This Week in Hematology — Sep 16, 2026

Welcome to This Week in Hematology. This week we're covering 10 notable papers spanning inherited red cell disorders and the anaemias, immune-based therapies moving earlier into plasma cell and acute leukaemia, and protein replacement strategies in bleeding and thrombotic microangiopathy. Let's dive in.

We start with the red cell, and the biggest news comes from The Lancet, where Cappellini and colleagues report ENERGIZE-T, a global double-blind phase 3 trial of mitapivat, an oral allosteric activator of pyruvate kinase, in adults with transfusion-dependent alpha or beta thalassaemia [1]. Two hundred and fifty-eight patients across nineteen countries were randomised two to one to mitapivat 100 milligrams twice daily or placebo for 48 weeks, and the primary endpoint was a transfusion reduction response, meaning at least a halving of transfused red cell units and a drop of at least two units over any consecutive twelve-week window. Roughly a third OF PATIENTS on mitapivat achieved that response compared with about one in eight on placebo, an absolute gain of eighteen percentage points and statistically significant. Ninety-two percent of participants completed the blinded period. Adverse events were common in both arms and mostly familiar ones, headache, upper respiratory infection, initial insomnia, diarrhoea and fatigue, with serious events actually numerically lower on mitapivat than placebo, though discontinuation for adverse events was more frequent on drug, six percent versus one percent, and there were no deaths. The practical significance is twofold. This is the first oral disease-modifying therapy for transfusion-dependent thalassaemia, and critically it includes alpha thalassaemia, a group with essentially nothing beyond transfusion and chelation. The honest framing for the clinic is that most patients did not meet the response threshold, so mitapivat is an option to trial rather than a transformation for everyone, but for a patient exhausted by a lifetime of transfusion, a pill that cuts units by half is worth attempting.

Staying with haemoglobinopathies, two studies in the British Journal of Haematology sharpen how we counsel patients about risk. Auger and colleagues analysed nearly 2.7 million pregnancies in Quebec over three decades and found that women with sickle cell anaemia who had crises, whether during pregnancy or outside of it, had roughly eight to nine times the risk of severe maternal morbidity compared with women without a sickling disorder [8]. The specific complications are striking, with acute renal failure elevated more than thirty-fold and peripartum sepsis more than twelve-fold, and crises severe enough to require delivery carried the highest risk of all, though even crises that resolved without delivery still conferred around a seven-fold excess. Preterm birth and caesarean delivery were also more frequent. The message is that a crisis history is not merely a marker of disease severity to be noted, it should trigger genuinely high-acuity multidisciplinary obstetric planning, with a low threshold for sepsis evaluation and close renal monitoring peripartum. Alongside that, Al-Agha and colleagues followed 195 Saudi patients with sickle cell anaemia or sickle beta-zero thalassaemia for a median of thirteen and a half years, covering over two thousand patient-years [9]. They observed fourteen strokes, thirteen of them ischaemic, at a median age of 28 years, and notably no overt stroke occurred before age ten, with incidence rising steadily into the fourth decade. By age thirty, cumulative incidence of overt ischaemic stroke was five percent while silent cerebral infarcts reached nearly nineteen percent. Higher baseline lactate dehydrogenase predicted ischaemic stroke, and higher fetal haemoglobin showed a protective trend that did not reach statistical significance. This cohort had high baseline fetal haemoglobin and hydroxyurea use that rose from under half to eighty-five percent over follow-up, so the absent childhood stroke peak may reflect a genuinely different phenotype rather than a universal truth, but it is a reminder that stroke surveillance in sickle cell disease cannot stop at adolescence, and that silent infarcts remain the dominant, under-detected burden.

Rounding out the anaemia theme, Annals of Internal Medicine published a synopsis of the KDIGO 2026 guideline on anaemia in chronic kidney disease, the first update since 2012 [4]. Eight graded recommendations are distilled here. Transferrin saturation and ferritin thresholds guide iron therapy, with intravenous iron preferred over oral in haemodialysis patients. Correctable causes should be addressed before reaching for an erythropoiesis-stimulating agent or a hypoxia-inducible factor prolyl hydroxylase inhibitor, and when pharmacotherapy is needed, the work group recommends an ESA rather than a HIF stabiliser as first-line, which is a pointed statement given how aggressively the oral agents have been marketed. Haemoglobin thresholds for starting an ESA should be individualised to symptoms and patient values, and the upper haemoglobin target during maintenance stays below 11.5 grams per decilitre. Forty-nine additional practice points cover areas where the evidence was too thin for formal grading.

Our second theme is immune-based therapy pushing into earlier disease and into harder targets. In Nature Medicine, Nadeem and colleagues report ImmunoPRISM, a randomised phase 2 trial of fixed-duration teclistamab, the BCMA-directed bispecific, versus lenalidomide and dexamethasone in high-risk smouldering myeloma [3]. Fifty-nine patients were treated, forty-five with teclistamab and fourteen with the doublet. The complete response rate was just under eighty percent with teclistamab and zero with lenalidomide-dexamethasone, with measurable residual disease negativity at ten to the minus five in about eighty-two percent of teclistamab-treated patients. At a median follow-up of two years, progression-free survival was 92 percent with teclistamab versus 49 percent with the doublet. Cytokine release syndrome was grade one to two only, there was no neurotoxicity, grade three infections were no more frequent than with the doublet, and no deaths occurred in either arm. This is immune interception, treating an asymptomatic precursor state with a potent T-cell engager, and the depth of response in a small randomised phase 2 is genuinely arresting. The caution is that we do not yet know whether eradicating residual disease in smouldering myeloma translates into cure or into overall survival, and the control arm here is a regimen many would consider modest. This is not yet standard of care, but it defines where the field is heading and it is a strong argument for trial referral.

