Incidence and Description of Autoimmune Cytopenias During Treatment with Ibrutinib for Chronic Lymphocytic Leukemia Autoimmune Cytopenias During Ibrutinib Treatment

Abstract

Chronic lymphocytic leukemia (CLL) is frequently complicated by secondary autoimmune cytopenias (AIC). Ibrutinib is an irreversible inhibitor of Bruton’s Tyrosine Kinase approved for treatment of relapsed CLL and CLL with del(17p). The effect of ibrutinib treatment on the incidence of AIC is currently unknown. We reviewed medical records of 301 patients treated with ibrutinib as participants in therapeutic clinical trials at the Ohio State University Comprehensive Cancer Center between July 2010 and July 2014. Subjects were reviewed with respect to past history of AIC, and treatment emergent AIC cases were identified. Prior to starting ibrutinib treatment, 26% of patients had experienced AIC. Information was available for a total of 468 patient-years of ibrutinib exposure, during which there were six cases of treatment emergent AIC. This corresponds to an estimated incidence rate of 13 episodes for every 1 000 patient-years of ibrutinib treatment. We further identified 22 patients receiving therapy for AIC at the time ibrutinib was started. Of these 22 patients, 19 were able to discontinue AIC therapy. We found that ibrutinib treatment is associated with a low rate of treatment emergent AIC. Patients with an existing AIC have been successfully treated with ibrutinib and subsequently discontinued AIC therapy.

Introduction

Autoimmune cytopenias (AIC) complicate the disease course of chronic lymphocytic leukemia (CLL) and small lymphocytic lymphoma (SLL) in 10–25% of cases longitudinally over the course of disease.^{1, 2} This can occur at any time, but most often is noted during times when CLL is clinically active.^{2, 3} Autoimmune hemolytic anemia (AIHA) is the most common AIC observed, followed by immune thrombocytopenia (ITP). Pure red cell aplasia (PRCA) and autoimmune granulocytopenia (AIG) are seen only rarely.^{2, 3} These secondary AIC require treatment with immunosuppression, which adds to the treatment burden, morbidity, and sometimes mortality experienced by CLL patients.^3–5 AIC in some cases can be triggered by treatment with standard therapies, notably chemotherapeutic agents such as fludarabine.^{2, 6, 7} While this risk with fludarabine can be abrogated by use in combination with cyclophosphamide, understanding the relationship of any new CLL targeted treatment with AIC remains imperative.^8–10

Inhibitors of Bruton’s tyrosine kinase (BTK) are an important new class of targeted agents approved for the treatment of CLL as well as other lymphoid malignancies. BTK is a critical molecule in the B cell receptor (BCR) signaling pathway. BTK inhibitors effectively abrogate BCR and integrin signaling, and are highly effective in the treatment of CLL.^11–13 The first-in-class agent, ibrutinib, was approved in 2014 by the United States Food and Drug Administration (FDA) for the treatment of CLL patients who have received previous treatment or who have the high risk cytogenetic feature of deletion 17p.¹⁴ Ibrutinib is now in widespread clinical use, and as the drug is currently administered continuously until disease progression, the prevalence of ibrutinib use over time will continue to increase.

The effects of ibrutinib treatment on the natural history of AIC have not yet been characterized. We conducted a retrospective cohort review of patients treated with ibrutinib for CLL or SLL to determine the incidence of AIC during treatment and describe the course of patients with a history of AIC, including those on chronic AIC therapy.

Methods

Patients who received ibrutinib at The Ohio State University Comprehensive Cancer Center on four sequential completed clinical trials with enrollment from July 2010 to July 2014 were included in this analysis. All studies were approved by the appropriate institutional review board. The treatment schemas for these trials are shown in Table 1. In three of these studies patients received single agent ibrutinib, whereas in one study enrolling a minority of the patients (21%), ibrutinib was given in combination with ofatumumab.

Table 1.

