Outcomes of Allogeneic Hematopoietic Cell Transplantation after Salvage Therapy with Blinatumomab in Patients with Relapsed/Refractory Acute Lymphoblastic Leukemia

Abstract

Historically, outcomes of adult patients with relapsed acute lymphoblastic leukemia (ALL) who fail to enter remission with conventional chemotherapy are very poor. Blinatumomab, a bispecific CD3/CD19 antibody, has shown remarkable activity in relapsed/refractory (r/r) ALL. Although allogeneic hematopoietic cell transplant (HCT) is the recommended consolidation therapy for patients with r/r ALL who respond to salvage therapy, HCT and toxicity outcomes for those who received blinatumomab salvage and HCT remain largely unknown. We treated 89 patients with r/r ALL with blinatumomab, of whom 43 patients (48%) achieved remission. Here we describe our single-center experience in the subset of patients who responded to blinatumomab salvage therapy for eradication of either gross (n = 24) or minimal residual disease (n = 11) before HCT. Overall survival at 1 and 2 years after allogeneic HCT was 77% and 52%, respectively. Leukemia-free survival at 1 and 2 years were 65% and 40%, respectively. Additionally, with blinatumomab administration pre-HCT, no unusual toxicities such as delayed neutrophil/platelet engraftment or graft failure were observed. Acute grades II to IV graft-versus-host disease (GVHD) at day +100 post-HCT was at 43% and 2-year chronic GVHD was 36%, both comparable with historic control subjects. Finally, results of our subset analysis based on pre-HCT minimal residual disease (MRD) status indicated no significant difference in survival outcomes among patients undergoing transplant in MRD-negative status and the entire cohort. In conclusion, based on results of this study, blinatumomab may be considered as a safe and effective agent for r/r ALL patients before HCT.

INTRODUCTION

With the currently available upfront chemotherapy regimens, most adult patients with B cell acute lymphoblastic leukemia (B-ALL) achieve a complete remission (CR). However, because of high rates of relapse [1–5], especially in high-risk groups, long-term overall survival (OS) for these patients remains low [6,7]. Poor outcomes after standard chemotherapy alone could be explained by disease-related factors, such as increased frequency of high-risk mutations [8] and lack of good-risk cytogenetics [9], and patient-related factors, such as age and comorbid conditions limiting the use of more intensive adolescent/young adult regimens [10,11]. In patients with relapsed ALL who are in their second or subsequent remissions, allogeneic hematopoietic cell transplant (HCT) is the only curative approach for long-term survival [7,12]. However, this treatment is only offered to select patients who are fit enough to undergo allogeneic HCT [13–15]. Allogeneic HCT outcomes depend to a great extent on achievement of morphologic and deep remissions (minimal residual disease [MRD] negative) before HCT, especially in patients with relapsed/refractory (r/r) ALL [16]. Unfortunately, most patients do not attain CR with traditional chemotherapy-based salvage approaches [17]. Therefore, to improve long-term outcomes in patients with r/r ALL and to achieve higher CR rates, novel salvage therapies that can bridge patients to HCT are required.

Blinatumomab is a bispecific T cell engager antibody with dual specificity for CD19 and CD3, which are expressed by B cells and endogenous cytotoxic T cells, respectively [18]. This immunotherapy drug has proven to be safe and effective in patients with r/r ALL and elimination of MRD after induction and consolidation therapy or after HCT [17,18]. In a pivotal phase II study of blinatumomab administration in adult patients with r/r B-ALL, CR was achieved in 43% of patients (33% CR and 10% CR with partial hematologic recovery) within 2 cycles of treatment [19]. In a subsequent phase III trial, treatment of adult patients with r/r B-ALL with blinatumomab resulted in significantly better OS compared with patients who received conventional chemotherapy [20]. In a more recent study by Gokbuget et al. [21], blinatumomab was proven to be highly effective in treatment of adult patients with MRD-positive B-ALL, with 78% of patients achieving a complete MRD response. Blinatumomab has also shown activity in the post-HCT relapse setting, with CR/CR with partial hematologic recovery rates of 45% with 2 cycles of treatment and median relapse-free survival of 7.4 months in responding patients [12]. Finally, results of a phase II study evaluating the efficacy and tolerability of blinatumomab in patients with Philadelphia chromosome-positive (Ph+) B-ALL showed antileukemic activity of the drug with CR/CR with partial hematologic recovery rates of 36% in the first 2 cycles of treatment [22]. Based on these promising outcomes, blinatumomab has been approved for treatment of Ph+ and Ph− B-ALL as well as elimination of MRD-positive residual disease postchemotherapy.