At the other end of the difficulty spectrum, Leukemia reports a phase 1 trial from Guru Murthy and colleagues of XmAb18968, a CD38-directed CD3 bispecific T-cell engager with an Fc domain modified to limit non-selective effector activation, in relapsed or refractory acute myeloid leukaemia and T-cell acute lymphoblastic leukaemia [10]. Twenty-two heavily pretreated patients, median age 63 with a median of three prior lines, received therapy. Tolerability was the headline, with no grade three or higher cytokine release syndrome or neurotoxicity and grade three cytopenias in the mid-teens. Efficacy was modest but real, with two MRD-negative complete remissions and one partial remission among eleven evaluable AML patients, and meaningful disease burden reduction in four of six patients with T-ALL. Median overall survival was around eight and a half months in both groups. In a population with essentially no options, a safe and active CD38-directed engager is a legitimate signal worth following. Also in the leukaemia space, the British Journal of Haematology published a retrospective comparison from Gruber and colleagues of reduced-dose 8 Gray versus standard 12 Gray total body irradiation before allogeneic transplant in 101 adults with ALL [7]. Despite the 8 Gray cohort being older, sicker and higher risk, unadjusted outcomes were no different, and in multivariable models 12 Gray was associated with roughly a six-fold higher non-relapse mortality with no evidence of reduced relapse. The overall survival disadvantage did not survive propensity overlap weighting, and pre-transplant measurable residual disease data were missing, so this is hypothesis-generating rather than practice-changing, but it makes a prospective trial of dose de-escalation look overdue. And for the rarer end of the spectrum, the American Journal of Hematology carries the 2026 hairy cell leukaemia update from Troussard and colleagues, reaffirming purine analogues first line, increasing use of cladribine with rituximab, and BRAF inhibitors, MEK inhibitors, BTK inhibitors and BCL-2 inhibitors in relapsed disease, with optimal sequencing still unresolved [6].

Our third theme is protein replacement, where two papers show how much of modern haemostasis care is now about convenience and burden rather than raw efficacy. In Blood, Scully and colleagues report the final analysis of the randomised phase 3 crossover trial of recombinant ADAMTS13 in congenital thrombotic thrombocytopenic purpura, with 48 participants aged three to sixty-eight crossing between six months of recombinant enzyme prophylaxis and six months of plasma-based therapy [2]. No participant had an acute TTP event on recombinant ADAMTS13 versus one on plasma-based therapy, subacute events and thrombocytopenia were both less frequent on the recombinant product, and the safety separation was dramatic, with treatment-related adverse events in about four percent of participants on recombinant enzyme versus close to half OF PARTICIPANTS on plasma. No neutralising antibodies were detected, activity rose roughly six-fold, and treatment satisfaction favoured the recombinant product. Complementing that, the Journal of Thrombosis and Haemostasis reports FRONTIER5 from Hermans and colleagues, a single-arm phase 3b study in which 61 patients with haemophilia A, with or without inhibitors, switched directly from emicizumab to denecimig, a factor VIII-mimetic bispecific antibody, with no washout and no loading dose [5]. All 61 completed the study. Treatment-emergent adverse events were mostly mild or moderate and largely unrelated to treatment, there were no thromboembolic events, no hypersensitivity reactions and no clinical evidence of neutralising antibodies, injection site reactions occurred in about one and a half percent of injections, and thrombin generation moved into the normal reference range within a week without an exaggerated haemostatic response. Nearly all patients rated the pen injector easy to use. The reassuring operational point for clinicians is that a direct switch appears feasible, with dosing options extending to monthly.

If you only have time for one paper this week, make it ENERGIZE-T in The Lancet [1]. It delivers the first oral disease-modifying option for transfusion-dependent thalassaemia, including alpha thalassaemia where we have had essentially nothing, and that changes the conversation you can have in clinic starting now.

Here are the key takeaways from this week in Hematology. Mitapivat halved transfusion burden in roughly a third of adults with transfusion-dependent thalassaemia, making it a reasonable trial in selected patients while acknowledging most will not respond. A crisis history in sickle cell anaemia marks pregnancies at roughly eight-fold risk of severe maternal morbidity, so those women need high-acuity multidisciplinary care with vigilance for sepsis and renal failure. Stroke risk in sickle cell disease does not end in childhood, and silent cerebral infarcts remain the dominant burden into the third decade. The new KDIGO anaemia guideline favours an erythropoiesis-stimulating agent over a HIF stabiliser first line, with maintenance haemoglobin kept below 11.5 grams per decilitre. And in congenital TTP, recombinant ADAMTS13 prophylaxis now has final phase 3 data showing prevention of acute events with a markedly cleaner adverse event profile than plasma.

That's your roundup for This Week in Hematology. The full transcript and references are available on the episode page in your AudioScholar library. This is an AI-curated summary — for clinical decisions, always consult primary sources and current guidelines. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And a quick rating on Apple Podcasts or Spotify helps other physicians discover the show.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.