Secondary ion mass spectrometry (SIMS) using a gas cluster ion beam (GCIB), in this case 40 keV (CO2)7k+, was used to map the intact lipid signals across 14 lymph node samples representing diffuse large B-cell lymphoma (DLBCL), a common and aggressive form of lymphoma, and nonmalignant controls. The analysis allowed the samples to be classified as malignant or nonmalignant and also highlighted additional aggressive cancer signature in a DLBCL sample with an unusually high proliferation index. A complementary, combined k-means/image PCA approach was used to interrogate the data, highlighting the pros and cons of the different approaches and potential sources for misclassification/diagnoses resulting from the heterogeneity of the DLBCL samples. Compared to other cancer types, lymphoma results in a reduction of non-neoplastic inflammatory cells and their characteristic signals that are often classed as cancer-related, highlighting the need to consider disease heterogeneity when examining MS data. While delivering new information regarding the chemistry of lymphoma, the results also highlight the need for cellular precision with high chemical specificity and sensitivity, and the challenges associated with spectral/spatial classification of such complex samples and data where differently aggressive cancer samples show different signatures and pockets of different cell types, in this case histiocytes, can show intermediate cancer/healthy lipid profiles.