Despite the encouraging CR rates post-treatment with blinatumomab, disease relapse remains high when blinatumomab is used as a single agent for treatment of r/r B-ALL. Results of a retrospective study showed that patients with r/r ALL who responded to salvage therapy with blinatumomab and subsequently underwent allogeneic HCT as consolidation therapy had significantly improved event-free survival and OS compared with patients who were unable to do so [23]. In a recently published pediatric case series [24], of 15 patients with ALL in morphologic remission who received blinatumomab for MRD elimination, 14 patients (93%) achieved MRD-negative status and successfully underwent allogeneic HCT with no delay in neutrophil engraftment. In this good-risk population, the 1-year OS was 93.3%. In another study by Jabbour et al. [25], comparison of survival outcomes in adult patients with r/r B-ALL who received standard of care chemotherapy or blinatumomab showed higher remission rates in favor of the latter arm (69% versus 39%). The 1-year OS outcomes were better in patients who achieved MRD-negative remission pre-HCT, and a greater proportion of patients were able to undergo allo-HCT in the blinatumomab arm because of higher CR/incomplete count recovery rates compared with the salvage chemotherapy arm (34% versus 16%, P < .001). Furthermore, MRD rates were higher in the blinatumomab arm (53% versus 44%). Data were lacking on graft-versus-host disease (GVHD) and other HCT-related outcomes, such as nonrelapse mortality ( NRM), in patients who underwent allo-HCT in this report.

In this single-center retrospective study, we sought to assess the toxicity profile and transplant outcomes in a homogenous patient population with r/r B-ALL after salvage-therapy with blinatumomab either for morphologic (n = 24) or MRD-positive disease (n = 11) status before allogeneic HCT. The primary endpoint of this study was to investigate the safety of blinatumomab as salvage therapy in this population of B-ALL patients. Our report is the first to describe HCT-related outcomes in adult patients who receive blinatumomab as salvage therapy before allogeneic HCT.

METHODS

Patients and Eligibility

Institutional review board approval was obtained to review medical records of patients with B-ALL who received blinatumomab salvage therapy from 2012 to 2018 at our institution. Eighty-nine patients were identified who received salvage with blinatumomab for r/r ALL; of these treated patients, 43 (48%) achieved remission. For this study we report data on those patients who responded to blinatumomab and subsequently underwent HCT (n = 35). Patients treated with blinatumomab both for gross (n = 24) or MRD-positive (n = 11) disease pre-HCT were included. Patients who did not enter remission with blinatumomab salvage therapy or did not have subsequent allogeneic HCT were not included.

Endpoint Definition

OS was defined as the time from HCT to death from any cause or censored on the last known date to be alive. NRM was defined as death from causes not related to disease relapse/progression, and relapse/progression was considered as a competing risk. NRM was censored at the time of last follow-up if patients were alive and remained relapse/progression free. Leukemia-free survival (LFS) and relapse were defined per Center for International Blood and Marrow Transplant Research criteria [26]. NRM was considered a competing risk event for relapse. LFS and relapse were censored at the time of last follow-up when patients remained alive and free of relapse/progression. Grades II to IV and III to IV acute GVHD were defined by the Glucksberg scale [27], and chronic GVHD was defined as limited or extensive chronic GVHD according to the Seattle criteria [28]. Severity of chronic GVHD was evaluated using the National Institutes of Health consensus criteria [29]. Relapse and NRM were considered as competing risk events for engraftment and GVHD. GVHD-free, relapse-free survival (GRFS) was defined as the time from HCT to the first occurrence of disease relapse, acute GVHD grades III to IV, or moderate to severe chronic GVHD. GRFS was censored at the last follow-up if patients were alive and did not experience disease relapse, acute GVHD grades III to IV, or moderate to severe chronic GVHD post-HCT.

