Prognostic Impact and Risk Factors of Infections in Patients with Chronic Lymphocytic Leukemia Treated with Ibrutinib

Simple Summary Ibrutinib demonstrated superior efficacy compared to chemoimmunotherapy in patients with chronic lymphocytic leukemia. However, adverse events are a frequent reason for treatment discontinuation. This study was aimed to evaluate the incidence, risk factors, and prognostic impact of infections in a large series of patients with chronic lymphocytic leukemia who received an ibrutinib-based therapy. Approximately one-third of patients developed pneumonia or a severe infection with an overall rate of 15.3% infections per 100 person-year. Patients who experienced a severe infection in the year before starting ibrutinib, those with chronic obstructive pulmonary disease, and those heavily pretreated showed greater vulnerability to infection. A scoring system based on these factors identified patients with a two- to threefold increase in the rate of infections. Infections showed an unfavorable impact in terms of treatment discontinuation and inferior survival. Our results demonstrate that infections are a relevant reason for treatment failure in patients treated with ibrutinib. Abstract Ibrutinib represents extraordinary progress in the treatment of chronic lymphocytic leukemia (CLL). However, treatment-related adverse events limit the benefit of this agent. This observational, multicenter study focused on the incidence, risk factors, and prognostic impact of infections in 494 patients with CLL treated with an ibrutinib-based treatment. Ibrutinib was given to 89 (18%) previously untreated patients (combined with rituximab, 24) and 405 (82%) relapsed/refractory patients. Pneumonia (PN), grade ≥3 non-opportunistic infections (NOI), and opportunistic infections (OI) were recorded in 32% of patients with an overall incidence rate per 100 person-year of 15.3% (PN, 10%; NOI, 3.3%; OI, 2%). Infections were the reason for the permanent discontinuation of ibrutinib in 9% of patients. Patients who experienced pneumonia or a severe infection showed a significantly inferior survival than those who were infection-free (p < 0.0001). A scoring system based on the three factors associated with a significant and independent impact on infections—PN or severe infection in the year before starting ibrutinib, chronic obstructive pulmonary disease, ≥2 prior treatments—identified patients with a two- to threefold increase in the rate of infections. In conclusion, the results of this study highlight the adverse impact of infectious events on the outcomes of CLL patients treated with ibrutinib.


Introduction
Chronic lymphocytic leukemia (CLL) is the most common leukemia in the adult population. According to the National Cancer Institute's Surveillance, Epidemiology, and End Results Program (SEER) [1], the estimated number of new cases of CLL in the United States in 2021 is 21,250. The incidence rapidly increases with increasing age, and the median age at diagnosis is about 70. The CLL-IPI combines genetic, biochemical, and clinical parameters into a prognostic model, discriminating four prognostic subgroups with 5-year survival probabilities ranging from 93% to 23% [2].
Several randomized trials have demonstrated the superiority of the B-cell receptor pathway inhibitor ibrutinib over chemotherapy in patients with chronic lymphocytic leukemia (CLL). [3][4][5][6]. The remarkable activity and the favorable tolerability profile have favored the widespread use of ibrutinib for the upfront treatment of patients with CLL. However, treatment discontinuation due to specific off-target toxicities of this agent, such as bleeding events and atrial fibrillation, is not negligible, although some treatment-related adverse events are now better managed [7][8][9][10][11][12][13][14][15].
It is well known that defects in humoral and cellular immunity due to CLL itself and chemoimmunotherapy result in an increased risk of infections that are a significant cause of morbidity and mortality in CLL patients [16]. Recently, an increased rate of infections has also been described in patients treated with ibrutinib, particularly in R/R patients. As a lower risk of infectious events was expected with targeted agents than chemoimmunotherapy, several reports have focused on infections in patients treated with ibrutinib [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34]. Although several inhibitory effects exerted by ibrutinib on immune effector cells have been described to explain the increased fragility to infections (particularly fungal infections [35][36][37][38][39][40][41][42]), little is known about the prognostic impact of infections in patients treated with ibrutinib. Moreover, the role of clinical and biological factors in favoring infections in patients treated with ibrutinib should be better defined.
This retrospective, multicenter study was carried out to define the incidence of pneumonia and severe infections in a large series of patients with CLL treated with ibrutinib.
This study was also aimed at investigating risk factors and the impact of infections on treatment discontinuation and survival of patients receiving ibrutinib.

