FlowCytoNews
Flow cytometry, gated weekly.
ISSUE #6 · MONDAY EDITION · THE LITERATURE · OCT 26, 2026
Welcome to the Monday edition of FlowCytoNews, a day late this week. Monday is The Literature: the cytometry papers and preprints that went online since our last Literature edition, who did the work, and the number that matters from each. Wednesday is On the Calendar and Friday is Industry Moves. This time: the share of naive-like T cells in 580 CAR-T products tracked with progression-free survival, flow MRD before blinatumomab predicted survival in Ph+ ALL even after it cleared, and a stem-cell gate flagged MDS relapse after transplant. Tell us what we missed — hit reply.
■ The Literature
No new OMIP in Cytometry Part A turned up in our search this fortnight, and neither did a first panel in Cytometry Part B's clinical series.
In 580 axi-cel patients, products with a high naive-like T-cell share went with longer PFS
Justin Budka and Davide Bedognetti (Kite, a Gilead Company, Santa Monica) analyzed the phenotype of axicabtagene ciloleucel products from 580 patients with non-Hodgkin lymphoma across several trials, with a median follow-up of 41.9 months. Patients whose product had a high share of naive-like (CCR7+CD45RA+) T cells had longer progression-free survival than those with a low share (hazard ratio 0.46, 95% CI 0.31–0.68), with no effect on safety. The association held after adjusting for product expansion, outperformed tumor burden and inflammatory markers in multivariate models, and did not depend on the tumor microenvironment in the 112 patients where that was assessed. The more prior lines of therapy a patient had, the lower the naive-like share, in both the product and the matched apheresis. An independent cohort of 203 patients treated with axi-cel-based therapy or brexucabtagene autoleucel confirmed the main findings, and mass cytometry on 116 products linked PFS to CD8 naive-like cells lacking TIGIT, LAG-3 and PD-1. Sixteen of the 18 authors are from Kite; the other two list Dana-Farber Cancer Institute, one of them also Sapporo Medical University.
▪ Nature Medicine · Oct 23 · 10.1038/s41591-026-04707-w
Why it matters: A CCR7/CD45RA tube is one of the oldest T-cell assays in the lab, and here it tracks progression-free survival in a sponsor's own 580-patient dataset. Expect product characterization to treat the naive-like share as a quality attribute, which turns a research gate into a number someone may want a specification on. That gate is less clean than it looks; see Bench Notes.
Flow MRD before blinatumomab predicted worse survival in Ph+ ALL, even after it cleared
Benjamin Frost (Brigham and Women's Hospital, Boston) and Evan Chen (Dana-Farber Cancer Institute), with co-authors at Memorial Sloan Kettering, Mass General Brigham and Yale, looked back at 81 adults with Ph+ B-ALL who were MRD-negative (below 10⁻⁴) in first remission after blinatumomab consolidation. They split them by flow MRD before blinatumomab: 56 negative and 25 positive, with similar baseline characteristics. At a median follow-up of 31.0 months, the pre-blinatumomab flow-negative group had better disease-free and overall survival, with multivariable hazard ratios of 0.27 (95% CI 0.08–0.87) and 0.17 (0.03–0.90). IKZF1-plus genotype and extramedullary disease at diagnosis also predicted worse disease-free survival. In exploratory analyses, the effect held for patients who did not have a transplant in first remission but not for those who did. No industry co-authors.
▪ Blood Advances · Oct 15 · 10.1182/bloodadvances.2026020724
Why it matters: In patients who did not go to transplant, a negative result after blinatumomab did not erase a positive one before it. The pre-consolidation flow result belongs in the history the clinician sees at the transplant decision. With 25 MRD-positive patients and upper confidence limits near 1, treat the direction as likely and the size as rough.
After transplant for MDS, a stem-cell flow gate split 3-year relapse 40.4% versus 6.5%
Wei-Bing Wu and Ying-Jun Chang (Peking University People's Hospital, Beijing), with He-Bing Chen (Academy of Military Medical Sciences, Beijing), prospectively followed 143 patients with MDS with excess blasts who had an allogeneic transplant. They used single-cell proteogenomics to characterize CD34+ MDS subsets, then measured a stem-like population defined by CD34, CD38 and a six-marker cocktail of aberrant antigens (CD56, CD22, CD11b, CD7, Tim-3, CLL-1) by multiparameter flow. Before transplant, a level of 0.1064% or more independently predicted relapse, while conventional flow measures of MDS cells did not. After transplant, patients positive for the stem-like cells (0.0081% or more) had 3-year relapse of 40.4% versus 6.5% and disease-free survival of 57.4% versus 85.1%. Conventional post-transplant flow MRD also separated patients (relapse 60.0% versus 12.8%); the authors report the stem-cell assay had higher sensitivity and C-index. No industry co-authors.
