Recovery of renal retention function by differentiating the Patlak-Rutland plot.

Introduction: It has been suggested that renal retention function can be estimated by differentiation of the Patlak-Rutland plot.
Aim: To evaluate the error resulting from this method.
Methods: A total of 5800 renograms were generated by convolving real input functions with artificial retention functions. Ten bi-exponential plasma disappearance curves of Tc mercaptoacetyltriglycine (Tc-MAG3) with varying renal clearances served as input functions and 580 retention functions with mean transit time between 3 and 60 min, and variable ratios of minimal to mean transit time served as original retention functions. The retention function was estimated by differentiation of the Patlak-Rutland plot of each renogram. A variant of this retention function was calculated by setting negative values to zero.
Results: Minimal transit time was estimated correctly in all cases but the recovered retention function systematically underestimated the original, and negative values were observed. Mean transit time was underestimated with a difference ranging from -22.05 to -0.06 min. By setting the negative values to zero, a less important underestimation was observed which ranged from -13.37 to 0.00 min. Maximal transit time was underestimated systematically with a difference ranging from -22.05 to 0.00 min. All differences were influenced by mean transit time, renal clearance and ratio of minimal to mean transit time.
Conclusion: Estimation of the renal retention function by differentiation of the Patlak-Rutland plot results in systematic and sometimes important underestimations. By setting the negative values of the recovered retention function equal to zero, important but still partial improvement can be obtained.