Materials and Methods
This study analyzed retrospectively the characteristics of 494 patients with CLL treated with ibrutinib at 16 Italian institutions, 12 academic centers, and four hospitals.
Inclusion criteria were CLL diagnosis according to the iwCLL criteria [43] and frontline or advanced-line treatment with ibrutinib-based therapy having started between February 2013 and February 2019. Exclusion criteria included known transformation from CLL to an aggressive lymphoma (i.e., Richter transformation) and clinically significant comorbidities potentially interfering with the regular administration of treatment.
Data of patients were collected from medical records by the referring physician and included demographics, comorbidities, infectious events within 12 months before starting ibrutinib, prior treatment, serum IgG level, antimicrobial prophylaxis, concomitant use of steroids for more than four weeks, the IGHV andTP53 mutation status, and the FISH profile.
Three types of infections were considered: pneumonia (PN), grade ≥ 3 non-oppor tunistic infections (NOIs), and opportunistic infections (OIs). When available, the etiologic agent was recorded. Infections of non-bacterial etiology with lung involvement were classified as NOI or OI, according to the identified agent. The severity of infections was graded according to the Common Terminology Criteria for Adverse Events, version 4.0.
The primary endpoint of this study was the incidence of PN, NOIs, and OIs in patients receiving ibrutinib. The secondary endpoints were the impact of infections on treatment discontinuation, survival, and factors associated with an increased rate of infectious events.
The database was locked on March 30, 2020, for analysis. The person-year duration of ibrutinib exposure was defined for each patient as the time from the start of ibrutinib to the last taken dose, last follow-up, or death. The person-year incidence rate of infections was defined as the number of infectious events observed during treatment divided by the total person-years of ibrutinib exposure. Survival was calculated from the start of treatment to the infectious event, disease progression, Richter syndrome, or the last follow-up or death. Survival curves were calculated according to the Kaplan and Meier method and differences in survival using the log-rank test in univariate analysis and the Cox regression model in multivariate analysis, after the assessment of the proportionality of hazards. Variables associated with a significant and independent increase in the infection rate were selected using a stepwise forward method, based on Wald statistics [44]. A scoring system to assess the risk of infections was developed by assigning to each significant variable in the final model a weighted point based on its hazard ratio, proportional to the regression β coefficient. The bootstrap method was used to validate the model. Prediction performance has been evaluated by the area under the curve (AUC) of the receiver operating characteristic (ROC) curve. Confidence intervals (CIs) have been calculated at the 95% level. All statistical tests were two-sided. A p-value of less than 0.05 has been considered significant. All analyses have been performed in SPSS v26.
This observational study was conducted in accordance with the Declaration of Helsinki and approved by the institutional review board and ethical committee.

Clinical Characteristics of Patients
The baseline characteristics of the 494 CLL patients included in this study are detailed in Table 1. The median time on ibrutinib was 35 months (range, 4-85 months) and the median duration of the exposure to ibrutinib per patient was 2.6 years. The median age of patients was 69 years (range, 32-92 years). A CIRS >6 was observed in 24% of patients and creatinine clearance <70 mL/min in 37.4%. A chronic obstructive pulmonary disease was recorded in 16% of patients, and 24% had a grade ≥3 severe infection or a pneumonia event in the year before starting ibrutinib. Twenty-one percent of patients showed IgG levels ≤ 400 mg/dl and were on immunoglobulin support and 8% had a baseline granulocyte count <1 × 109/L. Rai stage III-IV was present in 53% of patients, unmutated IGHV status in 73.5%, and TP53 disruption (del17p, and or TP53 mutation) in 55.2%. Eighty-nine (18%) of the patients received front-line therapy with ibrutinib, 420 mg once daily given continuously and 24 received also rituximab, 375 mg/sqm, weekly on day 1 of month 1 and then on day 1 of months 2-6. After a median number of one prior treatment (range, 1-9), 405 (82%) relapsed/refractory patients received ibrutinib single agent, 420 mg once daily given continuously. Concomitant steroids have been administered to 17% of patients.
Infection prophylaxis did not follow common guidelines but rather the guidelines of each institution. Seventy-four percent of patients were on Pneumocystis jirovecii (PJ) prophylaxis with trimethoprim-sulfamethoxazole (TMP-SMX), and 53 (8%) patients with a known risk of hepatitis B reactivation (HBc+ and, or HBsAg+) received lamivudine.

