How I treat CLL patients with ibrutinib

Ibrutinib has demonstrated marked efficacy in chronic lymphocytic leukemia (CLL) in clinical trials^1-3  and is approved by the US Food and Drug Administration for the therapy of any CLL patient in any line of therapy. Its use has rapidly become standard of care for relapsed CLL patients, as well as for many frontline high-risk or older patients. In the most recent update of the RESONATE registration trial, with 4-year follow-up, only 12% of relapsed CLL patients had discontinued for adverse events (AEs) that included atrial fibrillation (AF), bleeding, and infection.^1,2,4,5  Diarrhea, arthralgia, and skin toxicity can also be significant, and hypertension emerges over time.^6,7  Data from real-world use of ibrutinib indicate that these toxicities may limit ibrutinib use, however, with a recent retrospective report showing that 42% of 621 primarily relapsed patients had discontinued the drug, almost half because of toxicity, at a median of 6 months.⁸  The most common reasons were AF, infection, pneumonitis, bleeding, and arthralgia, all well-recognized toxicities of ibrutinib. The median progression-free survival (PFS) was 35 months, significantly less than the 52 months reported for more heavily pretreated patients on the phase 1b/2 trial.⁷ 

Age is a risk factor for AEs with ibrutinib. Studies from Ohio State University have identified increasing age as the primary predictor of discontinuation for reasons other than progressive disease,^9,10  with rates of >60% at 6 months in patients over 80 years of age. Recent population-based studies have also found that worse performance status¹¹  or higher Cumulative Illness Rating Scale comorbidity score¹²  was associated with worse outcomes, increasing the chances of discontinuation or death¹¹  or reducing PFS.¹²  In older patients for whom I am considering ibrutinib, I carefully weigh disease risk with comorbidities that may increase complications. The most important of these are a bleeding diathesis or cardiac disease.

Ibrutinib is associated with low-grade ecchymoses and petechiae in about half of patients¹³  and with major hemorrhage rates that vary from 1%¹⁴  to 9%,¹⁵  depending on the clinical study. Although some hemorrhages are periprocedural, spontaneous major bleeding also occurs, likely in several percent of patients.¹⁶  This susceptibility is related to the role of BTK and TEC kinases in platelet activation downstream of the collagen receptor glycoprotein VI. Addition of ibrutinib ex vivo to healthy donor platelets, or therapy in vivo, inhibits collagen-induced platelet aggregation^17-19  and platelet adhesion under high shear.^18,20,21  This in vitro effect correlates with low-grade bleeding in ibrutinib-treated patients.^17-19  Platelet aggregation defects have also been described in CLL patients at baseline in comparison with healthy controls, which were then further exacerbated by ibrutinib therapy.^17,19,22  Two studies have independently suggested pretreatment screening by either ristocetin-induced platelet aggregation²³  or low von Willebrand factor (VWF) activity or factor VIII levels,¹⁹  but neither approach has yet been validated.

In early ibrutinib trials, major hemorrhage including subdural hematoma occurred in the context of warfarin,²⁴  leading to the exclusion of these patients from ibrutinib clinical trials. The experience of combining long-term anticoagulation with ibrutinib is therefore extremely limited.¹⁴  Antiplatelet agents are more common, reported at 34% in one study,¹⁴  but dual antiplatelet therapy is very rare. Generally, the latter should be avoided given evidence of increased major bleeding with this combination, without ibrutinib.²⁵  A recent meta-analysis of published trial data on ibrutinib and bleeding demonstrates a 2.72 relative risk of any bleeding and a trend toward increased major bleeding at 1.66 relative risk, despite being underpowered for this outcome.²⁶  These data likely underestimate the actual risk of bleeding in an unselected patient population. For example, a center-based report of 71 ibrutinib patients with a median age of 73 found that 70% were also receiving an antiplatelet medication, 17% were receiving an anticoagulant, usually apixaban for atrial fibrillation, and 13% were receiving both.²⁷  Major bleeding occurred in 18% of patients, including 78% receiving both antiplatelet and anticoagulant therapy.²⁷  No major bleeds occurred in this study in patients who were not also receiving antiplatelet, anticoagulant, or a CYP3A4-interacting therapy, underscoring the need for care in combining these agents.