The MRD flow cytometry assay was sent out to an external laboratory with an assay of a detection sensitivity of more than .01% for leukemia blasts. The 18 fluorochromes used in this assay were CD4, CD5, CD7, CD13, CD14, CD15, CD16, CD19, CD33, CD34, CD38, CD45, CD56, CD64, CD71, CD117, CD123, and HLA-DR [30].

Statistical Analysis

Descriptive statistics were used to summarize baseline patient demographic, disease, and transplant-related characteristics. Kaplan-Meier curves and log-rank tests were used to evaluate OS, LFS, and GRFS. Cumulative incidence curves and Gray’s tests were used to evaluate relapse rates, NRM, and acute and chronic GVHD. All tests were 2-sided at a significance level of .05. SAS version 9.4 (SAS Institute, Cary, NC) was used to conduct the analyses.

RESULTS

Patient and Transplant Characteristics

Of the 89 patients who received blinatumomab for r/r ALL at our center, 43 patients responded to therapy, resulting in an overall response rate of 48%. Among the 43 patients who responded to blinatumomab, 35 patients (81%) underwent allogeneic HCT. Patient and transplant characteristics are summarized in Table 1. Briefly, the median age at the time of HCT was 35 years (range, 19 to 72), of which 12 patients (34%) were older than 50 years. HCT was done in first CR (n = 17, 49%) or at least second CR (n = 18, 51%), with 2 patients receiving blinatumomab salvage after prior HCTs (1 patient with a prior allogeneic HCT and 1 patient after a prior autologous HCT). Patients received transplant from a matched sibling donor (n = 13; 37%), matched unrelated donor (n = 13; 37%), haploidentical donor (n = 8; 23%), or double umbilical cord blood (n = 1; 3%) after myeloablative (n = 21, 60%) or reduced-intensity (n = 14; 40%) conditioning. The most common myeloablative regimen was total body irradiation and etoposide (n = 14; 40%) [31]. Fludarabine/melphalan (n = 9; 26%) was the most commonly used reduced-intensity regimen [32]. Graft source was peripheral blood stem cells in most patients (n = 31; 88.6%). GVHD prophylaxis was predominantly tacrolimus/sirolimus-based (n = 22, 63%) for matched sibling donor and matched unrelated donor recipients, and post-transplant cyclophosphamide with tacrolimus/mycophenolate mofetil was uniformly used for patients undergoing haploidentical transplant.

Table 1.

Patient and Transplant Characteristics (N = 35)

Characteristic	Value
Age at HSCT, yr
Median (range)	35 (19–72)
Interquartile range	24, 55
Recipient sex
Male	21 (60)
Female	14 (40)
Months from diagnosis to HCT
Median (range)	7 (3–46)
Interquartile range	5,10
Year of HCT
2012–2017	18 (51.4)
2018	17 (48.6)
Female donor to male recipient
Yes	7 (20)
No	28 (80)
Donor age, yr
Median (range)	31 (0–64)
Interquartile range	23,41
Disease status
First CR	17 (48.6)
Second or more CR	18 (51.4)
High-risk cytogenetics
Yes	18 (51.4)
No	17 (48.6)
Karnofsky performance status
90–100	18 (51.4)
70–80	14 (40)
Unknown	3 (8.6)
HCT-specific comorbidity index
0	9 (25.7)
1–2	15 (42.9)
>2	11 (31.4)
Donor type
Matched related	13 (37.1)
Matched unrelated	13 (37.1)
Haploidentical	8 (22.9)
Cord blood	1 (2.9)
Stem cell source
Peripheral blood stem cells	31 (88.6)
Bone marrow	3 (8.6)
Cord blood	1 (2.9)
ABO blood group compatibility
ABO compatible	24 (68.6)
Minor mismatch (donor is O)	5 (14.3)
Major mismatch (recipient is O)	4 (11.4)
Bidirectional (none is O)	2 (5.7)
Donor/recipient CMV serostatus
D−/R−	2 (5.7)
D−/R+	8 (22.9)
D+/R−	3 (8.6)
D+/R+	22 (62.9)
MRD pre-HCT
MRD negative pre-HCT (UW)	19 (54.3)
Morphologic remission	12 (34.3)
Morphologic remission, MRD +VE	3 (8.6)
Morphologic remission = 12	1 (2.9)
MRD
Positive	16 (45.7)
Negative	19 (54.3)
Conditioning regimen
Clofarabine/melphalan	1 (2.9)
Fludarabine/CTX/TMI	2 (5.7)
Fludarabine/melphalan	9 (25.7)
Fludarabine/melphalan/total body irradiation	2 (5.7)
FTBI/etoposide	14 (40)
Fludarabine/CTX/FTBI	3 (8.6)
FTBI/CTX	2 (5.7)
Fludarabine/FTBI	2 (5.7)
Conditioning regimen intensity (description)
Ablative	21 (60)
Nonmyeloablative/reduced Intensity	14 (40)
GVHD prophylaxis
mycophenolate mofetil/cyclosporine	1 (2.9)
CTX/tacrolimus/mycophenolate mofetil	7 (20)
Tacrolimus/mycophenolate mofetil	1 (2.9)
Tacrolimus/MTX	2 (5.7)
Tacrolimus/sirolimus	22 (62.9)
Tacrolimus/sirolimus/mycophenolate mofetil	1 (2.9)
Tacrolimus/sirolimus/MTX	1 (2.9)