▪ Transplantation and Cellular Therapy · Oct 14 · 10.1016/j.jtct.2026.10.016
Why it matters: Standard flow MRD still separated patients after transplant; the authors' claim is that the stem-cell gate is more sensitive. The bigger result is before transplant, where the conventional measure showed no association with relapse. The abstract does not say how many centers enrolled patients, and thresholds quoted to four decimal places will not transfer to another lab without its own validation. For the French network's approach to the same CD34+ compartment, see Also out.
In 502 AML marrows, the immunophenotype predicted genotype but not response to therapy
Tim Mocking and Costa Bachas (Amsterdam UMC), with co-authors from the HOVON and SAKK trial groups and centers across Europe, modeled more than 50 million cells from the diagnostic bone marrow of 502 AML patients, analyzing leukemic and normal cells together. Seven patient clusters emerged; they matched recurrent mutations and fusions, including t(8;21), and predicted genomic subgroups. The clusters were not associated with therapy response. Loss of HLA-DR on blasts at diagnosis was independently linked to worse outcome in patients with chromatin-spliceosome mutations, and a cluster of about 11% of patients with lymphocyte-rich marrow had outcomes that diverged by genetic risk. No industry co-authors.
▪ iScience · Oct 12 · 10.1016/j.isci.2026.117762
Why it matters: This is the second large group in a month to predict AML genetics from the diagnostic tube, after the 885-patient EuroFlow study we covered Oct 5. Both say the files you already acquire hold more than the report uses, and neither replaces sequencing. The practical item is HLA-DR: if blasts are HLA-DR-negative, say so explicitly in the report.
Celiac flow in children: new cut-offs, and the duodenal bulb picked up potential disease
Laura Escartín Madurga and Gerardo Rodríguez Martínez (Hospital Clínico Universitario Lozano Blesa and University of Zaragoza) reviewed duodenal biopsies analyzed by flow cytometry from 190 children aged 14 or under at several centers: 54 with active celiac disease, 69 with potential celiac disease and 67 controls. Optimized cut-offs for active disease were TCRγδ+ intraepithelial lymphocytes above 11.1% in both the bulb and the distal duodenum, and surface CD3− intraepithelial lymphocytes below 6.4% in the bulb and below 8.5% in the distal duodenum. These agreed better with serology and histology than the reference values, and sensitivity rose when both sites were analyzed. In potential celiac disease, the drop in CD3− cells appeared only in the bulb. No industry co-authors.
▪ Cytometry Part B · Oct 22 · 10.1002/cyto.b.70084
Why it matters: If your lab runs the intraepithelial lymphogram on distal duodenum only, ask for a bulb biopsy too; in potential celiac disease it was the only site that changed. The cut-offs come from a retrospective pediatric series, so validate them before applying them to adults.
Also out
AI reads the single-cell data behind routine CBCs. Ya-Lin Chen and Brody Foy (UW Medicine, Seattle) applied clustering and autoencoders to the cell-level measurements that hematology analyzers already record, across 242,584 CBCs from 127,543 patients. The new markers were associated with mortality, admission and later anemia, cancer or cardiovascular disease after adjusting for standard CBC indices. — Clin Chem Lab Med · Oct · 10.1515/cclm-2026-0822
TRBC1/2 skewing as a low-cost clonality check in HTLV-1. Koji Jimbo (University of Tokyo) and Shigeo Fuji (Osaka International Cancer Institute), with Imperial College London, combined HAS-Flow with TRBC1/2 staining in 53 infected people; the dominant TRBC ratio separated adult T-cell leukemia/lymphoma from other cases with an AUC of 0.913 and tracked integration-site clonality. — Br J Haematol · Oct 16 · 10.1111/bjh.70889
CD34+ progenitors in MDS: the French network's approach. Nicolas Chapuis (Hôpital Cochin, Paris) and Valérie Bardet (Hôpital Ambroise Paré, Boulogne-Billancourt) review quantitative, aberrancy and architecture changes in CD34+ cells and describe the CytHem-SMD group's integrated strategy and its efforts to standardize it across labs. — Cytometry Part B · Oct 23 · 10.1002/cyto.b.70082
Pediatric megakaryoblastic leukemia has its own phenotype. Xiaojin Cai and Wenyu Yang (Institute of Hematology & Blood Diseases Hospital, Tianjin), with one co-author from Tianjin Jinyu Medical Laboratory, compared 54 non-Down syndrome AMKL cases with 59 other AML: 14 AMKL cases were CD33-positive but CD13- and CD117-negative and 3 were negative for all three, patterns not seen in controls. — Cytometry Part B · Oct 26 · 10.1002/cyto.b.70085
"Spurious CD34 expression in mantle cell lymphoma due to non-specific binding to PerCP-Cy5.5." A letter from Abhilash Krishnan and Rakhee Kar (JIPMER, Puducherry). No abstract online, so title only. — Cytometry Part B · Oct 15 · 10.1002/cyto.b.70083