In practice I avoid ibrutinib in patients with a history of major or potentially life-threatening hemorrhage owing to an irreversible cause. Similarly, if a reasonable alternative therapy is available for a patient requiring long-term anticoagulation or dual antiplatelet therapy, I favor the alternative CLL therapy. If ibrutinib is uniquely the best option, then I evaluate the true necessity of the anticoagulation or dual-antiplatelet therapy. I also suggest that all patients stop vitamin E, fish oils, and nonsteroidal anti-inflammatory drugs, as in the ibrutinib clinical trials, and if on aspirin, consider stopping or reducing to the lowest available dose.

Ibrutinib should be held 3-7 days prior to and after any invasive procedures owing to the risk of periprocedural bleeding.^6,27,28  Therefore, if a patient needs a surgical procedure, I attempt to get it done before initiating ibrutinib. This is particularly relevant for patients with a large disease burden or aggressive disease who may experience disease regrowth if ibrutinib is held early after initiating therapy.^6,27  Once patients have been on ibrutinib for some time (>6-12 months) and achieved solid disease control, transient holds for procedures are not as problematic.

The best studied cardiac side effect of ibrutinib is AF, which occurs initially in about 6% of patients,^2,3,26  increasing to 10% to 15% over 2 years.²⁹  An analysis from 4 randomized registration trials identified risk factors for AF on ibrutinib as age >65 and a prior AF history. Hypertension and hyperlipidemia were significant in univariate but not multivariate analyses.²⁹  The recently described Shanafelt risk score for AF also stratified patients in this cohort into risks ranging from 0.4% to 17.9% (older age, male sex, valvular heart disease, and hypertension). In these clinical trial populations, most patients with 1 instance of AF stayed on ibrutinib, although multiple episodes were more common among ibrutinib-treated patients. This observation contrasts with a real-world experience in which complications, including congestive heart failure (CHF) and bleeding, were much more common and in which almost half the patients discontinued ibrutinib.³⁰  In evaluating patients for risk on ibrutinib, however, a history of major cardiac disease such as CHF or unrevascularized coronary artery disease is of greater concern to me than is AF, which can often be managed. Among the former patients, I have seen cases of elevated troponin or decompensated heart failure soon after starting ibrutinib. Our group recently reported rare cases of ventricular arrhythmias and sudden death associated with ibrutinib, and concern remains that risk of this serious outcome would be higher with significant cardiac disease.³¹  Prior to initiation, even difficult-to-control hypertension should be managed, because patients very commonly develop worsened hypertension on ibrutinib.^1,6 

Ibrutinib is <1% renally excreted, and exposure is not altered with mild-moderate renal impairment (creatinine clearance > 25 mL/min). A published case report describes the successful use of ibrutinib in a hemodialysis patient.³²  However, caution is in order, because ibrutinib has rarely been associated with causing renal impairment.

In mild, moderate, or severe hepatic impairment, single-dose ibrutinib pharmacokinetics showed mean increases in exposure of 4.1, 9.8, and 13.4 times, on the basis of the area under the curve (AUC), leading to the recommendation of 140 mg daily in patients with mild-moderate impairment, whereas a single daily dose was considered too high for severely impaired patients. Ibrutinib-induced acute liver failure has been reported,³³  as has reactivation of hepatitis B (HBV),^34,35  which should be screened for prior to initiation. Hepatologist referral and therapy for untreated chronic active hepatitis B (ie, surface antigen positive) is needed prior to ibrutinib, whereas patients who are hepatitis B surface-antigen negative and HBV DNA negative, but anti–hepatitis B core positive, can be monitored closely during therapy, including with HBV DNA testing.³⁶ 

Ibrutinib is hepatically metabolized, a major substrate of CYP3A4 and a minor substrate of CYP2D6. Modeling of ibrutinib drug-drug interactions show that coadministration with a strong inhibitor such as ketoconazole may increase ibrutinib AUC exposure by 24-fold, and moderate inhibitors led to a 4.9- to 7.5-times increase. Strong inducers such as rifampin can decrease AUC by 10-fold. Administration of ibrutinib with strong CYP3A4 inhibitors or inducers is therefore not recommended, though if a moderate inhibitor is absolutely required, the ibrutinib dose should be reduced to 140 mg daily.³⁷  Key moderate CYP3A4 inhibitors include diltiazem, verapamil, amiodarone, fluconazole, and voriconazole. Consumption of grapefruit and Seville oranges, as well as some supplements, also alter exposure of CYP3A4-metabolized drugs such as ibrutinib. A recent study from the Mayo Clinic found that 64% of 118 CLL patients treated with ibrutinib were on medications that could increase toxicity through drug interactions, the majority of which were antiplatelet medications, but 18% were drug interactions through CYP3A.³⁸  Thus a careful medication review prior to ibrutinib prescription is important to avoid AEs such as one reported with coadministration with the moderate CYP3A4 inhibitor verapamil.³⁹ 

