growth Curve (AUC). Here we report a study to identify the optimal mathematical method to calculate AUC measuring P. falciparum burden.
Parasitemia counts from 116 individuals inoculated with P. falciparum in the Induced Blood Stage Malaria (IBSM) model [1] were used to characterise parasitemia burden prior to antimalarial treatment. AUC was defined from qPCR patency to treatment (4-7 days post inoculation). Two mathematical approaches were used to calculate AUC: trapezoidal method, or integrating under the log-linear parasite growth curve. Different substitutions were trialled for ND points prior to patency, namely, ignoring the point or substituting 0 or half the limit of detection (32 parasites/mL). Intra-class correlations (ICC) were calculated to investigate consistencies between methods. Associations to other potential measures of burden were explored.
All methods differed numerically (logically, paired t-tests). ICC values were high (ICC>0.8) indicating that each method measures parasite burden similarly. Correlations were weak between parasite growth rate and AUC (r<0.35), and reasonable with peak parasitemia (0.7<r<0.9), suggesting that AUC is a separate measure of infection to the other widely used measure, growth rate, but more similar to peak parasitemia.
Our analysis indicates that either integrating the log-linear growth curve or the trapezoidal method with zero substituted for leading ND are acceptable methods for calculating AUC. Parasite growth rate, if available, can be integrated to derive AUC. The high ICCs imply this AUC is measuring a similar quantity to the trapezoidal measures, making it a promising choice for the measure of parasite burden. Future work would include investigating associations of AUC methods to immunological markers and simulation studies to discern an ‘ideal’ AUC measure.