Clinical Trial Treatment Plans

OSU 10032 (PCYC 1102) n=50	Ibrutinib 420 mg or 840 mg daily until disease progression
OSU 10053 (PCYC 1109) n=64	Cohort 1: Ibrutinib 420 mg daily starting C1D1 until disease progression; Ofatumumab starting C2D1 Cohort 2: Ibrutinib 420 mg daily starting C1D2 until disease progression; Ofatumumab starting C1D1 Cohort 3: Ofatumumab starting C1D1; Ibrutinib start C3D1 daily until disease progression
OSU 11133 n=150	Ibrutinib 420 mg daily until disease progression
OSU 12024 (PCYC 1112) n=37	Arm A: Ofatumumab starting C1D1; may cross over to ibrutinib upon disease progression (n=12) Arm B: Ibrutinib 420mg daily until disease progression (n=25)

All patients who received at least one dose of ibrutinib were included in these analyses. The diagnosis of AIC was made by the treating physician but was not independently verified. Diagnoses of AIHA, ITP, and PRCA were considered as AIC. Autoimmune granulocytopenia was not included due to the rarity of the condition, difficulty in confirming this diagnosis, and previously described relationship to treatment with anti-CD20 monoclonal antibodies.^15–17

Treatment emergent AIC were defined as episodes occurring any time between the first and last dose of ibrutinib received; were considered new if the type had never before been experienced by that patient, even if the patient had an antecedent AIC of a different type; and were classified as relapsed if the same type had been experienced by a patient prior to starting ibrutinib.

Patient characteristics were summarized with descriptive statistics and compared between those who had and had not previously experienced AIC using Fisher’s exact test or the Wilcoxon rank sum test. All tests were two-sided and statistical significance was declared at α=0.05. The person-time incidence rate of treatment-emergent AIC events was estimated by dividing the number of AIC episodes observed while on ibrutinib therapy by the total number of years patients were receiving ibrutinib and were at risk. The incidence rate is presented as the number of occurrences per 1 000 patients per year (i.e. per 1 000 person-years). Differences in incidence rates between subgroups were estimated with 95% confidence intervals and compared using the person-time chi-square test.¹⁸ Among the 22 patients who were on chronic AIC therapy at the time of beginning ibrutinib, time to discontinuation of AIC therapy was calculated from the date of first ibrutinib treatment to the date of discontinuation (event) or date last known to be on AIC therapy (censored). Median time to discontinuation was estimated using the method of Kaplan-Meier. Last follow up was in July of 2014 for all patients except those on a chronic AIC therapy who underwent approximately 6 additional months of follow up to monitor their chronic AIC therapy use.

Results

Cohort Characteristics

A total of 301 patients treated with ibrutinib were identified. Baseline characteristics of the cohort can be found in Table 2. The study cohort included primarily patients with relapsed CLL and was relatively enriched for CLL disease features associated with AIC including complex karyotype, deletion 17p, and unmutated IGHV status.^19–21 Prior to starting ibrutinib, 26% (78/301) of patients had experienced an episode of AIC. Of patients who had a history of AIC, 44 (56%) had AIHA, 25 (32%) had ITP, 8 (10%) had both AIHA and ITP either concomitantly or sequentially, and 1 (1%) had PRCA. They were treated with a variety of AIC therapies with many patients receiving multiple therapies. These AIC therapies are found in Table 3.

Table 2.

Baseline Characteristics of Cohort

Characteristics	n=301
Median age at diagnosis, years (range)	55 (20–85)
Median age at start of Ibrutinib, years (range)	65 (26–91)
Median duration from CLL diagnosis to start of Ibrutinib, years (range)	7.2 (0.15–32.8)
Men, n (%)	211 (70)
Rai stage, n (%)
Low (0/1)	80 (27)
Intermediate (2)	22 (7)
High (3/4)	199 (66)
Median prior CLL therapies, (range)	3 (0–13)
Prior Treatments (not mutually exclusive), n (%)
Fludarabine	218 (72)
Pentostatin	42 (14)
Alkylating agent	247 (82)
Anti-CD20 monoclonal Ab	288 (96)
Alemtuzumab	39 (13)
Allogeneic transplant	13 (4)
History of AIC, n (%)	78 (26)
FISH (not mutually exclusive), n (%)
BCL6 (3q27)	25 (8)
CMYC (8q24)	62 (21)
D12Z3 (12cen)	51 (17)
D13S319 (13q14·3 x0/13q14·3 x1)	154 (52)
ATM (11q22·3)	81 (27)
TP53 (17p13·1)	112 (38)
No FISH panel (unknown)	4
Karyotype, n (%)
Complex	165 (57)
Not complex	123 (43)
Unknown	13
IGHV Mutational Status, n (%)
Mutated	54 (20)
Unmutated	212 (80)
Unknown	35

Table 3.

