Simultaneous estimation of the cellular water exchange rate, intracellular volume fraction, and longitudinal relaxation rate in cancer cells

The purpose of the current study was to investigate the feasibility of simultaneously estimating the cellular water efflux rate (kie$$ {k}_{ie} $$), intracellular longitudinal relaxation rate (R10i$$ {R}_{10i} $$), and intracellular volume fraction (vi$$ {v}_i $$) of a cell suspension using multiple samples with different gadolinium concentrations. Numerical simulation studies were conducted to assess the uncertainty in the estimation of kie$$ {k}_{ie} $$, R10i$$ {R}_{10i} $$, and vi$$ {v}_i $$ from saturation recovery data using single (SC) or multiple concentrations (MC) of gadolinium-based contrast agent (GBCA). In vitro experiments with 4 T1 murine breast cancer and SCCVII squamous cell cancer models were conducted at 11 T to compare parameter estimation using the SC protocol with that using the MC protocol. The cell lines were challenged with a Na+/K+-ATPase inhibitor, digoxin, to assess the treatment response in terms of kie$$ {k}_{ie} $$, R10i$$ {R}_{10i} $$, and vi$$ {v}_i $$. Data analysis was conducted using the two-compartment exchange model for parameter estimation. The simulation study data demonstrate that the MC method, compared with the SC method, reduces the uncertainty of the estimated kie$$ {k}_{ie} $$ by decreasing the interquartile ranges from 27.3% +/- 3.7% to 18.8% +/- 5.1% and the median differences from ground truth from 15.0% +/- 6.3% to 7.2% +/- 4.2%, while estimating R10i$$ {R}_{10i} $$ and vi$$ {v}_i $$ simultaneously. In the cell studies, the MC method demonstrated reduced uncertainty in overall parameter estimation compared with the SC approach. MC method-measured parameter changes in cells treated with digoxin increased R10i$$ {R}_{10i} $$ by 11.7% (p = 0.218) and kie$$ {k}_{ie} $$ by 5.9% (p = 0.234) for 4 T1 cells, respectively, and decreased R10i$$ {R}_{10i} $$ by 28.8% (p = 0.226) and kie$$ {k}_{ie} $$ by 1.6% (p = 0.751) for SCCVII cells, respectively. vi$$ {v}_i $$ did not change noticeably by the treatment. The results of this study substantiate the feasibility of using saturation recovery data of multiple samples with different GBCA concentrations for simultaneous measurement of the cellular water efflux rate, intracellular volume fraction, and intracellular longitudinal relaxation rate in cancer cells.