Infections, in particular pneumonia, are common in the clinical trials of ibrutinib, with grade 3+ pneumonia in 25% and grade 3+ infection in 51% of relapsed refractory patients on the phase 1b study.⁵  The infection rate has been reported to decline by more than half after 6 months of therapy, although what happens after many years is unknown.^1,6,40  Ibrutinib has been reported to allow for partial reconstitution of normal B cells,⁴⁰  and immunoglobulin G (IgG) levels remain stable for the first 6 months of therapy^5,40  before declining thereafter.⁴⁰  IgA levels increase,^5,40  and a greater increase was associated in one study with a lower rate of infections.⁴⁰  Preclinical data also suggest that ibrutinib may help repair the T-cell defects in CLL.⁴¹  Multiple studies have reported that ibrutinib decreases Th2 cytokines in vivo,^41-43  normalizes total T-cell numbers,⁴²  and decreases T-regulatory cells⁴³  during the first 6 months of therapy.

Despite these observations, as ibrutinib is widely used, it is clear that susceptibility to infection remains significant. For example, cases of invasive aspergillosus, including 3 central nervous system (CNS) cases,⁴⁴  and Cryptococcus have been reported.^45,46  Seven of 18 patients with primary CNS lymphoma treated on a phase 1b study of ibrutinib followed by chemotherapy developed invasive aspergillosus, 2 while on ibrutinib alone and 4 with CNS disease.⁴⁷  In this study, BTK knockout and wild-type mice were infected with Aspergillus, and the former had significantly greater mortality, suggesting that BTK is needed for immune control of Aspergillus.⁴⁷  Increased vigilance for fungal infection is therefore warranted in ibrutinib-treated patients, but prophylaxis would be challenging given the duration of therapy and the CYP3A4 interactions with azoles.

Other infections are also seen. Our infectious diseases group has recently reported a 8.1% rate of opportunistic infections among hematologic malignancy patients receiving first-line BTK inhibitors.⁴⁸  In the ibrutinib phase 1b/2 study, 5 patients experienced varicella-zoster virus (VZV) reactivation despite antiviral prophylaxis at some point in 85 out of 101 patients⁶ ; subsequent reports confirm this.⁴⁹  Mild cases of pneumocystis jiroveci pneumonia (PJP) identified predominantly by polymerase chain reaction were reported in 5 patients out of 96 receiving single-agent ibrutinib, at a median of 6 months on therapy.⁵⁰  Although PJP was clearly the cause of symptoms and related to ibrutinib, 3 of the patients did not immediately receive secondary prophylaxis, and none recurred, leading the authors to suggest increased vigilance as opposed to universal prophylaxis.⁵⁰  Even so, I have continued my practice of using prophylaxis for both VZV and PJP in any relapsed CLL patient on any therapy. Given the cases noted above, I now usually do so in previously untreated patients receiving ibrutinib who are without contraindication, but many other CLL specialists do not. More data are needed to clarify the risk prior to a universal recommendation.

What about vaccinations? One study of influenza vaccination in 19 patients on single-agent ibrutinib found that 26% had seroconversion to at least 1 vaccine strain.⁵¹  A subsequent study looking at 13 relapsed CLL patients on ibrutinib for a median of 7.5 months found no responders to influenza vaccine.⁵²  A similar study of 13-valent pneumococcal conjugate vaccination found that all 4 untreated patients responded, whereas none of the 4 ibrutinib patients did, suggesting that if it is possible to vaccinate prior to starting ibrutinib, this is preferred.⁵³  My practice is to vaccinate prior to ibrutinib if feasible, but if not, to still provide vaccinations to patients on ibrutinib, because data are limited, and any response is helpful.