Pre-Ibrutinib AIC Therapies

AIC Therapy, n (%)	n=78
Steroids	64 (82)
Rituximab	47 (60)
Intervenous immune globulin (IVIG)	25 (32)
Cyclosporine	13 (17)
Splenectomy	12 (15)
Other*	16 (21)

^*Other AIC therapies included: Rho(D) immune globulin, danazol, romiplostim, azathioprine, cyclophosphamide, mycophenolate mofetil, and epoetin alfa.

Characteristics were not significantly different between patients who had previously experienced AIC and those who had not, except for a positive direct anti-globulin test (DAT) at study entry. A DAT was not required for all studies, but was available for 252 patients. In those patients a positive DAT was more common in the group who had previously experienced AIC (40%) compared to those who had not (8%) (p<0.0001).

With a median follow-up of 1.8 years (range 0.2–4.1), 216/301 (72%) patients remained on ibrutinib therapy. The median time to discontinuation had not been reached.

Patients Receiving Continuous AIC Therapy at Ibrutinib Initiation

Of the 78 patients with a history of AIC, 22 (28%) were receiving AIC therapy when ibrutinib was started, although all had controlled disease. One study (OSU 12024) specifically excluded patients with uncontrolled AIHA or ITP defined as those requiring >20mg of prednisone daily or equivalent dose of steroids or declining hemoglobin or platelet count due to immune destruction. The remaining studies functionally excluded uncontrolled AIC by excluding immune therapies or cytopenias not related to CLL infiltration of bone marrow.

In the group of patients taking an AIC therapy at the time ibrutinib was started, the majority of patients had only experienced one prior AIC episode (12/22, 55%). An additional five patients had experienced two episodes, two had three prior episodes, and one patient each had four, five, and six prior episodes. Median time from the first AIC episode to start of ibrutinib treatment was 14.6 months (range 0.3–85.0), and these patients had been on continuous AIC therapy for a median of 2.9 months (range 0.3–34.2). Of these patients, 14 (64%) were receiving prednisone alone, 3 (14%) were receiving prednisone plus cyclosporine, 2 (9%) romiplostim, 2 (9%) cyclosporine, and 1 (5%) monthly rituximab. The patient on monthly rituximab discontinued this treatment upon starting ibrutinib and did not experience an AIC relapse.

AIC therapy was discontinued in 86% of patients, at an estimated median of 4.7 months (95% CI: 0.7–7.4) from ibrutinib initiation. Time to AIC therapy discontinuation was variable and is shown in Figure 1 along with timing of AIC episodes in relation to ibrutinib treatment. Three patients were known to be on an AIC therapy at the last time treatment records were available. One patient with ITP remained on 2.5 mg of prednisone daily and was censored at last follow up. Two others were censored once AIC treatment records were no longer available. Only one patient with a history of refractory AIHA experienced a relapse after discontinuing chronic AIC therapy with prednisone three weeks into ibrutinib treatment. He was unable to remain on ibrutinib due to high-dose steroid treatment for AIHA and recurrent infections.

Figure 1. Relationship of AIC episodes and therapy to ibrutinib treatment.

Time from the start of most recent AIC therapy to AIC therapy discontinuation is shown for the 22 patients who were on a continuous AIC therapy at the time of starting ibrutinib. All AIC episodes experienced by these patients are shown. Two patients were censored when they were no longer followed for their AIC therapy. One patient was censored at last follow up when known to be taking prednisone. * Episode of AIHA treated with steroids, with a second episode after ibrutinib was discontinued. ** Episode of ITP treated with IVIG.

Treatment Emergent AIC

Six cases of treatment-emergent AIC were identified over a total of 468 patient-years of ibrutinib exposure, corresponding to an estimated incidence rate of 13 (95% CI: 3–23) AIC episodes per 1 000 patient-years of ibrutinib exposure. Of these episodes, three were new and three were relapses (Table 4). Only one episode occurred in the group receiving concomitant anti-CD20 monoclonal antibody treatment estimating an incidence rate of 7 (95% CI: 0–21) AIC episodes per 1 000 patient-years of ibrutinib treatment in this group, compared to an estimated 15 (95% CI: 2–29) AIC episodes per 1 000 patient-years in the ibrutinib monotherapy group. Given the rarity of these events and with observation over a limited period of time, the difference in rates did not demonstrate statistical significance.

Table 4.

