We thank Gleeson et al 1 for their comments.We agree that immunohistochemistry (IHC) methods for cell of origin (COO) determination have had variable reproducibility reported in many studies, as well outlined by Gleeson et al.However, when performed in a central laboratory and interpreted by an experienced hematopathologist, a significant level of concordance can be achieved.In recent published work by the group from the British Columbia Cancer Agency (BCCA), original members of the Lymphoma/Leukemia Molecular Profiling Project who developed the Hans algorithm compared the IHC-based Hans algorithm to gene expression profiling (GEP) and achieved greater than 90% concordance. 2In our study, tumor biopsies from a subset of 40 patients were submitted for repeat IHC and independent review by an author and member of the BCCA (R.D.G.) who was blinded to the original Hans algorithm subclassification performed at the Mayo Clinic.In these patients, 93% concordance was seen between the Mayo Clinic and BCCA hematopathology laboratories, as stated in the paper, 3 which reflects a high level of reproducibility between the two centers.Thus, with established a priori scoring criteria and methodology, IHC can be reproducible in experienced hands. 4In addition, IHC methods to demonstrate the differential activity of lenalidomide in relapsed/refractory diffuse large B-cell lymphoma (DLBCL) also were performed at Mayo Clinic by using the same methodology. 5Arguably the most important, albeit indirect, validation of the Hans algorithm in our study rests with the analysis of the R-CHOP (rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisolone) -treated control cohort.In this cohort of patients, COO determination on the basis of IHC was performed in the same laboratory at Mayo Clinic, and COO was predictive of outcome both for progression-free and overall survivals.This finding contrasts with the lack of a survival difference in patients treated with R-CHOP plus lenalidomide (R2CHOP) on the basis of COO determination with the Hans algorithm, which establishes that this methodology does identify a clinically relevant subgroup of patients who benefit from the addition of lenalidomide.We agree with Gleeson et al 1 that COO determination on the basis of IHC suffers from issues related to reproducibility. [6]7][8] Therefore, we believe that studies that claim to demonstrate a differential benefit across COO subtypes of DLBCL must include a control cohort to establish that the IHC algorithm used distinguishes different outcomes for germinal center B-cell (GCB) versus non-GCB types after treatment with R-CHOP.Although this does not address the issue of ground truth for COO determination, it does demonstrate that the methodology used is able to identify a subset of patients who derive benefit from the novel agent being tested.As Gleeson et al noted, 1 determination of the COO in DLBCL has more than 15 years of use now, and we have witnessed a significant evolution of COO assessment methods.The initial requirement for fresh tissue for Affymetrix (Santa Clara, CA) GEP and difficulties in reproducibility of IHC methods have hampered the introduction of COO testing into routine clinical practice.GEP that is based on the