Values are n (%) unless otherwise defined. UW indicates ; VE, ; CTX, ; TMI, ; FTBI, ; MTX, methotrexate.

Most patients had Ph− ALL disease (n = 31). High-risk cytogenetic markers were seen in 18 patients (51%) and included Ph (n = 4), mixed lineage leukemia rearrangements (n = 2), complex karyotype (n = 5), and other high-risk cytogenetics (n = 7). The mean blast percentage was 48% (range, 7.5% to 90%) in patients who received blinatumomab. In patients with morphologic disease, the mean MRD level before blinatumomab salvage therapy was .8% (range, .01% to 2.5%).

Blinatumomab Treatment and Toxicities

At the time of salvage with blinatumomab, patients had either grossly morphologic relapse (n = 24, 68%) or MRD alone, defined by flow cytometry analysis (n = 11, 31%). Negative MRD remission status was achieved in pre-HCT bone marrow biopsy in 19 patients, including 9 of 11 patients (81%) who received blinatumomab for MRD eradication before HCT.

Sixty cycles of blinatumomab were administered with the mean of 1.8 cycles per patient (range, 1 to 5). All but 1 patient with morphologic relapse achieved remission after 1 cycle of blinatumomab. Cytokine release syndrome was seen in 13 patients, and all patients responded to supportive care measures without requiring cessation of blinatumomab infusions. One patient developed grade 3 neurotoxicity requiring cessation of therapy. Grade 1 cytokine release syndrome manifested with fever (n = 10), neurotoxicity (n = 2), or transaminitis (n = 2). Before salvage therapy with blinatumomab, 20 patients (57%) received upfront treatment with an adolescent/young adult regimen. The mean number of chemotherapy regimens before blinatumomab salvage was 1.3 per patient (range, 1 to 3).

Transplant Outcomes

With a median follow-up of 18.0 months (range, 5.7 to 70.2), 1-year OS and LFS were 77% (95% confidence interval [CI], .58 to .88) and 65% (95% CI, .46 to .78), respectively, and 2-year OS and LFS were 52% (95% CI, 29 to 72) and 40% (95% CI, 20 to 60), respectively (Figure 1A). Cumulative incidence of relapse at 1 and 2 years was 18% (95% CI, .07 to .32) and 36% (95% CI, .16 to .57), respectively (Figure 1B). Incidences of grades II to IV and III to IV acute GVHD at day +100 post-HCT were 43% (95% CI, .26 to .59) and 20% (95% CI, .09 to .35), respectively (Figure 1C). The cumulative incidences of all chronic GVHD and extensive chronic GVHD at 24 months were 36% (95% CI, .20 to .52) and 30% (95% CI, .15 to .46), respectively (Figure 1D). At 1 year post-HCT NRM was 18% (95% CI, .07 to .32) (Figure 1E). Finally, the composite endpoint of GRFS at 1 year was at 42% (95% CI, .25 to .57) (Figure 1F). No cases of sinusoidal obstruction syndrome (SOS) or primary or secondary graft failure were seen in this cohort, and stable neutrophil and platelet engraftment was achieved in >90% of patients by 26 and 42 days, respectively (Supplementary Figure 1).

Figure 1.