Incidence of Infections
One hundred and fifty-six (32%) patients experienced at least one infectious event and 35 (7%) had an additional infection. The total number of infections we recorded was 193 with an overall incidence rate of 15.3% infections per 100 person-year. Pneumonia was the most common infection with an incidence per 100-person per year of 10%, while it was 3.3% for grade ≥3 non-opportunistic infections, and 2.0% for opportunistic infections. The median time from the start of ibrutinib and the onset of infections was six months, seven for pneumonia events, nine for grade ≥3 non-opportunistic infections, and three for opportunistic infections (Table 2).

Type of Infections
The incidence and type of infections are described in detail in Table S1. One hundred (20%) patients experienced at least a pneumonia event, 32 (6.5%) a grade ≥3 non-opportunistic infection, and 24 (5%) a grade ≥3 opportunistic infection. Among the 32 patients who developed a grade ≥3 non-opportunistic infection, 7 viral infections were recorded. HBV reactivation was described in three patients in whom the serological profile for HBV had not been evaluated at baseline while none of the 41 patients on HBV prophylaxis due to a known risk for HBV reactivation developed an HBV reactivation. Symptomatic COVID-19 pneumonia was diagnosed in two young patients with hypogammaglobulinemia and was fatal in one. Fungal infections were the most frequent type of opportunistic infection (14/24; 58%), and, in turn, Aspergillus infection was the most frequent type of fungal infection (11/14; 79%). Aspergillus infection was associated with SNC involvement in three cases and was the direct cause of death for 6/11patients. Other opportunistic infections included Cryptococcus infection in two patients, CMV infection in three, disseminated HVZ infection in five. Pneumocystis jirovecii pneumonia (PJP) was diagnosed in a previously treated patient who received trimethoprimsulfamethoxazole irregularly. This accounts for 0.8% (1/129) of PJP cases among patients who did not receive appropriate prophylaxis. Mycobacterium infection was described in two asymptomatic patients who underwent the biopsy of an isolated lung nodule. A fatal, progressive multifocal leukoencephalopathy was recorded in a 79-year-old man with advanced disease treated with ibrutinib and rituximab.

Outcomes of Patients Who Developed Infections
Forty-three (9%) patients discontinued ibrutinib permanently due to an infection. Treatment discontinuation due to infection was an early event observed in 20 (4%) patients in the first year of treatment, in 10 (2%) in the second year, and 13 (3%) thereafter.
The other most frequent reasons leading to permanent ibrutinib discontinuation were disease progression in 16% of patients, Richter syndrome (RS) in 5%, atrial fibrillation in 5%, and second malignancies in 4% (Table S2).
Patients who experienced pneumonia or a severe infection showed a significantly inferior survival than those who were infection-free (36-month OS: 82% vs. 63%; p < 0.0001; Figure 1A). The survival probability of patients who discontinued ibrutinib due to an infection or Richter syndrome was significantly inferior to that of those who discontinued treatment due to disease progression (median survival: 2 vs. 6 vs. 9 months; p = 0.006) ( Figure 1B).

