31.08.2013 Views

Experimental infection and protection against ... - TI Pharma

Experimental infection and protection against ... - TI Pharma

Experimental infection and protection against ... - TI Pharma

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Efficacy of pre-erythrocytic <strong>and</strong> blood-stage malaria vaccines can be assessed in small<br />

sporozoite challenge trials in human volunteers<br />

calculated the geometric mean multiplication rate <strong>and</strong> SD. Samples with a Q-PCR<br />

CT >45 were estimated at half the detection limit (10 or 50 Pf/ml).<br />

Geometric mean parasitemia in the first cycle <strong>and</strong> parasite multiplication rates<br />

showed a normal distribution, so power calculations were performed using a<br />

two-sample t test power analysis with α=0.05 <strong>and</strong> β=0.80. We separately<br />

determined the power using all three daily time points if available (8:00am,<br />

4:00pm <strong>and</strong> 10:00pm, n=33), two daily time points (8:00am <strong>and</strong> 4:00pm, n=48)<br />

or one daily time point (8:00am, n=48).<br />

Power calculations estimated the group size required to detect significant<br />

differences between vaccinees <strong>and</strong> controls. Vaccine <strong>and</strong> control group were<br />

assumed of equal size. Pre-erythrocytic vaccines were assumed to reduce the<br />

geometric mean parasitemia of the first multiplication cycle, but not alter the<br />

parasite multiplication rate. Erythrocytic stage vaccines were assumed to reduce<br />

the parasite blood-stage multiplication rate, reducing parasite load from the first<br />

cycle onwards. Vaccine effects were simulated by beta distribution with an<br />

assumed mean inhibition of 70, 80, 90 <strong>and</strong> 95% <strong>and</strong> SD of 5 or 10%. Vaccine<br />

inhibition was simulated 100 times for every individual subject <strong>and</strong> repeated ten<br />

times based on r<strong>and</strong>om sampling from the distribution. We assumed a<br />

continuous infectious dose of merozoites, as opposed to the actual release of<br />

discrete batches of merozoites from infected hepatocytes. Interindividual<br />

variation masked the discrete nature of <strong>infection</strong> in empirical situations with<br />

high numbers of infected hepatocytes. In simulations with low numbers of<br />

infected hepatocytes, where a discrete variable would give a binary outcome<br />

(i.e. protected or non-protected), the resulting blood-stage parasitemia was<br />

always below the Q-PCR detection limit. In such cases, values were corrected to<br />

a st<strong>and</strong>ard value of half the Q-PCR detection limit. The number of subjects with<br />

simulated parasitemia ≤ 1Pf/ml or multiplication rate ≤1 was counted. For some<br />

subjects second cycle data were missing because they had already received<br />

antimalarial treatment for safety reasons: eight volunteers in the group with<br />

samples thrice daily (24%) <strong>and</strong> nine volunteers in the group with samples twice<br />

daily (19%). The group size for erythrocytic vaccines was corrected for these<br />

missing data by addition of the respective proportion of volunteers. Similarly,<br />

the power for pre-erythrocytic vaccines was corrected for the number of<br />

volunteers with the first cycle data points below the detection limit: two<br />

volunteers in the group with samples once daily (4%).<br />

121

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!