The most likely reasons are 1) you have not gone out far enough in x or 2) you have not written a logistic expression.
Note that r must be negative to have your logistic increase asymptotically, and >0 to decrease asymptotically to 0. Perhaps you just need to put a negative in front of r?
C = K*C0*exp(-r*t)./(K-C0+C0*exp(-r*t));
plot(t,cumsum(C),'--bo','LineWidth',0.5,'MarkerSize',5,'MarkerEdgeColor','black')
title( 'Brain Tumor Population Against Time' )
ylabel('Brain Tumor Population (cells)')
Still, I don't believe you should have to use a cumsum to get the results, and they still don't look like a typical logistic plot, so perhaps check your equation? For example, here's an equation that I think is more typical of logistic growth. At  , C=C0 and at
, C=C0 and at  , C=K. It is also able to obtain the expected results without using cumsum. I do make it go to 1400 to see the assymptote.
, C=K. It is also able to obtain the expected results without using cumsum. I do make it go to 1400 to see the assymptote. plot(t,C,'--bo','LineWidth',0.5,'MarkerSize',5,'MarkerEdgeColor','black')
title( 'Brain Tumor Population Against Time' )
ylabel('Brain Tumor Population (cells)')