Transplant outcomes. (A) OS at 1 and 2 years post-HCT. (B) Cumulative incidence of relapse at 1 and 2 years post-HCT. (C) Cumulative incidence of acute GVHD at 100 days post-HCT. (D) Cumulative incidence of chronic GVHD at 2 years post-HCT. (E) Cumulative incidence of NRM at 100 days and 1 year post-HCT. (F) Probability of GRFS at 1 year post-HCT.

Causes of death post-HCT were GVHD (n = 3), NRM (n = 5; 4 sepsis and 1 regimen-related toxicity), and relapse (n = 4). Of the 4 patients who relapsed, 2 patients had bone marrow only relapse and 2 patients had systemic and extramedullary relapse in the central nervous system. CD19 expression was retained in 3 patients who relapsed, and only 1 patient relapsed with CD19-negative disease.

The most common viral infections occurring within 100 days post-HCT were cytomegalovirus (CMV; n = 17, 49%) and human herpesvirus-6 (n = 5, 14%). Bloodstream infections were reported in 9 patients (26%). Of the bacterial species isolated, gram-negative bacteremia was seen in 5 patients (Escherichia coli in 2 and Pseudomonas aeruginosa, Klebsiella pneumoniae, and Serretia marcescens in 1 patient each). Grampositive bacteremia was seen in 4 patients (Enterococcus faecalis and Bacillus cereus in 1 patient each and coagulase-negative Staphylococcus in 2 patients), and 5 patients had Clostridium difficile infection. Other rare infections included BK virus (n = 2), adenovirus (n = 1), and mold infection (n = 1).

Subgroup Analysis Based on Pre-HCT Patient Characteristics

Of the 35 patients in this cohort, 11 patients received blinatumomab for MRD clearance only, of which 9 achieved MRD-negative status (81%). Overall, pre-HCT MRD-negative status was seen in 19 patients (including the 9 patients who received the drug for MRD eradication as noted above). Of the 24 patients who received blinatumomab for morphologic relapse, 10 achieved MRD-negative remission and 14 achieved morphologic remission pre-HCT. Subgroup analysis based on MRD status did not show a significant difference in the 1– and 2-year OS and NRM in patients with MRD-negative (n = 19) status at the time of HCT compared with patients with MRD-positive status (n = 16) (Figure 2).

Figure 2.

Subgroup analysis based on pre-HCT MRD status. (A) OS at 1 and 2 years post-HCT. (B) NRM at 1 and 2 years post-HCT.

When we examined the associations between age at HCT (<40 versus ≥40 years), time from diagnosis to HCT, conditioning intensity, and cytogenetics risk with OS, LFS, NRM, and relapse, only conditioning intensity was significantly associated with OS (log-rank P = .038). Patients who received myeloablative conditioning survived longer compared with those who received reduced-intensity conditioning (Supplementary Figure 2).

We assessed OS and LFS outcomes in patients with standard-risk (n = 17) and high-risk (n = 18) cytogenetic abnormalities. There was no difference in 1-year OS between patients with standard-risk (77%; 95% CI, 50.3 to 90.9) and high-risk (75%; 95% CI, 47 to 90.1) cytogenetic abnormalities (P = .706). One-year LFS was 66.2% (95% CI, 39.6 to 83.2) and 63.3% (95% CI, 35.8 to 81.6) in patients with standard- and high-risk cytogenetics, respectively (P = .712).

DISCUSSION

Our report demonstrates encouraging HCT-related clinical outcomes of adult patient with r/r B-ALL who received allogeneic HCT after salvage therapy with blinatumomab. The 1-year OS of 77% was favorable in this high-risk group of patients who otherwise carried a poor prognosis for long-term survival. Consistent with results of a recent study by Jabbour et al. [25], our NRM outcomes after blinatumomab therapy pre-HCT were reassuringly low. Acute and chronic GVHD outcomes were similar to results reported in large registry studies [33,34]. Importantly, no unusual or unexpected toxicities were observed and no cases of graft failure or delayed neutrophil engraftment were noted in patients of this cohort. Interestingly, no cases of SOS were seen in patients who were otherwise at high risk of this complication, given their extensive prior exposure to chemotherapy. This outcome is in contrast with the SOS rate of 13% reported in patients who underwent allogeneic HCT after salvage with inotuzumab ozogamycin in the INO-VATE study [35,36]. Based on the overall safety profile of blinatumomab pre-HCT, this drug represents an attractive choice for first-line salvage therapy, especially in heavily pretreated patients who are at a higher risk of hepatic complications arising from HCT conditioning. Seventeen patients in our study developed CMV reactivation, reflecting a potentially high risk of viral reactivation in the heavily pretreated patient population. No death was attributed to CMV disease, and hopefully with approval of levermovir in CMV-seropositive patients in the early post-transplant setting the incidence of CMV reactivation will decline [37].

