We report on the impact of MATV and MATV-dynamics during the bridging period upon CAR T-cell treatment in patients with r/r LBCL who were treated with axi-cel at a single institution.

In this analysis, patients with a low baseline MATV had the best TTP and OS. In addition, patients with an initial high baseline MATV, but in which tumor volume could be reduced or controlled during the bridging period, had a better outcome compared to patients for whom bridging therapy did not lead to tumor reduction.

Nowadays, in real-world, a significant number of patients receive bridging therapy prior to CAR T-cell infusion [12,13,14]. Patients responding to bridging therapy, regardless of the applied regimen, had a reduced risk of progression after CAR T-cell infusion [4]. In addition, others have found that a higher MATV determined prior to CAR T-cell infusion is associated with disease progression after CAR T-cell therapy [11, 24,25,26]. Breen et al. discovered that especially an increased tumor volume from apheresis to pre-LD is associated with an increased risk of death [11]. These results align with our findings that effective reduction or control of tumor volume results in an improved survival.

Of note, it is difficult to distinguish the effect of disease reduction from underlying differences in disease biology to establish these results. Patients with a high baseline MATV that showed effective control of tumor volume during the bridging period could represent a subset of patients that is more sensitive to bridging therapy and CAR T-cell therapy, caused by a preferable, less aggressive disease biology. However, in our analysis we did not find any significant differences in the distribution of well-known markers of aggressive disease, such as lymphoma histology, advanced stage, a high number of extranodal sites and a high IPI-score between these subgroups. Moreover, no significant differences were observed in the percentage of patients that were primary refractory to first and second line immunochemotherapy between the different subgroups. Altogether, despite the unknown molecular composition of the tumor itself, these data suggest that disease reduction in patients with a high disease burden at baseline has a positive effect on CAR T-cell outcome, pointing to an opportunity for further exploration of the implementation of more effective (aggressive) bridging regimens.

Furthermore, in our cohort radiotherapy was chosen in the majority of patients in need of bridging therapy. Radiotherapy was highly effective as a single-modality bridging regimen in controlling MATV in patients with a low-risk MATV at baseline. In addition, patients bridged with radiotherapy were most likely to shift from the high-risk MATV at baseline to the low-risk group at pre-LD. Both of these findings are in line with the previously found advantages of radiotherapy bridging for patients with a high tumor load [27]. This highlights the potential of radiotherapy as a bridging strategy in this heavily pretreated patient group.

In our cohort, baseline and pre-LD 18F-FDG PET/CT-scans were conducted with limited deviation in timing from apheresis and infusion. Uniquely, these scans were performed irrespective of the application of bridging therapy or clinical indication for progression. Therefore, we were able to provide MATV-dynamics of both patients who did and did not receive bridging therapy. It is known that patients in need of bridging therapy during this period are more likely to have a high IPI-score, elevated LDH and bulky disease, and have a worse outcome after CAR T-cell infusion compared to the patients without an apparent need for bridging therapy [5]. In alignment, in our cohort patients with a high baseline MATV had indeed a higher LDH (Supplementary Table 2), suggesting more aggressive disease.

Even though most patients who did not undergo bridging therapy showed an increase in MATV, only a few patients transitioned from low-risk at baseline to high-risk at pre-LD (Fig. 2). Moreover, our results showed a clear survival benefit for patients who start with low baseline MATV, as reported previously [5]. This emphasizes the importance of rapid screening, apheresis and treatment if patients have an indication for CAR T-cell therapy.

Moreover, we found that a high MATV baseline (> 190 cc) was associated with a higher risk of CRS grade ≥ 2. Other studies have found similar results for CRS grade ≥ 3 [6, 10]. Due to low incidence of CRS grade ≥ 3 in our cohort, this association was not evaluated. On the contrary, MATV at pre-LD was not associated with CRS grade ≥ 2. Data regarding pre-LD MATV and CRS are conflicting [24, 26]. Similarly, inconsistent findings for the relation of pre-LD MATV with ICANS are reported [11, 26, 28]. We observed no association of both baseline and pre-LD MATV with ICANS grade ≥ 2.

In our cohort, median MATV increased from baseline to pre-LD in general, indicating the tendency for progressive disease in these patients despite the bridging strategies that were applied in the majority of patients. The optimal cut-off values differed between baseline and pre-LD. This can partly be explained by the median increase of MATV from baseline to pre-LD and fast progressive disease in patients not responding to bridging therapy. On the other hand, radiotherapy is most commonly used within our cohort, which is a local treatment and disease progression is expected outside the radiotherapy treatment field. Also, response assessment after radiotherapy with 18F-FDG PET/CT-scan was performed after less than 2 weeks. Normally, response assessment is performed after more than 6 weeks, suggesting that the effect of radiotherapy might be underestimated.

We found optimal cut-off values of 190 cc at baseline and 480 cc at pre-LD. This is substantially higher compared to other studies. The studies of Dean et al., Galtier et al., Hong et al. and Iacoboni et al. reported optimal cut-offs of 147 cc [6], 80 cc [7], 26 cc [10] and 25 cc [9], using MATVs based on all pre-infusion 18F-FDG PET/CT-scans. These differences can be explained by the heterogeneity in used segmentation methods (Syngo-volume-counting-program, manually segmented, the relative SUV threshold 41% SUVmax, and the fixed threshold SUV 2.5 we used). In our recent publication comparing several segmentation methods, the application of the fixed threshold SUV 2.5 resulted in the highest median MATV, while the 41% SUVmax method led to the lowest median MATV, explaining our higher cut-off values [18]. The 41% SUVmax method is recommended by the EANM, but only for higher tumor-to-background values and homogenous tracer uptake [16]. For LBCL, several studies preferred other SUV thresholds due to known heterogeneity of this disease, leading to an underestimation of the tumor volume when 41% SUVmax method is used [19, 20]. This phenomenon has also been described by Dean et al. in the setting of CAR T-cell treatment for LBCL, resulting in their preference of manually segmented analyses above the 41% SUVmax method [6]. To maintain robustness and avoid the exclusion of tumor areas, a fixed threshold of SUV 2.5 was used in our research.

However, validation of our optimal cut-off point for MATV at baseline and pre-LD using the same segmentation method needs to be performed. Nevertheless, our results align with previous studies in terms of the survival benefit of reduced and controlled tumor volume resulting from effective bridging. Ultimately, a standardized and optimal segmentation method for LBCL is needed to reduce these variations between studies.

A limitation of these kind of studies including the current, is that bridging modalities were chosen according to the treating physician and not randomized. This could induce the occurrence of selection bias, as disease location, spread, and tumor volume could influence the choice of bridging regimen and thereby also the MATV dynamic. Additionally, performing detailed analyses on the response to bridging in combination with MATV-dynamics as well as other FDG PET characteristics would be of interest. Especially since Breen et al. already showed that other increased FDG PET characteristics, such as total lesion glycolysis (TLG), from pre-apheresis to pre-LD are associated with worse survival outcomes [11].

In conclusion, we demonstrate that r/r LBCL patients treated with CD19-directed CAR T-cell therapy with low baseline tumor volume measured using MATV, had a better TTP and OS and a lower incidence of severe CRS compared to high volume patients. In addition, effective reducing and controlling tumor volume during the bridging period in patients with a high baseline volume improved survival outcomes, providing rationale for the use of more aggressive bridging therapy regimens.