Autoimmune complications in CLL are common. Ibrutinib is thought to have the potential to control autoimmunity because of its activity in a rheumatoid arthritis mouse model,⁵⁴  and because it decreases differentiation of Th17 T cells⁴²  and normalizes expanded CD8 T-cell compartments⁵⁴  in patients. Multiple case reports demonstrate ibrutinib effectively controlling steroid refractory autoimmune hemolytic anemia,^55,56  and a report from the RESONATE trial⁵⁷  supports safe administration of ibrutinib with the suggestion of enhanced control of autoimmunity. Another study reported acute flare of autoimmunity right after ibrutinib initiation, with longer-term control in patients maintained on ibrutinib.⁵⁸  My experience largely recapitulates this report, namely, occasional flare of autoimmunity after ibrutinib initiation but control over the long term. I therefore treat the autoimmunity first to optimize control prior to initiation of ibrutinib and then overlap the autoimmune therapy with ibrutinib for some time. Similarly, if a patient flares shortly after starting ibrutinib, I will institute therapy for the autoimmune complication while continuing ibrutinib and slowly taper the other autoimmune therapy.

This question is challenging, because choice of therapy always requires balancing the risk of the patient’s CLL with the risk of a given therapy. Ibrutinib is sufficiently effective that in very high risk CLL, it may need to be tried regardless of contraindication, particularly because it is typically more readily available than venetoclax or idelalisib. That being said, the following constitute circumstances under which I would think very carefully before starting or resuming ibrutinib: (1) ongoing, poorly controlled bleeding; (2) history of major life-threatening bleeding, especially if not due to a reversible cause; (3) requirement for high treatment dose anticoagulation that cannot be stopped even briefly; (4) older patient (>70-75 years old) with cardiac comorbidity, including systolic or valvular dysfunction, recurrent uncontrolled congestive heart failure or AF, symptomatic unrevascularized coronary artery disease, or history of ventricular arrhythmia without automatic implantable cardioverter-defibrillator; (5) unexplained syncope concerning for cardiac etiology after initiation of ibrutinib; (6) ongoing invasive fungal infection requiring therapy; and (7) refractory autoimmune complications developing soon after ibrutinib initiation, particularly bone marrow aplasia.

Patient 1 was a 90-year-old woman who presented with CLL/small lymphocytic lymphoma (SLL) in 2001 and received 6 lines of therapy prior to January 2014, when she had a >25-cm abdominal mass, and was started on ibrutinib 420 mg daily. She experienced mild diarrhea, increased low-extremity edema, and arthralgia, which resolved after a few months.

Symptoms after starting ibrutinib include diarrhea, which is usually self-limited and easily managed with supportive care, as well as occasional nausea; diffuse body, joint, or muscle aches that vary from mild to severe; and ecchymoses or petechiae with minimal trauma (particularly when combined with antiplatelet drugs). Some patients also report significant fatigue and loss of concentration or attention on ibrutinib. Taking ibrutinib at night helps prevent the gastrointestinal side effects. The body aches or migratory arthralgias that affect 1 joint 1 day and an entirely different joint the next can be quite troublesome and are a frequent reason for ibrutinib discontinuation in clinical practice.⁵⁹  These aches often abate without major intervention. Magnesium supplements or tonic water containing quinine can be helpful. For severe arthralgias, management is anecdotal and variably effective, with options including a brief methylprednisolone taper, anti-inflammatory drugs (without antiplatelet effects), or drug holiday or dose reduction. The latter is the only real option for fatigue or attention problems.

The ibrutinib clinical trials recommended holding ibrutinib for grade 3+ toxicities until resolution to grade 1, followed by resuming at full dose at least once. The goal was to maintain full BTK occupancy (considered >95%), which is predicted in only 26% to 51% of patients treated at 140 to 280 mg daily.⁶⁰  Clinical outcome data at lower doses are limited. One retrospective investigation of 197 ibrutinib-treated patients reported similar objective response rate, PFS, and overall survival in 37 patients (19%) receiving a reduced median dose of 4.3 mg/kg/d.⁶¹  The UK and Ireland expanded access program also did not report reduced OS in patients who had dose reductions.¹¹ 

The literature suggests that drug holds may be more problematic, with decreased PFS among ibrutinib patients who missed >8 consecutive doses on the RESONATE registration trial.⁶²  This subgroup of patients had reduced creatinine clearance and advanced stage disease, however, and may have been likely to do poorly anyway. Progression was also defined by the independent review committee solely based on imaging.⁶²  The expanded access programs show conflicting data on drug holds, with no effect in the Polish experience but worse outcomes in the UK-Ireland experience, with greater than 14-day holds.¹¹ 

Although overall, these data appear reassuring, particularly for dose reductions, follow-up time remains short, and caution is still in order because of concern that inadequate BTK occupancy could potentially drive resistance. Ideally, an assay to measure BTK occupancy would be available clinically, but this does not exist. A key question is whether dose reduction helps with symptoms, and in my personal experience, patients may experience some improvement but not always a marked difference. Systematic studies addressing these issues are very limited and desperately needed. In the meantime, I encourage patients to stay at full dose for the first few months, because many of these early side effects resolve, and only after that to consider dose reduction if the symptoms remain challenging.

