IMW 2019 | A.R.R.O.W.: Subgroup analysis by frailty score

On Saturday 14^th September, 2019, Professor (Prof.) Maria-Victoria Mateos, University Hospital Salamanca, Salamanca, ES, presented a subgroup analysis from the phase III A.R.R.O.W study at the XVII International Myeloma Workshop (IMW) in Boston, US. The results of this analysis are provided here by the Multiple Myeloma (MM) Hub.¹

Background

The A.R.R.O.W. study evaluated the safety and efficacy of once-weekly versus twice-weekly carfilzomib (K), in combination with dexamethasone (d, Kd), in patients with relapsed/refractory MM (RRMM). Patients with RRMM who had received two to three prior lines of treatment, including exposure to a proteasome inhibitor (not K or oprozomib), and an immunomodulatory drug were enrolled.¹ Patients (n= 478) were randomized (1:1) to receive K once-weekly (n= 240) as a 30-minute intravenous (IV) infusion (70mg/m², Kd70) or twice weekly (n= 238) as a 10-minute IV infusion (27mg/m², Kd27).^1,2 The results from a pre-planned interim analysis of the A.R.R.O.W study were presented at the American Society of Clinical Oncology (ASCO) meeting in 2018 and can be viewed here.²

A subgroup analysis of patients < 65 years vs ≥ 65, showed hazard ratios (HRs) favored Kd70 (HR: 0.6, 95% CI, 0.42–0.86) over Kd27 (HR: 0.84, 95% CI, 0.61–1.15) for progression-free survival (PFS) in both patient groups.³ Aside from age, frailty is an important factor to consider when determining the appropriate treatment of patients with MM. Prof. Mateos and colleagues, therefore, conducted a subgroup analysis which aimed to evaluate efficacy and safety, post-hoc, by frailty status in the A.R.R.O.W. study.¹

Frailty subgroup analysis¹

Prof. Mateos and colleagues used a modified frailty index to explore patient outcomes in the A.R.R.O.W. study. The frailty algorithm considered age, medical history (derived Charlson comorbidity index [CCI]), and Eastern Cooperative Oncology Group (ECOG) score. Patients were evaluated and assigned frailty scores (Table 1), which were balanced between arms.

Table 1. Frailty scores in the Kd70 and Kd27 groups

CCI, Charlson comorbidity index; d, dexamethasone; ECOG, Eastern Cooperative Oncology Group ECOG; K, carfilzomib
	Kd70 (n= 240)	Kd27 (n= 238)
Age group, years, n (%)
< 75	194 (81)	206 (87)
75–80	35 (15)	29 (12)
≥ 80	11 (5)	3 (1)
Modified CCI score, n (%)
≤ 1	124 (52)	138 (58)
> 1	105 (44)	92 (39)
Missing	11 (5)	8 (3)
ECOG performance status, n (%)
0	118 (49)	118 (50)
1	121 (50)	120 (50)
≥ 2	1 (0.4)	0
Frailty score, n (%)
0 (fit)	60 (25)	66 (28)
1 (intermediate)	89 (37)	103 (43)
≥ 2 (frail)	80 (33)	61 (26)
Missing	11 (5)	8         (3)

• PFS analysis:
  • In the total cohort, PFS was improved with Kd70 compared to Kd27²
    • Median PFS (Kd70 vs Kd27): 11.2 vs 7.6 months (HR: 0.69, 95% CI, 0.54–0.83)²
  • This benefit was maintained across the subgroups according to frailty status (Table 2)¹
    • The HR for progression was 0.76 in the frail group, which is similar to the total cohort¹
• ORR was similar in the Kd70 group, irrespective of frailty status (fit vs intermediate vs frail: 67% vs 64% vs 56%), which was in line with the total cohort ORR in the Kd70 arm of 63%¹

Table 2. PFS by frailty score and K-dosing regimen¹

	Fit		Intermediate		Frail
	Kd70	Kd27	Kd70	Kd27	Kd70	Kd27
n	60	66	89	103	80	61
Median PFS (months)	15.7	5.7	11.7	7.7	10.3	6.6
HR (95% CI)	0.53 (0.33–0.86)		0.81 (0.55–1.19)		0.76 (0.49–1.16)
HR, hazard ratio; PFS, progression free survival

Table 3. Best overall response by frailty score¹

CR, complete response; d, dexamehasone; K, carfilzomib; PR, partial response; sCR, stringent complete response; VGPR, Very good partial response
	Fit		Intermediate		Frail
N (%)	Kd70	Kd27	Kd70	Kd27	Kd70	Kd27
n	60	66	89	103	80	61
sCR	1 (2)	0	2 (2)	0	0	0
CR	5 (8)	2 (3)	5 (6)	2 (2)	3 (4)	0
VGPR	18 (30)	4 (6)	23 (26)	14 (14)	20 (25)	9 (15)
PR	16 (27)	13 (20)	27 (30)	33 (32)	22 (28)	16 (26)
Overall response rate, % (95% CI)	67 53–78	29 18–41	64 53–74	48 38–58	56 45–67	41 29–54

• Safety analysis:¹
  • Kd70 was well-tolerated with similar incidences of hematological and non-hematological AEs as in Kd27 arm (Table 4)
  • AEs of interest were low in incidence across all subgroups and K-dosing regimens

Table 4. Safety by frailty status in the safety population¹

d, dexamethasone; K, carfilzomibTEAE, treatment-related adverse events
	Fit		Intermediate		Frail
N (%)	Kd70	Kd27	Kd70	Kd27	Kd70	Kd27
n	60	66	88	101	79	60
Grade ≥ III TEAE	33 (55)	41 (62)	60 (68)	58 (57)	61 (81)	42 (70)
Grade ≥ III TEAEs of interest
Peripheral neuropathy	0	1 (2)	0	0	0	0
Acute renal failure	0	3 (5)	6 (7)	6 (6)	3 (4)	4 (7)
Cardiac failure	1 (2)	1 (2)	3 (3)	3 (3)	3 (4)	5 (8)
Ischemic heart disease	1 (2)	0	0	1 (1)	0	1 (2)
Pulmonary hypertension	0	0	0	0	0	1 (2)
TEAE leading to discontinuation of K	2 (3)	5 (8)	11 (13)	11 (11)	16 (20)	11 (18)

Conclusion

Prof. Mateos concluded that both fit and frail patients experience a clinical benefit with Kd70 (once-weekly dosing of K) compared to Kd27 (twice-weekly). Additionally, no new safety signals were detected by subgroup analysis indicating a similar profile in fit and frail patients as per the general population. It was also noted that grade III or higher TEAEs of interest were uncommon across all subgroups.

This analysis supports the use of once-weekly carfilzomib for both fit and frail patients with RRMM.