Evaluation Of Lactate As A Predictive Marker Of Survival And Local Response To Radiation Therapy In Patients With Gbm

Glioblastoma Multiforme (GBM) is associated with poor prognosis as local relapse occurs several months after chemo-radiation therapy (RT). There is considerable heterogeneity in response to therapy between patients, prognostic markers are thus necessary to understand why some patients present a relapse earlier than others. Because Lactate (Lac) is both a marker of hypoxia and malignancy, our aim was to assess the prognostic value of Lac in GBM before RT, detected by in vivo three-dimensional Magnetic Resonance Spectroscopic Imaging (MRSI). We also evaluated the spatial correlation between pre-RT Lac volumes and contrast enhanced (CE) volumes visualized on T1 gadolinium (T1Gd) weighted MRI before RT and at relapse. The analysis concerned MRI and MRSI data from 14 patients with GBM included in a prospective phase II clinical trial associating treatment with farnesyltransferase inhibitor and conformational RT. For each patient, pre-RT Lac/N-Acetyl Aspartate (NAA, neurotransmitter) ratio maps were built. As no specific threshold is defined yet for Lac/NAA ratio, several thresholds (between 1 and 2) were considered for survival analysis. For the spatial analysis, pre-RT areas with an abnormal Lac/NAA ratio were contoured and spatially compared to volumes of anatomic abnormalities before RT and CE volumes at relapse. Patients with a maximum Lac/NAA ratio above 1.6 had a significantly shorter median time to progression (4 months) compared to patients with a maximum ratio below 1.6 (9 months, p < 0.05). Lac/NAA ratio > 1.6 (LNR1.6) was then considered abnormal. The spatial analysis showed that pre-RT LNR1.6 volumes were found predominantly in CE areas, confirming the presence of increased Lac signal in areas of hypoxia (blood brain barrier breakdown) and neoplastic proliferation (Warburg effect). The volumes of intersection with CE at relapse were significantly smaller for pre-RT LNR1.6 volumes compared to pre-RT CNR2 (Choline/NAA > 2) volumes (p < 0.01), a metabolic index used as a reference to predict the site of local relapse. Therefore, the spatial distribution of Lac before RT did not provide additional information on the site of local tumor growth. Our study indicated that pre-RT LNR1.6 in GBM is predictive of time to progression, but does not provide insights about the local evolution of the tumor. This metabolic tool may be used in future clinical trials to stratify patients so that new therapeutic strategies could be tailored to each patient.