Glioblastoma multiforme (GBM) is the most common malignant primary brain tumors, with over 10,000 cases diagnosed in the United States each year. Current gold-standard treatment with surgery, radiation, and chemo-therapy have significantly improved 2 year survival but not 5-year survival remains less than 10%. Immunotherapy is a very promising treatment, based on the use of antibodies against immune checkpoints such as cytotoxic T lymphocyte-associated protein 4 (CTLA4) and programmed cell death 1 (PD-1)^1,2. As these immunotherapies enter the clinic, improved assessment of response has become a critical need because immunotherapy induced inflammation produces edema and contrast enhancement difficult to distinguish from progressive tumo. We report preliminary data from a retrospective analysis designed to assess the predictive value of quantitative and semi-quantitative MRI biomarkers including volumetric analysis of contrast enhanced T1, T2 FLAIR, and diffusion weighted imaging (DWI) to distinguish patients with treatment response from non-responders.Longitudinal MRIs were performed on patients diagnosed with recurrent GBM. Patients had MRIs every 2 months with contrast enhanced T1, T2 FLAIR and DWI during the treatment. Tumor volumes-of-interest (VOIs) were drawn on T1 and T2 FLAIR images. The FLAIR VOIs were copied to the registered apparent diffusion coefficient (ADC) maps. Inside the VOIs, only voxels with ADC values in the range of (0.7~1.1)×10^-3 mm²/s were calculated into the ADC volume (Fig 1). The volume of the tumor was also measured on T1 and T2 FLAIR images. The abnormal tissue volumetrics were analyzed along with tumor pathologies and clinical trial length as a primary outcome measure of clinical benefit.Including patients who had (a) progression free survival on the trial >6 months, or (b) trial duration of >2 months with surgical pathology following conventional imaging progression, yielded 10 patients with analyzable data. Among the 10 patients, 5 were deemed to have received benefit from treatment based on progression free survival on trial for more than 6 months. One of this group had pathology demonstrating predominant inflammation and necrotic tumor. The other 5 patients were deemed to have received no substantial benefit from the treatment based on removal from trial for clinically worse symptoms and/or MRI followed by surgical resection revealing substantial viable tumor (Fig 2). The average time on trial for the no benefit group was 82 days, as comparing to 229.2 days for the benefit group. We also plotted the volume change curves for T1, FLAIR and ADC during immunotherapy treatment. For all members of the no benefit group, all three volumes increased, except one patient whose FLAIR volume was stable. For the benefit group, the ADC volumes all increase initially and then decrease or stabilized. Contrast enhanced T1 volumes showed inconsistent patterns: increasing in 2 patients (patient 6 and 8) and decreasing in the other 3. FLAIR volume change pattern was very similar to ADC except for patient 9. The volume of abnormal tissue on ADC seems to correlate better with patient benefit than abnormality on FLAIR or contrast enhanced T1. Figure 3 and 4 show typical cases from the two groups. Time from start of immunotherapy to the point when ADC and FLAIR volume curves began to decrease was 2~6 months for the patients who received benefit. Immunotherapy is known to cause inflammation, making assessment of response difficult. Our preliminary data derived from anti-PD1 treatment of recurrent GBM suggest that all three volumes increase for the first 2~3 months even in patients who ultimately received significant benefit. Although early increase in volume of intermediate/low ADC tissue has been demonstrated to correlate with response and survival in patients treated with XRT and chemoradiation^3,4, it does not seem to predict progression in anti-PD1 immunotherapy patients, at least in the first 2 months. Among the possible explanations for this are non-exclusive possibilities: (a) that immunotherapy may have a delayed onset of cytotoxic effect and (b) that immunotherapy initially produces an inflammatory cell swelling that decreases ADC. Further investigation in animal models seems indicated. This raises the possibility that some patients who receive no benefit from the treatment might have benefitted if they had remained on treatment longer. Also of note, all 3 patients who received combined anti-PD1 and anti-CTLA4 drugs were in the benefit group with a survival of at least 8 months (244 - 320 days) from the start of immunotherapy and two out of the three patients continue on therapy at this time. Whether this reflects improved tolerance of the therapy, improved efficacy or both is under ongoing investigation.