Treatment Emergent AIC Cases

	Patient 1	Patient 2	Patient 3*	Patient 4	Patient 5	Patient 6
Age at study entry, Gender	50’s M	70’s F	60’s M	70’s M	70’s M	70’s F
Number of prior CLL Treatments	5	0	3	6	2	0
Rai Stage	IV	I	II	IV	I	IV
Complex Karyotype	Y	N	Y	Y	Y	N
IGHV mutational status	mutated	mutated	unmutated	unknown	unmutated	unmutated
FISH panel abnormalities	BCL6 (3q27) CMYC (8q24) TP53 (17p13·1) D13S319(13q14·3)	D12Z3 (12cen)	D13S319(13q14·3) ATM (11q22·3)	none	D13S319(13q14·3) TP53 (17p13·1)	D13S319(13q14·3)
AIC prior to ibrutinib	none	none	AIHA	AIHA	AIHA	ITP
AIC therapies prior to ibrutinib	n/a	n/a	Prednisone, IVIG, Rituximab, Dexamethsone, Aranesp	Rituximab, oral cyclophosphamide, mycophenolate mofetil, IVIG, prednisone, cyclosporine	Prednisone, IVIG	None
On an AIC therapy when starting ibrutinib	n/a	n/a	No	Yes, Prednisone 10mg	Yes, Prednisone 10mg	n/a
Treatment emergent AIC type	New AIHA	New AIHA	Relapsed AIHA	Relapsed AIHA	New ITP	Relapsed ITP
AIC therapy received	Prednisone, IVIG	IVIG	IVIG	Prednisone	Prednisone	Prednisone, Dexamethasone
AIC controlled	Yes	Yes	Yes	No	Yes	No
On ibrutinib at last follow up	No	Yes	Yes	No	Yes	No
Patient outcome	Died from infection	Alive and well	Alive and well	Stopped ibrutinib due need for steroids and infections, died from infection	Alive and well	Discontinued ibrutinib due to rash

Characteristics are at time of ibrutinib initiation unless otherwise specified. Patient 3 is the only patient who received concomitant treatment with ofatumumab.

^*Developed autoimmune granulocytopenia which was treated with cyclosporine.

All treatment-emergent AIC patients received treatment with single agent ibrutinib except one who received ibrutinib in combination with ofatumumab. The patients were diverse in terms of age (range 53–77 years-old), Rai stage (range I–IV), and number of prior CLL treatments (range 0–6). AIC episodes did not appear to be temporally associated with ibrutinib initiation, occurring at 0.2, 1, 6, 7.5, 22, and 23 months after the first dose of ibrutinib. The treating physician suspected that the episode of AIHA diagnosed one week after starting ibrutinib may have antedated ibrutinib initiation, but confirmatory laboratory testing was not available.

Treatment-emergent cases received AIC therapy with corticosteroids, IVIG, or both. Ibrutinib treatment was temporarily suspended in four of the six cases to allow for doses of prednisone excluded by the clinical trial protocol. Patients restarted ibrutinib once steroid dose was lowered to the permitted level. The three patients who subsequently discontinued ibrutinib did so for reasons that were not directly related to their AIC. Two patients died from infections and one had a rash which may have been unrelated to study or AIC treatment. Of the three patients able to continue ibrutinib, all recovered and AIC therapy was able to be discontinued without relapse.

Of note, the patient who received combination ofatumumab and ibrutinib treatment and experienced a relapse of AIHA, developed isolated neutropenia six months after starting ibrutinib (eight months after starting ofatumumab). His neutropenia was initially treated with G-CSF and a bone marrow biopsy demonstrated improvement in CLL (60–70% CLL prior to therapy vs 30% on ibrutinib). There was no morphologic evidence of myelodysplastic syndrome and karyotype and FISH panel were unchanged from study entry. A monoclonal T cell population was identified by PCR and he was started on cyclosporine with improvement in the neutropenia. The diagnosis of autoimmune granulocytopenia or late onset neutropenia after anti-CD20 monoclonal antibody therapy was considered. Later in his ibrutinib treatment course he was successfully treated for a relapse of his AIHA with IVIG without discontinuing ibrutinib. At last follow up he remained on cyclosporine and ibrutinib with a normal neutrophil count and no evidence of hemolysis.

Discussion

Here we provide the first detailed description of the incidence and outcomes of AIC in patients receiving ibrutinib. This adds to previous abstracts by our group and others.^{22, 23} Our study cohort was at increased risk for the development of AIC given the long median duration of CLL, high median number of prior therapies, and relatively high percentage of patients with unmutated IGHV.²⁴ As an indicator of this risk, 26% of the patients had experienced a least one episode of AIC prior to study enrollment, which is one of the higher reported AIC rates in studies of CLL patients.^{2, 25} Despite this, a low rate of treatment emergent AIC was seen, suggesting that ibrutinib reduces the risk of AIC development.