A 67-year-old man with immunoglobulin heavy chain variable region (IGHV) unmutated, del(11q) deleted CLL with complex karyotype had a 4-year remission after fludarabine, cyclophosphamide, and rituximab in 2011, but by mid-2016 had white blood cell (WBC) count of 108 000/μL, hemoglobin of 10 g/dL, and platelet count of 69 000/μL. He started ibrutinib, and his WBC rose to 250 000/μL before slowly declining. He had a history of premature ventricular contractions (PVCs), which increased despite metoprolol. An event monitor demonstrated PVCs, premature atrial contractions, and brief atrial tachycardia, and he was put on aspirin, which in combination with ibrutinib and platelets at 60 000/μL, resulted in significant bruising.

An increase in circulating lymphocyte count is common and peaks at 1 to 2 months, followed by a slow decline, but a subset of patients never resolve their lymphocytosis fully.⁶³  Data show that patients with prolonged lymphocytosis have outcomes as good as those who do not (likely because prolonged lymphocytosis is associated with favorable prognostic factors).⁶³  This sharp early increase in lymphocyte count should not be mistaken for progressive disease in a patient who is symptomatically improved and whose nodes are decreased. It is also important to note that the improvement in anemia on ibrutinib can be quite slow, sometimes requiring 4 to 6 months, although stabilization is usually seen early. Thus lymph node shrinkage is the simplest early indicator of response. Progression in the first 1 to 2 years on ibrutinib is often Richter’s transformation, while CLL progressions tend to occur later.¹⁰  A slow, steady increase in lymphocyte count at later times should raise the possibility of progressive disease. The identification and management of progressive disease on ibrutinib has been well covered in a previous “How I Treat” article.⁶⁴ 

Also very common after initiation of ibrutinib is a drop in platelet count, which is typically not problematic, because it is usually of small magnitude and remains stable prior to slowly improving over time. Because of this drop and the antiplatelet effect of ibrutinib, I am cautious when initiating ibrutinib in a patient with a platelet count below 30 000/μL (clinical trial cutoff), and I will usually use a higher platelet transfusion threshold of about 25 000/μL shortly after initiation. These patients, in particular, should have all other anticoagulant drugs stopped.

A few months later, patient 2 presented with fatigue and different palpitations and was in AF, which converted spontaneously. His CHADS2 (congestive heart failure, hypertension, age ≥ 75, diabetes, previous stroke/transient ischemic attack) VASc (vascular, age 65-74, sex) score was 2 (age and hypertension), potentially warranting anticoagulation. He already had experienced significant bruising, with platelets still in the 60 000s/μL. Cardiology therefore felt that his bleeding risk likely exceeded his stroke risk but, given the ongoing need for ibrutinib in a high-risk CLL patient, decided to suppress the burden of AF with amiodarone. Given the drug interaction, the ibrutinib dose was reduced to 140 mg daily. With amiodarone, the patient has not had AF or PVCs and feels well. His CLL remains in good control, and he continues on low-dose aspirin.

Patient 3 was a 70-year-old man with CLL, with del(11q) and del(13q) deletions and unmutated IGHV, who was treated in early 2009 on CALGB 10404 with fludarabine and rituximab, with complete response. His disease progressed, and he started ibrutinib in June 2016, complicated by easy bruising. At a routine preoperative evaluation, he was in AF and started on metoprolol and apixaban, and ibrutinib was held. He underwent cardioversion successfully, and ibrutinib was resumed, but 2 weeks later, he was again in AF. He was asymptomatic and restarted on low-dose apixaban without concurrent antiplatelet agents. He was counseled about increased risk of bleeding, and he elected to continue both drugs.

Several months after starting ibrutinib, patient 1 presented with syncope and was found to have severe bradycardia, for which a pacemaker was placed. She continued on ibrutinib with the pacemaker.

