Impact of salvage treatment phase on inotuzumab ozogamicin treatment for relapsed/refractory acute lymphoblastic leukemia: an update from the INO-VATE final study database

Adults with relapsed or refractory acute lymphoblastic leukemia (R/R ALL) have a poor prognosis. The median overall survival (OS) is 3–8 months with standard of care chemotherapy (SC) [1–3], with more favorable outcomes among first salvage (S1) patients with a longer duration of first complete remission (CR1) [3–5]. Inotuzumab ozogamicin (InO), a calicheamicin-conjugated CD22 antibody, is approved for R/R B-cell ALL treatment. In the phase 3 INO-VATE study (NCT01564784), patients in the InO vs. SC treatment arm (in S1 or second salvage [S2]) had a higher remission rate and longer OS [6,7]. Our analysis aimed to provide an updated, in-depth analysis of INO-VATE outcomes by salvage phase, including secondary endpoints, safety outcomes, and 24-month follow-up data from the final study database. We also analyzed a subgroup of S1 patients with an extended CR1 (≥12 months). This analysis could inform clinicians about optimizing InO use, maximizing its benefits, and ultimately improving outcomes for R/R ALL patients in S1 or later salvage (S2+).

Methods and study design for the global, open-label, randomized, phase 3 INO-VATE trial were previously described [6,7]. Patients (N = 326) aged ≥18 years with CD22-positive, Philadelphia chromosome-positive (Ph+) or negative R/R ALL in S1 or S2 were randomized 1:1 to InO (n = 164) or SC (n = 162) treatment (stratification factors, dosing regimens, and populations in Supplemental Methods). We report the primary endpoints (OS and complete remission [CR]/CR with incomplete hematologic recovery [CRi]) and secondary outcomes, including hematopoietic stem cell transplantation (HSCT) rate, minimal residual disease (MRD) negativity rate, duration of remission (DoR), progression-free survival (PFS), and safety outcomes – according to salvage treatment phase (S1 vs. S2) up to last patient/last visit (4 January 2017). We also present an analysis of S1 patients with a pre-study extended CR1 (≥12 months).

Baseline characteristics were generally balanced between InO and SC for S1 and S2 patients (Supplemental Table 1), and between InO and SC for S1 patients with an extended CR1. In both arms, a higher proportion of S2 vs. S1 patients had received prior HSCT and were Ph+.

CR/CRi rates and OS were higher with InO vs. SC for S1 and S2 patients; these improvements were more pronounced in S1. CR/CRi rates were higher in the InO vs. SC arm (S1: p < .0001; S2: p = .0012), with a greater between-arm difference in S1 vs. S2 (Table 1). Similarly, 24-month OS probabilities and OS hazard ratios (HRs) show more pronounced improvements with InO in S1 vs. S2 (Table 1, Figure 1(A)). These results confirm and extend previously reported improved outcomes with InO [6,7], with our analysis suggesting a greater relative benefit of InO in S1. Despite this, the benefit of InO on remission rate in S2 is clinically meaningful, particularly since >30% of S2 patients in both treatment arms had received prior HSCT (Supplemental Table 1). Outcomes were generally improved among S1 vs. S2 patients in both treatment arms (Table 1). These data may suggest that patients would benefit from earlier treatment with the most efficacious agents available, and may therefore support the use of InO in S1. Compared with prior literature, SC arm CR/CRi rates were higher in S2 (prior literature vs. our study: 21% [95% confidence interval: 16–26%] vs. 35.6% [23.6–49.1%]) and lower in S1 (40% [37–44%] vs. 28.4% [19.9–38.2%]) [4]. The increased SC-arm CR/CRi rate was not expected in S2 vs. S1, and may be due to limited sample size or baseline differences. Despite these limitations, InO showed consistent benefits compared with SC for both S1 and S2 patients.

Figure 1. OS and PFS by salvage treatment phase. (A) OS and (B) PFS. Salvage treatment phase by clinical report form, intent-to-treat population. CI: confidence interval; InO: inotuzumab ozogamicin; m: median; mo: months; OS: overall survival; PFS: progression-free survival; NE: not evaluable; S1: salvage 1; S2: salvage 2; SC: standard of care chemotherapy.

Table 1. Responses and outcomes stratified by salvage treatment phase (ITT population).