There are several mechanisms of AIC development specific to CLL, which are potentially targeted by ibrutinib. First, any treatment that is able to control the CLL clone will reduce secondary AIC. In AIHA, the leukemic cells have been demonstrated to be more efficient at presenting Rh antigens than other antigen-presenting cells.²⁶ Reduction in the ability of the CLL cells to perform this function by effective therapeutic treatment with ibrutinib would reduce Rh antigen presentation and therefore AIC. Beyond that, it is known that in the majority of cases of AIHA and ITP the pathogenic antibody is an IgG produced by non-leukemic poly- or oligoclonal “bystander” B cells.²⁷ By inhibiting BCR signaling, the ability of these non-leukemia B cells to respond to autoantigens and produce the pathologic antibody may be inhibited, halting antibody mediated AIC. This mechanism is less clearly implicated in PRCA or AIG that is not mediated by an antibody. Ibrutinib has the potential to inhibit AIC mechanisms beyond its effect on B cells. The environment of immune dysregulation in CLL plays a role in AIC development with changes in T cell subsets associated with AIC development.²⁸ Ibrutinib can impact this with inhibition of ITK, which could diminish response of auto reactive T cells. Additionally, ibrutinib has been shown to shift T helper immunity towards a Th1 response that is less favorable for the development of AIHA, which primarily exhibits Th2.^{29, 30} Lastly, ibrutinib may reduce clearance of antibody coated cells from circulation by decreasing function of the reticuloendothelium. This has not yet been demonstrated, but ibrutinib does alter macrophage function.³¹ If our findings are reinforced in laboratory models and prospective clinical study, ibrutinib would represent a therapeutic alternative to existing therapies for AIC.

Our analysis is limited by the retrospective design. Many of the patients did not receive medical care at our institution from the time of their initial CLL diagnosis, and therefore incomplete records may underestimate the incidence of historical AIC. As well, patients were prospectively followed for ibrutinib efficacy and safety and not specifically for AIC, so new AIC diagnoses were not always extensively evaluated, potentially over- or underestimating the incidence of treatment-emergent AIC. However, given that ibrutinib is relatively new to clinical practice and our center has a large cohort of patients with long follow-up, more robust data will not likely be available for a long period of time. There is also a possibility of ascertainment bias as patients with active AIC were excluded from enrollment on the included clinical trials. However, given the number of patients with prior AIC episodes, receiving ongoing AIC treatment when starting ibrutinib, or experiencing an episode of AIC in the months prior to starting ibrutinib as shown in Figure 1, we would still anticipate a much higher incidence rate than we report.

Ibrutinib is already changing the way CLL is treated by providing effective treatment that is tolerable and avoids the risks of conventional chemotherapy. As CLL treatment moves away from chemotherapy and more patients receive ibrutinib, additional benefits in terms of immune modulation can be realized. At a minimum, we offer preliminary evidence that control of AIC may be an important secondary benefit of ibrutinib that can assist in selection of appropriate patients for this treatment. Given the cost and indefinite duration of treatment, ibrutinib is unlikely to replace steroids as a first-line AIC therapy. However, ibrutinib may offer a new therapeutic option for AIC patients with concomitant treatment indications for their underlying CLL, refractory AIC, or experiencing toxicities from long-term immunosuppressant therapy should our findings be confirmed in future prospective clinical trials in this patient population.

We showed an absolute difference in the incidence rates of AIC episodes during ibrutinib treatment compared to treatment with ibrutinib and ofatumumab with a trend towards lower rates in the combination arm. This difference did not reach statistical significance due to the rarity of AIC events and may reach significance with longer-term follow up or a larger sample size. While the addition of anti-CD20 monoclonal antibodies has not been proven to increase the efficacy of ibrutinib treatment for CLL, the combination may have a benefit as AIC therapy in select patients.

In conclusion, ibrutinib may have an immune stabilizing effect which precludes the development of new AIC and facilitates discontinuation of chronic AIC therapies. This information is potentially significant for CLL patients taking or considering ibrutinib treatment, including those with a history of AIC where our findings do not show a high rate of AIC relapse. The effect of ibrutinib on AIC which we observed warrants further investigation in prospective clinical trials, particularly for patients with refractory AIC or those experiencing toxicities from long-term immunosuppression where there is an unmet clinical need for new therapies.