Patient 1 illustrates that older patients, in particular, can develop a range of cardiac complications not well represented in the younger trial patients. The most common cardiac complication is AF, though, and I will typically hold ibrutinib at least briefly while controlling the AF. The long-term management decisions are complex, as in the 2 cases, and cardiologists are helpful with the risk-benefit assessment and decision making. Systematic data to guide management are lacking, but recent reviews have proposed management guidelines consonant with my approach.^65,66  Although not studied in these patients, the CHADS2 VASc score for clotting risk and the HAS-BLED (Hypertension, Abnormal renal/liver function, Stroke, Bleeding history or predisposition, Labile INR, Elderly, Drugs/alcohol concomitantly) score for bleeding risk can provide some guidance when considering anticoagulation. Whether or not to continue ibrutinib also depends on the patient’s remission status and the underlying aggressiveness of his or her disease, as well as patient preference. Ibrutinib might in principle provide some helpful antiplatelet activity in cardiac disease, but at present this is completely unknown. For those patients who do continue ibrutinib and anticoagulation, our group has favored apixaban, given the cardiology and thromboembolism literature, which suggest that apixaban may result in fewer bleeding events than warfarin or other new oral anticoagulants.^67-70  That being said, almost no safety data exist for this combination, and ibrutinib is a P-glycoprotein inhibitor that may increase the concentration of apixaban, leading us to use the lower 2.5-mg twice-daily dose in many patients also on ibrutinib. I typically maintain the ibrutinib dose, because most patients in the clinical trial series tolerated continuing full dose.²⁹ 

She had near resolution of her bulky abdominal mass on ibrutinib but developed a pulmonary embolism after a cross-country plane flight in December 2015 and received anticoagulation without holding ibrutinib. She also had an infected leg ulcer with osteomyelitis and experienced clinically significant recurrent bleeding from this wound until ibrutinib was held. She was then quite debilitated and stayed off ibrutinib without a problem for nearly 2 years, at which time she was off anticoagulation and resumed ibrutinib with good response of her recurrent abdominal adenopathy.

Despite all precautions, some patients develop clinically significant bleeding. My management typically involves holding ibrutinib and providing platelet transfusions as needed, regardless of platelet count, to overcome the platelet function defect induced by ibrutinib. Case reports of the benefit of platelet transfusion have been published, although data are sparse.⁷¹  A recent randomized trial in the context of other antiplatelet agents suggested that platelet transfusion may actually increase mortality in CNS bleeding.⁷²  Given the absence of data with ibrutinib, management should be individualized.⁶⁵  Holding ibrutinib alone will typically substantially resolve the antiplatelet effects in 2 to 3 days.^17,18  I planned for patient 1 to complete 6 months of anticoagulation for her pulmonary embolus, and during that time I would not have added ibrutinib back. Whether or not to continue ibrutinib in a patient with a bleed is an individualized decision that depends on the bleed severity, the reversibility of its cause (periprocedural or transient anticoagulation, for example), the quality of the CLL response, and the aggressiveness of the underlying CLL. This patient did well off of ibrutinib and was subsequently able to resume ibrutinib when her other issues had resolved.

This 75-year-old woman first presented with CLL, complicated by cold agglutinin disease, in 1992, and despite multiple rounds of therapy, was dependent on weekly transfusions by July 2014, when she started ibrutinib. She had an excellent response and became transfusion independent. In December 2016 she presented with acute bilateral vision loss, right hemiparesis, and dysarthria after 1 week of sinus symptoms and headache. She was diagnosed with cavernous sinus thrombosis and suffered rapid neurologic deterioration. Invasive mucormycosis involving both cavernous sinuses was confirmed at autopsy.

Infection is one of the most common AEs for patients on ibrutinib, especially in the first 6 months of therapy.⁶  Any patient presenting with likely pneumonia or systemic infection should have a complete workup, including evaluation for opportunistic infections. Typically, if a patient presents acutely with fever and is hospitalized, I will hold ibrutinib until a diagnosis is established and the patient is clearly improving. Ibrutinib can cause a drug-related pneumonitis responsive to prednisone, reported in a small series⁷³  and seen occasionally in clinical practice. Lung biopsies demonstrated organizing pneumonia and interstitial inflammation or granulomas without organisms.⁷³  Whether ibrutinib can be resumed in these patients is not clear; 1 patient in this series resumed ibrutinib and developed recurrent interstitial pneumonitis.⁷³  For most patients with typical bacterial or viral pneumonia, however, ibrutinib can be resumed once the patient is improving. An exception may be patients who develop invasive fungal infections on ibrutinib, who will need to be comanaged with infectious disease. Ibrutinib has significant drug interactions with both voriconazole and posaconazole. The alternative agent isavuconazole has less severe drug interactions but is less proven. In limited anecdotal experience, these patients often do need to discontinue ibrutinib for an extended period to enable full control of their infection.^44,47  More data on the natural history of opportunistic infections and long-term incidence of infections on ibrutinib are desperately needed.

