The problem with diffuse large B-cell lymphoma is that it keeps wearing fake mustaches. Doctors call it one disease, but under the microscope and inside the genome, it behaves like a lineup of suspects all insisting they were "just in the neighborhood."
Diffuse large B-cell lymphoma, or DLBCL, is the most common aggressive lymphoma. It starts in B cells, the immune system's antibody-making professionals, which normally behave like well-trained security staff. In DLBCL, some of those staff members go rogue, smash the filing cabinets, and start expanding the operation without permits.
For years, researchers have tried to sort DLBCL into meaningful groups. One classic split asks whether the cancer resembles a germinal center B cell, a B cell in training, or an activated B cell, a cell already getting worked up for immune action. Useful? Yes. Complete? Not even close. Many patients still relapse, and about a third need more treatment after first-line therapy, even in the modern era of R-CHOP, pola-R-CHP, CAR T cells, and bispecific antibodies [1,2].
Enter Enssle and colleagues, who treated 478 DLBCL tumors like a crime scene with three evidence bags: DNA, RNA, and proteins [3]. DNA tells you what mutations are written into the suspect's record. RNA tells you which instructions the cell is reading out loud. Proteins tell you what machinery the cell actually built. That last part matters because cancer cells, like shady contractors, do not always build according to the blueprint.
The Evidence Board Gets Bigger
The team integrated genomic, transcriptomic, and proteomic data and found seven "proteogenotypes," or PGs. Think of these as biological neighborhoods, each with its own local troublemakers, bad lighting, and suspicious van parked by the alley.
The big reveal was PG4.
PG4 was not just another label for the old categories. It cut across activated B-cell-like tumors, germinal center B-cell-like tumors, and cases that had previously refused to sit in any assigned chair. More troubling, PG4 predicted poor outcomes independently of established risk factors like cell-of-origin, International Prognostic Index, and known genetic features [3]. In detective terms, PG4 was the fingerprint nobody had checked for, and it was on the murder weapon.
PG4: The Suspect With a Metabolic Alibi
What made PG4 look so suspicious?
First, these tumors carried a "dark-zone" B-cell flavor. The dark zone is part of the germinal center, where B cells normally proliferate and mutate their antibody genes during immune training. It is supposed to be a temporary boot camp. PG4 seems to borrow that high-growth energy and then forgets to leave. Very relatable, if your relatable phase involves lymphoma biology and bad decisions.
Second, PG4 tumors showed enrichment for BTG1 mutations that can activate MYC [3]. MYC is one of cancer biology's repeat offenders: a growth-driving gene that tells cells to divide, build, burn fuel, and generally behave like they have a motivational poster taped to the nucleus. The twist here is that PG4 showed enhanced MYC activity even without classic MYC translocations. No obvious getaway car, still somehow across town with the jewels.
The team also found heightened TCF3 or TCF4 transcription factor activity, depending on tumor context, plus changes in protein translation that were visible at the proteomic level [3]. That is the kind of clue you miss if you only read the DNA file. The tumor's genome may say "nothing to see here," while the protein machinery is in the basement printing counterfeit growth signals.
The Neighborhood Was Not Helping
Cancer is never only about the cancer cell. The surrounding tumor microenvironment is the local precinct, the streetlights, the nosy neighbor, and sometimes the guy pretending not to see anything.
In PG4, single-cell sequencing and spatial transcriptomics pointed to exhausted CD8 T cells [3]. CD8 T cells are supposed to be the body's tiny bodyguards, the ones that recognize and kill dangerous cells. Exhausted T cells are not dead, exactly. They are more like security guards at hour 19 of a double shift, staring at the monitor while the tumor walks past wearing a fake badge.
That matters because immune exhaustion may help explain why some tumors resist chemoimmunotherapy and why future treatment might need to account for both the tumor's internal wiring and its immune surroundings. Recent DLBCL work has increasingly shown that single-cell biology and tumor ecosystems can reveal risk patterns missed by bulk profiling alone [4].
Why This Case Matters
If PG4 holds up in more cohorts and clinical trials, it could change how high-risk DLBCL gets recognized. Today, clinicians already use clinical risk scores, cell-of-origin, double-hit biology, and genetic subtype information to guide thinking. PG4 suggests another layer: some dangerous tumors may share a protein-and-immune program even when their genetics look different.
That could help identify patients who need something smarter than standard treatment escalation. Maybe PG4 tumors need therapies aimed at B-cell receptor signaling, MYC-linked biology, protein translation, immune exhaustion, or combinations that do not politely wait for the tumor to adapt. Prior studies already show that molecular subtypes can influence response to targeted additions like ibrutinib in selected DLBCL groups [5]. PG4 adds a new suspect profile to test.
The catch, because science always makes you sign paperwork before the car chase, is that this is not yet a bedside test for choosing therapy tomorrow morning. Proteogenomics is powerful, but it is more complicated than routine pathology. The next job is validation, simplification, and clinical trial proof.
Still, this paper gives the field a sharper magnifying glass. DLBCL has been getting away with biological impersonation for years. PG4 may not solve the whole case, but it gives investigators a new lead, a motive, and a very suspicious protein trail.
References
-
Ansell SM, Nowakowski GS. Current treatment algorithm: diffuse large B cell lymphoma. Blood Cancer Journal. 2026;16:32. DOI: 10.1038/s41408-026-01458-2
-
Tilly H, Morschhauser F, Sehn LH, et al. Polatuzumab vedotin in previously untreated diffuse large B-cell lymphoma. New England Journal of Medicine. 2022;386:351-363. DOI: 10.1056/NEJMoa2115304
-
Enssle JC, Häupl B, Qoku A, et al. Pathogenesis of diffuse large B cell lymphoma proteogenotypes. Cancer Cell. 2026. DOI: 10.1016/j.ccell.2026.05.008
-
Wang B, Wright G, Enssle JC, et al. Axes of biological variation in diffuse large B cell lymphoma. Cancer Cell. 2026;44:567-585.e12. DOI: 10.1016/j.ccell.2025.12.015
-
Wilson WH, Wright G, Huang DW, et al. Effect of ibrutinib with R-CHOP chemotherapy in genetic subtypes of DLBCL. Cancer Cell. 2021;39:1643-1653.e3. DOI: 10.1016/j.ccell.2021.10.006. PMCID: PMC8722194
Disclaimer: The image accompanying this article is for illustrative purposes only and does not depict actual experimental results, data, or biological mechanisms.