Abstract 2701

Because of the wealth of information obtained using MR technology, this technique promises to be one of the best imaging methods in cancer. In addition, MR technology is fairly noninvasive and does not require ionizing radiation or radioactivity, making it one of the safest imaging modalities available. This study aims to evaluate the predictive and diagnostic capabilities of a multimodality MR exam to study non-Hodgkin lymphoma (NHL) patients.

Experimental Design:

Under ethical review board approval, a pretreatment MR exam was obtained in NHL patients, which included clinical MRI, diffusion weighted imaging (DWI), 1H MR spectroscopy (MRS) and/or 31P MRS. Using 3D-localized, 1H-irradiated 31P MRS the phosphomonoester (PME) levels normalized to nucleoside triphosphates (PME/NTP) were measured. Using single voxel 1H MRS the tumor total choline to water (tCho/H2O) ratio was determined. DWI was acquired using echo-planar imaging with fat suppression and b-values of 0 and 1000 s/mm2 to obtain the apparent diffusion constant (ADC) of water in the tumor. In regard to clinical information, treatment response was assessed six-months after treatment and the patients grouped as those with a complete response (CR) and those without a complete response (NCR). Time to treatment failure (TTF), was measured as the time between the end of one treatment and the start of a new one.


Pretreatment PME/NTP values were measured in 59 newly diagnosed NHL patients. Twenty-seven of these patients were diffuse large B-cell lymphoma (DLBCL) patients. From these patients, 20 were treated with CHOP or equivalent therapy. The significantly different mean pretreatment PME/NTP values (SD, n) for CR and NCR were 1.42 (0.41, 13) and 2.46 (0.40, 7), respectively (p<0.00001), with sensitivity and specificity of 0.85 in a Fisher test (p<0.01). PME/NTP correlated with TTF by Cox (p<0.02) and Kaplan-Meier tests (p<0.00001). The additional seven DLBCL patients were treated with added rituximab also showing significantly different mean pretreatment PME/NTP values: CR, 1.56 (0.61, 4) vs. NCR, 3.18 (0.41,3), (p<0.02), as well as Kaplan-Meier curves for TTF (p < 0.01). The remaining 32 patients with other forms of NHL were studied together. Again in this group of patients, the PME/NTP was significantly different in the response groups: CR, 1.41 (0.35, 5) vs. NCR, 2.11 (0.16, 27), p<0.04. In seven NHL patients the correlation of the PME/NTP values determined by 31P MRS vs. the tCho/H2O values determined by 1H MRS prior to receive treatment showed a statistically significant linear regression (y = 0.16× – 0.77, r2 = 0.7, p < 0.005). DWI was obtained in 15 NHL patients refractory to previous treatments demonstrating that the mean ADC value of refractory NHL patients is lower than the one in normal volunteers (0.91 [0.23, 15] vs. 1.13 [0.14, 10], p<0.01), but similar to those reported from newly diagnosed patients.


Our data validate the capabilities of a multimodality MR exam to study NHL patients by demonstrating that: 1) the pretreatment tumor PME/NTP levels predict both responses to treatment and TTF in different histological subtypes of NHL patients; 2) the correlation of PME/NTP with tCho/H2O that suggests that prediction of treatment outcome may also be possible using the ratio determined by 1H MRS, which will allow to use the increased sensitivity of 1H MRS; 3) the potential additional information provided by the simultaneous determination of 31P and 1H MRS; and 4) our preliminary DWI results suggesting that for newly diagnosed and refractory NHL patients, the quantitative ADC evaluation can aid assessing malignant masses and differentiate them from normal lymph nodes.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.