Tumour T(1) changes in vivo are highly predictive of response to chemotherapy and reflect the number of viable tumour cells – a preclinical MR study in mice

BACKGROUND: Effective chemotherapy rapidly reduces the spin–lattice relaxation of water protons (T(1)) in solid tumours and this change (ΔT(1)) often precedes and strongly correlates with the eventual change in tumour volume (TVol). To understand the biological nature of ΔT(1), we have performed studies in vivo and ex vivo with the allosteric mTOR inhibitor, everolimus. METHODS: Mice bearing RIF-1 tumours were studied by magnetic resonance imaging (MRI) to determine TVol and T(1), and MR spectroscopy (MRS) to determine levels of the proliferation marker choline and levels of lipid apoptosis markers, prior to and 5 days (endpoint) after daily treatment with vehicle or everolimus (10 mg/kg). At the endpoint, tumours were ablated and an entire section analysed for cellular and necrotic quantification and staining for the proliferation antigen Ki67 and cleaved-caspase-3 as a measure of apoptosis. The number of blood-vessels (BV) was evaluated by CD31 staining. Mice bearing B16/BL6 melanoma tumours were studied by MRI to determine T(1) under similar everolimus treatment. At the endpoint, cell bioluminescence of the tumours was measured ex vivo. RESULTS: Everolimus blocked RIF-1 tumour growth and significantly reduced tumour T(1) and total choline (Cho) levels, and increased polyunsaturated fatty-acids which are markers of apoptosis. Immunohistochemistry showed that everolimus reduced the %Ki67(+) cells but did not affect caspase-3 apoptosis, necrosis, BV-number or cell density. The change in T(1) (ΔT(1)) correlated strongly with the changes in TVol and Cho and %Ki67(+). In B16/BL6 tumours, everolimus also decreased T(1) and this correlated with cell bioluminescence; another marker of cell viability. Receiver-operating-characteristic curves (ROC) for everolimus on RIF-1 tumours showed that ΔT(1) had very high levels of sensitivity and specificity (ROC(AUC) = 0.84) and this was confirmed for the cytotoxic patupilone in the same tumour model (ROC(AUC) = 0.97). CONCLUSION: These studies suggest that ΔT(1) is not a measure of cell density but reflects the decreased number of remaining viable and proliferating tumour cells due to perhaps cell and tissue destruction releasing proteins and/or metals that cause T(1) relaxation. ΔT(1) is a highly sensitive and specific predictor of response. This MRI method provides the opportunity to stratify a patient population during tumour therapy in the clinic.