Kinetic analysis of 2-[carbon-11]thymidine PET imaging studies: Compartmental model and mathematical analysis

David A. Mankoff, Anthony F. Shields, Michael M. Graham, Jeanne M. Link, Janet F. Eary, Kenneth A. Krohn

Research output: Contribution to journalArticle

100 Scopus citations

Abstract

Carbon-11-thymidine is a PET tracer of DNA synthesis and cellular proliferation. Quantitative analysis of [11C]thymidine images is complicated by the presence of significant quantities of labeled metabolites. Estimation of the rate of thymidine incorporation into DNA using [11C]thymidine requires a kinetic model that is capable of describing the behavior of thymidine and labeled metabolites. Methods: Based on previous studies with labeled thymidine, we constructed a five-compartment model describing the kinetic behavior of 2-[11C]thymidine and its labeled metabolites. In addition, we have performed a series of calculations and simulations to calculate the sensitivity and identifiability of model parameters to estimate the extent to which individual parameters can be estimated; to determine appropriate model constraints necessary for reproducible estimates of the constant describing flux of thymidine from the blood into DNA, i.e., thymidine flux constant; and to determine the potential accuracy of model parameter and thymidine flux constant estimates from PET imaging data. Results: The underlying assumptions in the thymidine compartmental model lead to a description of the thymidine flux constant for DNA incorporation in terms of model parameters. Sensitivity and identifiability analyses suggest that the model parameters pertaining to label metabolites will be difficult to estimate independently of the thymidine parameters. Exact evaluation of the kinetic parameters of the labeled metabolites is not the principal goal of this model. Simulations were performed that suggest that it is preferable to tightly constrain these parameters to preset values near the center of their expected ranges. Although it is difficult to estimate individual thymidine model parameters, the flux constant for incorporation into DNA can be accurately estimated (r > 0.9 for estimated versus true simulated flux constant). Flux constant estimates are not affected by modest levels of local degradation of thymidine that may occur in proliferating tissue. Conclusion: By using a kinetic model for thymidine and labeled metabolites, it is possible to estimate the flux of thymidine uptake and incorporation into DNA and, thereby, noninvasively estimate regional cellular proliferation using [11C]thymidine and PET.

Original languageEnglish (US)
Pages (from-to)1043-1055
Number of pages13
JournalJournal of Nuclear Medicine
Volume39
Issue number6
StatePublished - Jun 1 1998

    Fingerprint

Keywords

  • Cancer
  • Carbon-11-thymidine
  • Kinetic modeling
  • PET

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging

Cite this