	Salvage 1		Salvage 2
Parameter	InO (n = 111)	SC (n = 102)	InO (n = 51)	SC (n = 59)
CR/CRi, n (%) [95% CI]	87 (78.4) [69.6–85.6]	29 (28.4) [19.9–38.2]	33 (64.7) [50.1–77.6]	21 (35.6) [23.6–49.1]
CR, n (%) [95% CI]	45 (40.5) [31.3–50.3]	17 (16.7) [10.0–25.3]	10 (19.6) [9.8–33.1]	9 (15.3) [7.2–27.0]
CRi, n (%) [95% CI]	42 (37.8) [28.8–47.5]	12 (11.8) [6.2–19.6]	23 (45.1) [31.1–59.7]	12 (20.3) [11.0–32.8]
MRD negativity in pts with CR/CRi, n/N (%) [95% CI]	67/87 (77.0) [66.8–85.4]	10/29 (34.5) [17.9–54.3]	24/33 (72.7) [54.5–86.7]	9/21 (42.9) [21.8–66.0]
Median OS (95% CI), mo	8.6 (6.1–10.3)	6.9 (4.7–8.8)	7.0 (4.2–9.2)	5.2 (3.6–9.1)
OS HR (97.5% CI), p value	0.75 (0.53–1.07), .0350		0.92 (0.57–1.49), .3525
Probability of OS at 24 mo (95% CI), %	25.9 (18.1–34.4)	8.9 (4.1–15.8)	16.8 (8.0–28.4)	12.2 (5.0–22.8)
Median PFS (95% CI), mo	5.4 (4.5–6.3)	1.8 (1.5–2.4)	3.9 (2.7–5.1)	1.4 (1.2–2.4)
PFS HR (97.5% CI), p value	0.44 (0.31–0.64), <.0001		0.49 (0.30–0.81), .0006
Subsequent HSCT at any time, n (%) [95% CI]	58 (52.3) [42.6–61.8]	27 (26.5) [18.2–36.1]	20 (39.2) [25.8–53.9]	9 (15.3) [7.2–27.0]

Salvage treatment phase was according to the clinical report form.

CI: confidence interval; CR: complete remission; CRi: complete remission with incomplete hematologic recovery; HR: hazard ratio; HSCT: hematopoietic stem cell transplantation; InO: inotuzumab ozogamicin; ITT: intent-to-treat; mo: months; MRD: minimal residual disease; OS: overall survival; PFS: progression-free survival; pts: patients; SC: standard of care chemotherapy.

Outcomes were assessed in S1 patients with a pre-study extended CR1 (≥12 months, N = 88, n = 48 [InO], n = 40 [SC]). InO- vs. SC-arm patients in this subgroup showed an improved CR/CRi rate (85.4% [41/48] vs. 27.5% [11/40], p < .0001) and improved OS (HR 0.55 [p = .0086]; Supplemental Figure 1). Compared with previous reports for InO-arm patients in S2 and/or with a shorter CR1 (<12 months), this remission rate is numerically higher [6,7]. Although patient numbers were small, this subgroup analysis indicates that, as with SC, CR1 duration is a positive prognostic factor for InO-arm patients. Overall, our analyses demonstrate that InO provides significant clinical benefit both in S1 and S2, which may be maximized when used in S1, including patients with an extended CR1.

Secondary outcomes, including MRD negativity, HSCT rate, PFS, and DoR, further demonstrate the benefits of InO in S1 and S2. MRD negativity is a key prognostic factor for OS, event-free survival, and post-transplant outcomes, including in the salvage-therapy setting [8,9]. Among InO-arm patients who achieved CR/CRi, MRD negativity rates were >70% and higher than SC-arm patients in S1 and S2 (Table 1). MRD negativity rates were also higher in the InO vs. SC arm in S1 patients with an extended CR1: 90.2% (37/41) vs. 45.5% (5/11), p = .0032, further supporting the prognostic significance of CR1 duration among InO-arm patients. HSCT rate was higher in the InO vs. SC arm (S1: p < .0001; S2: p = .0022) and numerically higher for both arms in S1 vs. S2 (Table 1). This is clinically relevant, especially for older adults, since HSCT remains the established, potentially curative R/R ALL treatment. Along with improved MRD negativity rates, increased HSCT rates may have contributed to the improved survival outcomes seen, particularly at later time points, with InO in S1 treatment. InO-arm post-transplant survival did not appear to depend upon salvage treatment phase; however, outcomes were improved among those proceeding directly to follow-up HSCT (i.e. without receiving additional therapy) after achieving CR/CRi (Supplemental Figure 2). PFS was significantly improved in the InO vs. SC arm in S1 and S2 (Table 1 and Figure 1(B)), while DoR was significantly improved with InO vs. SC in S2 (S1: HR 0.736 [p = .1213], median 5.8 vs. 5.2 months, S2: HR 0.518 [p = .0199], median 4.2 vs. 3.0 months). Another novel compound that shows superiority over chemotherapy in S1 and S2+ is blinatumomab [10]. With blinatumomab vs. SC, OS was significantly improved in S1 but not S2+, and remission rates and MRD response were significantly improved in S2+, but not S1 [10]. In the blinatumomab study, chemotherapy arm remission rates were closer to historical expectations (S1: 36.5%; S2+: 14.1%), with lower remission rates in S2+ that may reflect inclusion of later lines of therapy. The blinatumomab study differs from our study in including later lines of therapy, excluding Ph+ patients, and in having a generally younger patient population (median age: 37 vs. 47 years in the blinatumomab study [11] vs. our study). This is important because salvage status, Ph+ status, and age are prognostic factors in adult R/R ALL [3,12,13]. Overall, InO treatment provided significant clinical benefit for S1 and S2 patients compared with SC, with potentially greater improvements among those with an extended CR1.

