Clonal hematopoiesis and CAR T-cell therapy outcomes

Clonal hematopoiesis (CH) is defined as the clonal expansion of hematopoietic stem cells, resulting in blood cells with the same genetic mutations. CH of indeterminate potential (CHIP) refers to this process in the presence of leukemia-associated mutations with a variant allele frequency (VAF) of ≥2% and has been found to be associated with increased mortality in both patients with cancer and healthy individuals. CHIP is common in patients with non-Hodgkin lymphoma (NHL) and multiple myeloma (MM) and is associated with worse outcomes and reduced mortality.¹

Chimeric antigen receptor (CAR) T-cell therapy, although highly effective for patients with relapsed or refractory lymphoid malignancies, results in frequent complications owing to cytokine release syndrome (CRS). Miller et al. recently published their study in Blood Advances, which investigated CHIP in CAR T-cell recipients to determine whether there was an association with poorer outcomes.¹ They hypothesized that CHIP mutations may influence outcomes in patients receiving CAR T-cell therapy due to their ability to enhance inflammatory signaling via interleukin-6 (IL-6), a known mediator of CRS. Also, DNMT3A and TET2 genes influence CAR T-cell programs and are known to be frequently mutated in CHIP.

Study design and patient characteristics

This retrospective study included a cohort of patients with NHL or MM who received CAR T-cell treatment between 2016 and time of study start, and for whom blood samples were available. Next-generation sequencing (NGS) on a targeted gene panel was preferentially performed on blood samples that had been collected the day of, within the 3 months before, or as close as possible to CAR T-cell infusion. Primary endpoints were progression-free survival (PFS) and overall survival (OS), and secondary endpoints were complete response (CR) rate and CRS Grade ≥2.

There were 154 patients with available blood specimens for genetic analysis with a mean age of 63 years. The cohort were 41% female with 30% having received hematopoietic stem cell transplant (HSCT) previously (Table 1).

Table 1. Patient characteristics of cohort*

CRS, cytokine release syndrome. *Adapted from Miller et al.1
Characteristic, % (unless otherwise stated)	Whole cohort (N = 154)
Diagnosis, n
Non-Hodgkin lymphoma	144
Multiple myeloma	10
Median age (range)	63 (24−83)
Female gender	41
Best overall response
Complete response	65
Partial remission	21
Stable disease	3
Progressive disease	11
Prior lines of treatment
1−2	46
3−5	44
6−10	10
Prior hematopoietic transplant
Autologous	27
Allogeneic	3
CRS grade
0	14
1−2	74
3−4	10
5	1

Results

CH was identified in 76% of patients with a VAF ≥0.004 and was found to be associated with age; 85% of patients >70 years of age had CH with VAF >0.01. Patients with CH were also found to have a median of two mutations most commonly occurring in PPM1D, DNMT3A, TP53, TET2, and SRCAP.

To determine whether there was an association with CRS and CHIP, Miller, et al. compared CH status of VAF <0.02 with CH status of VAF ≥0.02 (CHIP) in terms of CRS grade across the whole cohort and in NHL patients only (Table 2). This analysis revealed a significant association between CHIP status and Grade ≥2 CRS in younger patients (<60 years old) in both the whole cohort (p = 0.032) and NHL patients (p = 0.042).

Table 2. Association between CH and CRS in the whole cohort and in NHL patients*

CH, clonal hematopoiesis; CRS, cytokine release syndrome; NHL, non-Hodgkin lymphoma; VAF, variable allele frequency. *Adapted from Miller et al.1 Bold font indicates statistically significant p values.
Characteristic	CH status	n	% CRS <2	% CRS ≥2	p value
Whole cohort	No CHIP	80	40.0	60.0	0.5
CHIP	74	33.8	66.2
Age < 60	No CHIP	40	52.5	47.5	0.032
CHIP	22	22.7	77.3
Age ≥ 60	No CHIP	40	27.5	72.5	0.37
CHIP	52	38.5	61.5
NHL Patients	No CHIP	76	39.5	60.5	0.49
CHIP	68	33.8	66.2
NHL patients aged < 60	No CHIP	37	54.1	45.9	0.042
CHIP	18	22.2	77.8
NHL patients aged ≥ 60	No CHIP	39	25.6	74.4	0.26
CHIP	50	38.0	62.0

In NHL patients only, Miller, et al. assessed whether CR was associated with CH status. They found that patients with CHIP (VAF ≥0.02) were more likely to achieve CR (p = 0.013; Table 3), but this was only true for those patients aged <60 years.

Table 3. Association between CH and CR in the whole cohort and in NHL patients*

CH, clonal hematopoiesis; CR, complete response, non-Hodgkin lymphoma; VAF, variable allele frequency. *Adapted from Miller et al. †One patient could not be evaluated for CR. Bold font indicates statistically significant p values.
Characteristic	CH status	n	% CR	% No CR	p value
NHL Patients	No CHIP	76	57.9	42.1	0.13
CHIP	67†	77.6	22.4
NHL patients aged < 60	No CHIP	37	48.6	51.4	0.0067
CHIP	18	88.9	11.1
NHL patients aged ≥ 60	No CHIP	39	66.7	33.3	0.64
CHIP	49†	73.5	26.5

Despite the differences observed between CR and CRS in relation to CHIP in these patients, there was no association between CH status and PFS, with patients with CHIP having a PFS of 25.6 months (95% confidence interval [CI], 10.1–infinity) vs 21.5 months (4.5–infinity; p = 0.6). There was also no association between CHIP and OS (median not achieved in either group; p = 0.6), and this did not change when stratifying the outcomes by age.

The study did find a higher prevalence of CHIP in the study cohort than in patients treated for solid tumors (48% vs 30%; p = 0.001), or in patients undergoing autologous-HSCT for NHL (48% vs 30%; p = 0.0001) or multiple myeloma (48% vs 14%; p = 0.001).

Conclusion

Miller et al. concluded that the data from their study highlight that hematopoietic mutations can affect inflammatory pathways, but that further work is needed to understand these mechanisms and how they impact resistance or loss of CAR T-cell activity. This work supports an association between CHIP and CR and CRS but found that it did not impact longer term outcomes. Therefore, CHIP should not preclude patients from CAR T-cell therapy.