The benefit of allogeneic HCT in achieving long-term cure in r/r ALL has been demonstrated in 2 large, multicenter, retrospective studies [38,39]. However, until recently the major barrier to allogeneic HCT in patients with r/r ALL was the low rates of remission with conventional therapies. Blinatumomab is the first US Food and Drug Administration—approved bispecific T cell engager therapy for patients with r/r ALL. In a pivotal phase II study that led to approval of blinatumomab, an NRM of 11% was achieved in patients who underwent allogeneic HCT [19]. However, details of HCT outcomes (ie, GVHD, SOS, and graft failure) were not available in the initial report. In this study we showed low NRM rates consistent with earlier reports and comparable acute and chronic GVHD rates in patients who received salvage chemotherapy before allogeneic HCT [14,39].

In our study the 1-year LFS was 65%, which declined to 40% at 2 years in surviving patients, indicating the need for continued vigilance for relapsed disease after allogeneic HCT and possible administration of maintenance agents in this setting [40]. Given the potential role of blinatumomab in augmenting the graft-versus-leukemia effect [41] and eradication of MRD after chemotherapy [21,42], maintenance therapy with blinatumomab after allogeneic HCT is currently being investigated in a prospective clinical trial (NCT02807883).

In our cohort, most patients who were MRD-positive after chemotherapy achieved MRD-negative status after blinatumomab therapy (9/11 patients; 81%). Of the 35 patients in our cohort, 19 patients achieved MRD-negative status pre-HCT. Unlike previous reports, achievement of MRD-negative status before allogeneic HCT did not impact survival or NRM, suggesting that patients can be offered allogeneic HCT when in morphologic remission, and additional therapy to eliminate MRD-positive disease may not be beneficial in this specific setting. Given the small sample size of our study and the currently available data from previously published studies [43–45], these findings should be confirmed in a larger prospective study.

After blinatumomab approval, 2 other classes of immunotherapy agents, tisagenlecleucel [46–48] and inotuzumab ozogamycin [49], have also been approved for treatment of patients with r/r ALL. Despite these advances, allogeneic HCT is the only curative option for this high-risk patient population [50,51]. With advances in graft cell engineering, GVHD prophylaxis, availability of alternative donors [52], and adoption of novel conditioning regimens [53], allogeneic HCT can be offered to most patients with r/r ALL who are in remission with these newly approved agents. Relapse rates after allogeneic HCT remain high for patients not in remission; hence, the advent of immune therapies in the salvage setting [54] will most certainly improve clinical outcomes of patients who undergo allogeneic HCT in deep remission. Importantly, the sequencing of these therapies in patients with r/r ALL is not yet well characterized and may be influenced by malignant cell surface antigen expression (CD22 or CD19), age, SOS risk, disease status, physician preference, and resources available at the treating institution [55,56]. Improvements in the design of chimeric antigen receptor–based cellular therapy may lead to less toxicity and more effective targeting of tumor cells, further altering the treatment landscape of r/r ALL patients in the future [57].

Based on its retrospective nature, our study has several limitations, including a small sample size, heterogeneity in transplant type (allogeneic, haploidentical, and cord blood), disease status (gross and MRD positive only), and transplant conditioning intensity (myeloablative and reduced intensity). Additional multicenter studies with larger patient populations may help to better explore the role of these newly approved immunotherapy agents pre-HCT.

In conclusion, results of our study indicate that in adult patients with r/r ALL, blinatumomab-based salvage therapy before allogeneic HCT is associated with excellent 1-year OS of 71% and no unexpected toxicities in terms of GVHD, SOS, and engraftment failure post-HCT. Our data suggest that blinatumomab is a safe and effective agent for r/r ALL patients before HCT.