Rash is a common ibrutinib side effect in 2% to 27% of patients^3,5,24,28  but is often self-limited or easily managed with topical steroids.^74,75  A case series from Stanford University of 14 patients has been published, in which 2 categories of rash were identified.⁷⁴  The first was a nonpalpable asymptomatic petechial rash, perhaps related to ibrutinib-induced platelet dysfunction, and the second was a palpable pruritic rash, which on biopsy showed an inflammatory infiltrate. The latter sometimes required ibrutinib interruption and corticosteroids to resolve, but all patients were eventually able to resume ibrutinib.⁷⁴  More recently, erythema nodosum has been seen in ibrutinib-treated patients and may require systemic steroids or extended time off of ibrutinib for resolution.

Finally, brittle nails and hair are common on ibrutinib and increase over time. In a National Institutes of Health report, 67% of patients reported brittle fingernails at a median of 6 months on ibrutinib, with about a quarter reporting brittle toenails or hair changes at a median of 9 months.⁷⁶  We generally recommend nail oil to the cuticle bed.

Over time, hypertension increases in patients on ibrutinib and often requires initiation or increase of drug therapy. The incidence is in the 5% to 14% range^1,2  at 1 to 2 years but rises to 25% to 30% with 5-year follow-up in clinical trials.^6,7  This incidence may be even higher outside of trials and with longer follow-up, requiring vigilance on the part of treating physicians.

Ibrutinib is a highly effective therapy for CLL, which at present is planned to be given indefinitely, yet we still have much to learn about its side-effect profile over time. Managing toxicities to keep patients on the drug long enough to achieve a deep response is critical to ibrutinib benefit, and at present, little is known about the potential durability of response after ibrutinib discontinuation for AEs. For patients who come off of ibrutinib for AEs, I typically try to delay subsequent therapy as long as possible while the AEs resolve; sometimes this observation period can last years. For patients who received ibrutinib only briefly frontline, I would typically favor trying chemoimmunotherapy (CIT) next, although we have almost no data on its efficacy after ibrutinib. In the relapsed setting, I would likely use either venetoclax or idelalisib at progression, the former more likely if the patient had been on ibrutinib for some time or progressed while taking ibrutinib. I have, however, had good success with idelalisib, particularly in older patients with significant cardiac disease who are pretreated with CIT. Just as we need a better understanding of ibrutinib side effects over time, we need more information on the outcomes of these subsequent therapies, particularly CIT, in patients who received ibrutinib frontline. Ultimately, the development of combination shorter-duration regimens that result in deep remissions may allow treatment breaks that could reduce both short- and long-term toxicities in those who develop them. These ibrutinib combinations could be based on anti-CD20 antibodies such as obinutuzumab,⁷⁷  particularly in patients who are frailer, but will likely involve venetoclax in fit patients because of its efficacy in inducing complete response and minimal residual disease negativity.^78,79  Minimizing toxicity is just one of the benefits of moving toward time-limited therapy, along with patient preference, reduced cost, and reduced resistance. Another potential advance will be the upcoming availability of more specific BTK inhibitors, including acalabrutinib⁸⁰  and BGB-3111,⁸¹  which may have fewer side effects, although experience at present remains limited. Meanwhile, however, we need systematic population-based studies of the natural history of ibrutinib therapy over time to optimally inform our patient management.

The author thanks Jeffrey Zwicker for the initial idea for this article.

Contribution: J.R.B. developed the concept, drafted the paper. and approved the final manuscript.

Conflict-of-interest disclosure: J.R.B. has served as a consultant for Janssen, Pharmacyclics, Astra-Zeneca, Sun, Redx, Sunesis, Loxo, Gilead, TG Therapeutics, Verastem, and Abbvie and receives research funding from Sun and Gilead.

Correspondence: Jennifer R. Brown, Department of Medical Oncology, Dana-Farber Cancer Institute, 450 Brookline Ave, Boston, MA 02215; e-mail: jennifer_brown@dfci.harvard.edu.