Patient-specific DNA PCR MRD for myeloid disease, where flow and RNA qPCR fall short. Libby Huang and Toby Trahair (Children's Cancer Institute, Sydney) built genomic PCR-MRD assays for 97.6% of patients and monitored 40 prospectively; serial MRD tracked outcome. — Br J Haematol · Oct 20 · 10.1111/bjh.70880
How blood immune subsets shift with age in healthy adults. Liyuan Zheng and Taisheng Li (Peking Union Medical College Hospital, Beijing) ran two 12-color panels on 146 adults aged 20 to 92; CD8 TEMRA accumulation, CD28 loss and naive CD8 depletion had the strongest age correlations in the 43-person second panel (ρ beyond ±0.5). — Front Aging · Oct 9 · 10.3389/fragi.2026.1847033
T-cell count as a stand-in for lymphopenia grading. Shicheng Li (Suzhou City University) and Junjun Zhang (Second Affiliated Hospital of Soochow University) studied 2,896 hospitalized adults: T-cell cut-offs of 700, 573 and 356 cells/µL agreed with standard grading (weighted kappa 0.792), but no subset discriminated CRP-defined inflammation well (best AUC 0.667). — J Clin Lab Anal · Oct 15 · 10.1002/jcla.70375
Three functional clinical assays, written up as protocols. Caleb Cornaby and Maurice O'Gorman (Children's Hospital Los Angeles) cover EdU lymphocyte proliferation, DHR-123 neutrophil oxidative burst and leukocyte adhesion deficiency testing for inborn errors of immunity. — Curr Protoc · Oct · 10.1002/cpz1.70462
Preprints
Not peer reviewed.
A 27-color whole-blood panel for platelet-leukocyte aggregates on 40+ subsets. Haiyuan Ding and corresponding author Scott Brakenridge (University of Washington) found aggregate frequency ranged from 1.7% on memory CD4 T cells to over 50% on monocyte subsets. Anticoagulant, centrifugation, dilution and processing time decided whether the counts were reproducible, and imaging flow confirmed them. — bioRxiv · Oct 7 · 10.64898/2026.09.30.755388
Sorting T cell–monocyte doublets that are actually touching. N. Kang and corresponding author Julie Burel (La Jolla Institute for Immunology) used spectral image-guided sorting to split circulating doublets from tuberculosis patients into physically interacting and coincidental pairs; interacting pairs showed monocyte-to-T cell RNA transfer and stronger T-cell cytotoxicity signatures. — bioRxiv · Oct 16 · 10.64898/2026.10.09.758094
A 21-color spectral panel for rat tissues, with PE left open. Z. Liang and corresponding author Jie Zhang (Peking University Third Hospital, Beijing) built a 21-color phenotyping panel and a 9-color cytokine panel for rat blood, spleen and lymph node, reserving PE for antibodies or probes users add. — bioRxiv · Oct 13 · 10.64898/2026.10.13.758716
■ Bench Notes
Before naive-like T cells become a specification, pin down the gate
Kite's 580-patient analysis (10.1038/s41591-026-04707-w) ties progression-free survival to the share of CCR7+CD45RA+ T cells in the product. If you characterize cell products or run immune monitoring for a sponsor, someone will ask you for that number. Make it reproducible first:
Write the definition into the SOP. Name the parent gate (live CD3+, and CAR+ or all T cells) and report CD4 and CD8 separately. Kite's mass cytometry signal was in CD8 naive-like cells lacking TIGIT, LAG-3 and PD-1, so say whether exhaustion markers are part of your definition.
Know what CCR7+CD45RA+ contains. Stem cell memory T cells sit inside the naive-like compartment and are told apart by high CD95 (Gattinoni et al., Nat Med 2011, 10.1038/nm.2446). A preprint we covered Oct 5 found CCR7/CD45RA gating mixes transcriptionally distinct CD8 states (10.64898/2026.09.15.751807). If the difference matters for your product, add CD95.
Place the boundaries from controls, then lock them. Set the CCR7 and CD45RA gates from fluorescence-minus-one controls and fix them in the analysis template. Moving them case by case changes the number you report.
Measure the starting material too. Kite found the naive-like share fell with prior lines of therapy in both the product and the matched apheresis. If you receive apheresis material, record the same gate there so the two can be compared.
Run a reference sample with every batch. Gosling et al. (Front Immunol, Oct 5, 10.3389/fimmu.2026.1959895) kept instrument and unmixing performance stable for more than two years with rainbow beads and antibody spectra. Against a stabilized control, they found gating built on dimensionality reduction more reproducible between operators than two-dimensional gating. Chart your reference sample's naive-like percentage over time.
What did we miss? Hit reply — corrections run at the top of the next edition. If a colleague characterizes cell therapy products or signs out leukemia MRD, forward this to them.
— Flow Novice