The incidence of treatment-emergent adverse events (TEAEs) in the safety population was similar for InO vs. SC, although there was an increased veno-occlusive disease (VOD) risk with InO, especially in S2. For InO- vs. SC-treated patients, grade ≥3 TEAE incidence was 91.0% vs. 95.7% (S1) and 90.2% vs. 98.0% (S2). The most common grade ≥3 TEAEs were blood and lymphatic system disorders followed by infections and infestations, the incidence of which appeared lower with InO vs. SC for S1 and S2 (Supplemental Table 2). With InO, the VOD incidence appeared greater in S2 vs. S1, either during study treatment/follow-up without HSCT (5.9% [3/51] vs. 1.8% [2/111]) or following HSCT (35.0% [7/20] vs. 19.0% [11/58]). The post-HSCT VOD rate in S2 vs. S1 with SC was 11.1% (1/9) vs. 7.7% (2/26). These findings are consistent with previous literature showing that VOD most commonly occurs following HSCT [14,15], and may have been influenced by the higher rate of prior HSCT in S2 vs. S1 patients in both treatment arms. These findings reinforce the need for careful patient monitoring post-HSCT [14,15]. Similar TEAE trends were seen among InO- vs. SC-treated S1 patients with an extended CR1 (grade ≥3 blood and lymphatic system disorders: 83.3% vs. 91.9%; infections and infestations: 16.7% vs. 56.8%; post-HSCT VOD: 14.3% [4/28] vs. 0), suggesting that CR1 duration does not significantly alter the safety profile of InO. Overall, InO treatment had an acceptable safety profile for patients in either S1 or S2, although S1 patients had a reduced VOD risk.

Overall, this updated analysis of INO-VATE confirms and extends previous reports of improved outcomes with InO vs. SC in S1 and S2. This analysis suggests that the benefits of InO vs. SC may be greater in S1 compared with S2, with InO-arm patients in S1 showing more pronounced improvements in remission rates, long-term survival, HSCT rates, and PFS, along with a possible reduction in VOD incidence. Furthermore, both salvage status and first remission duration appeared to be prognostic factors in InO-arm patients. Utilizing InO in S1 may therefore maximize the chances of a successful treatment outcome in R/R ALL.

Acknowledgements

This study was sponsored by Pfizer. Medical writing support was provided by Susan Tan, PhD, of Engage Scientific Solutions, and was funded by Pfizer. The first draft of this manuscript was based on poster PS950 presented at the 24th Congress of the European Hematology Association (EHA), 2019, and poster 7029 presented at the American Society of Clinical Oncology (ASCO) Annual Meeting, 2019.

Disclosure statement

EJ has received research grants and consultancy fees from AbbVie, Adaptive Biotechnologies, Amgen, Bristol-Myers Squibb, Pfizer, and Takeda. MS declares research support and honoraria from Pfizer. ASA declares research funding and honoraria from Pfizer, research funding from Amgen, and honoraria from AbbVie and Kite Pharmaceuticals. DJD declares honoraria from Pfizer. NG declares research support and honoraria from Amgen, Novartis, and Pfizer. DIM declares consultancy with Pfizer. WS declares research funding and honoraria from Pfizer, honoraria from Research to Practice and Up to Date, and has served on advisory boards for Agios, Astellas, and Kite Pharmaceuticals. SO declares research support from Acerta, Kite Pharmaceuticals, and Regeneron; research support and consultancy from Gilead, Pfizer, Pharmacyclics, Sunesis, and TG Therapeutics; and consultancy from AbbVie, Alexion, Amgen, Aptose Biosciences Inc., Astellas, Celgene, Eisai, GlaxoSmithKline, Janssen Oncology, Vaniam Group LLC, and Verastem. RDC has received research support from Amgen, Incyte, Kite/Gilead, Merck, Pfizer, and Vanda Pharmaceuticals; has received personal fees from Amgen and Pfizer; and has a spouse who is an employee of Seattle Genetics. TW, AN, and EV are employees of and own stock in Pfizer. HK has received research support and honoraria from AbbVie, Agios, Amgen, Immunogen, and Pfizer; has received research support from Ariad, Astex, Bristol-Myers Squibb, Cyclacel, Daiichi-Sankyo, Jazz Pharmaceuticals, and Novartis; has served on advisory boards for Actinium; and has received honoraria from Takeda.

Data availability statement

Upon request, and subject to certain criteria, conditions and exceptions (see https://www.pfizer.com/science/clinical-trials/trial-data-and-results for more information), Pfizer will provide access to individual deidentified participant data from Pfizer-sponsored global interventional clinical studies conducted for medicines, vaccines, and medical devices (1) for indications that have been approved in the United States and/or European Union, or (2) in programs that have been terminated (i.e. development for all indications has been discontinued). Pfizer will also consider requests for the protocol, data dictionary, and statistical analysis plan. Data may be requested from Pfizer trials 24 months after study completion. The deidentified participant data will be made available to researchers whose proposals meet the research criteria and other conditions, and for which an exception does not apply, via a secure portal. To gain access, data requestors must enter into a data access agreement with Pfizer.