ASH 2025 | Optimal ferritin threshold to diagnose IEC-HS following CAR T-cell therapy in R/R myeloma

So this study looked at the optimal ferritin threshold to diagnose immune effector cell-associated hemophagocytic syndrome, IEC-HS, following CAR-T therapy in relapsed refractory myeloma. So IEC-HS is a life-threatening inflammatory toxicity following CAR T-cell therapy, and in a previous abstract that was published at ASH, we described a 5% provider-delineated IEC-HS diagnosis in 1,502 relapsed/refractory myeloma patients. And it was associated with a worse progression-free survival and overall survival. 

So timely diagnosis is very important to improving patient outcomes when it comes to IEC-HS. But the challenge is that IEC-HS may overlap with CRS, ICANS, infection, or tumor-driven inflammation. So it’s unclear whether the timing of IEC-HS is going to be different depending on the CAR-T product or the underlying tumor type and whether or not there will be overlap with some of these other inflammatory conditions. So real-time biomarkers are important and could help identify IEC-HS cases to help clinicians more closely monitor and start immunosuppression. 

So this slide outlines what our objectives of this study were. So we wanted patients were first identified by their home physician as having IEC-HS or not. And so we had 74 patients whose physicians had said that they felt like their clinical picture was consistent with IEC-HS and 1,428 patients where the provider did not feel that they had a diagnosis of IEC-HS. So the first objective was really simple. We just wanted to answer the question, in the provider adjudicated cases of IEC-HS, was there a ferritin threshold that could discriminate patients with IEC-HS versus without IEC-HS? The second objective was to compare baseline clinical and disease characteristics in patients with a ferritin value at or above the threshold versus below the threshold. 

So for exploratory objectives, we wanted to focus on subset analyses within the full patient cohort. So we evaluated whether peak ferritin and the timing of IEC-HS diagnosis occurred together, which we were able to assess in about 64 patients who had both dates available. And the second, we examined whether the ASTCT IEC-HS criteria in patients with ferritin levels above 5,000 nanograms per milliliter, whether they had those criteria. So we had detailed criteria for about 134 patients, regardless of whether they were diagnosed with IEC-HS or not. 

So to determine the optimal ferritin threshold, we use the receiver operating characteristic analysis. We calculated the area under the curve overall and separately for cilta-cel and ide-cel. AUCs above 0.9 indicate excellent discriminative performance. So this slide shows that the ROC analysis used to identify the optimal ferritin cut point for IEC-HS. And for cilta-cel, this optimal threshold was 7,540 with an AUC of 0.931, indicating excellent discrimination. For ide-cel, the threshold was nearly identical at 7,470 with an even higher AUC of 0.956. And across both products, ferritin levels around 7,500 provided strong sensitivity and specificity for physician-determined IEC-HS. 

So this slide shows our comparison of baseline characteristics between patients with ferritin above versus below the threshold of 7,500. And so you can see that patients in the high ferritin group were slightly younger, more often male. They were also more likely to have performance status of ECOG-2 or higher at apheresis and had significantly higher baseline ferritin at the start. We also saw higher rates of penta-refractory disease and greater use of bridging therapy in the high ferritin group. Finally, these patients had much higher baseline plasma cell percentages in the bone marrow. 

So we looked at the timing of physician-identified IEC-HS relative to peak ferritin, and in the subgroup of 64 patients who had both dates available, about 87% had their peak ferritin occur concurrently with IEC-HS diagnosis. And the median peak ferritin among the IEC-HS cases were very high at 25,800 with a wide range, and only three cases had peak ferritins below 7,500. So the table at the bottom shows that both IEC-HS onset and peak ferritin occurred around day nine after infusion. 

And finally, in our last analysis, we examined the Hines et al. criteria in a subset of patients to really understand what we were seeing in terms of their prevalence in this patient population. And the slide highlights some of the key elements of the ASTCT IEC-HS criteria. So it really entails the development of a hyperinflammatory syndrome that is independent from CRS and ICANS and which has features of macrophage activation. It has to be attributable to IAC therapy, and it is associated with progression or nuance of cytopenias, high ferritin, coagulopathy, and or transaminitis. So when stratified by peak ferritin, patients with ferritin above 7,500 demonstrated higher frequencies of several of the ASTCT IEC-HS criteria compared to those in the category of 5,000 to 7,500. And except for cytopenias and fever, which were common in both groups, all other dimensions reflected a notable difference across groups. The slide shows the co-occurrence of Hines criteria in patients with the ferritin below the 7,500, and what’s really striking here is that there’s really very little clustering that we see. In contrast, above that 7,500 threshold, we see extensive clustering, and multiple criteria are frequently occurring together, including neurotoxicity, low fibrinogen, pulmonary manifestations, and renal insufficiency. So this really demonstrates that once ferritin arises above a certain threshold, patients accumulate multiple IEC-HS criteria simultaneously, reflecting a much more systemic inflammatory syndrome. 

So to summarize, using provider-adjudicated IEC-HS diagnoses, we identified 7,500 nanograms per milliliter as the optimal real-time ferritin threshold for both cilta-cel and ide-cel, and ferritin at or above this level aligns closely with our clinical diagnosis timing, it strongly associates with IEC-HS criteria, and it reflects distinct baseline characteristics. This really suggests that ferritin may serve as a very simple, accessible biomarker that clinicians have access to for earlier identification of IEC-HS. 

There were limitations to the study. It is a retrospective design where the diagnoses of IEC-HS were provider adjudicated. There’s also inter-provider and center variability and how they diagnosed IEC-HS. And in addition, providers may have used ferritin as part of their diagnostic reasoning, although the Hines et al. criteria does not outline an optimal threshold for diagnosis. The other limitations include the fact that the timing of the max ferritin IEC-HS diagnosis and detailed Hines criteria were only really available in a subset of patients. So looking ahead, we really want to expand this investigation. So the 7,500 threshold, we are going to look at a prospective validation cohort, and in future evaluations, we hope to understand if the 7,500 nanograms per ml can be applied to other CAR-T products and also in other disease states.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.