Impact of Baseline Characteristics of Patients on Infections
The incidence rates per 100 person-year of infections according to patients' clinical and biologic characteristics are summarized in Table 3. Baseline factors associated with a

Impact of Baseline Characteristics of Patients on Infections
The incidence rates per 100 person-year of infections according to patients' clinical and biologic characteristics are summarized in Table 3. Baseline factors associated with a significantly higher rate of infections were chronic obstructive pulmonary disease (p <0.0001), pneumonia or grade ≥3 infection in the year before starting ibrutinib (p < 0.0001), IgG levels ≤400 mg/dl (p = 0.002), neutrophil count <1 × 109/L (p = 0.001), Rai stage III-IV (p < 0.0001), prior treatment (0 vs. ≥1, p = 0.004; 1 vs. ≥ 2; p < 0.0001), and the addition of steroids to ibrutinib (p = 0.002) ( Table 3). The addition of rituximab to front-line therapy with ibrutinib, the IGHV mutational status, and the presence of TP53 aberrations did not reveal any impact on the infection rate.
In multivariate analysis, three factors maintained significance: a pneumonia event or a severe infection in the year before starting ibrutinib (HR, 2.69 (95% CI, 1.90-3.76)), the presence of a chronic obstructive pulmonary disease (HR, 1.52 (95% CI, 1.03-2.25)), two or more prior treatments (HR, 1.63 (95% CI, 1.17-2.28)) ( Table 4).   To predict the risk of infections we designed a scoring system based on the HR values of factors with a significant and independent impact on infections. Two points were assigned for the presence of a pneumonia event or a severe infection in the year before starting ibrutinib, and one for chronic obstructive pulmonary disease, and two or more prior treatments. The sum of these scores showed an AUC of 0.65 (95% CI: 0.60-0.71). Three subgroups of patients were identified according to the score. The low-risk group (score, 0-1 points) identified 352 (71.2%) patients of whom 81 (23.0%) experienced an infectious event. The intermediate-risk group (score, 2 points) included 67 (13.6%) patients of whom 28 (41.8%) experienced an infection. Finally, the high-risk group (score, ≥3 points) identified 75 (15.2%) patients of whom 47 (62.7%) developed an infection. (Table 5). Table 5. Scoring system to assess the risk of infections in CLL patients treated with ibrutinib ± rituximab.

Discussion
Infections are a well-known cause of morbidity and mortality in CLL patients treated with chemotherapy. This study was carried out to evaluate the prognostic impact of pneumonia and severe infections in patients who receive ibrutinib.
Overall, with a median follow-up of 35 months, about a third of patients experienced at least one infectious event with an overall rate of 15.3% infections per 100 person-year. Given the large proportion of unselected and previously treated patients included in this study, the high infection-related morbidity was not unexpected as compared to that described in controlled trials [3][4][5][6]9,10,45,46]. Previous chemotherapy treatment revealed a significant impact in promoting greater fragility to infections in patients receiving ibrutinib.
As observed in other studies, infections were an early event that occurred after a median time of 7 months from starting ibrutinib [10,46]. This observation suggests careful monitoring of patients during the first months of treatment with ibrutinib. Pneumonia was recorded in 20% of patients with an incidence rate per 100 person-years of 10%. Variable rates of pneumonia events, ranging between 10% and 28%, have been reported and this could be related to the different characteristics and follow-up of patients included in other studies [3][4][5][6]9,32,34,46]. Two young, R/R patients with hypogammaglobulinemia developed symptomatic COVID-19 infection, discontinued treatment, and one did not survive the infection. The small number of patients with COVID-19 disease we recorded was due to the locking of the database for analysis before the pandemic became widespread in our country. At present, there is no agreement on the impact of ibrutinib on the outcomes of CLL patients who developed COVID-19 disease. Some studies report a lower number of COVID-19 infections than expected and more favorable outcomes in patients treated with ibrutinib [47][48][49][50][51]. However, the same favorable impact of ibrutinib has not been confirmed in a multicenter study [52].
The rate of opportunistic infections per 100 person-year recorded in this study, 2%, is in line with that reported by Rogers et al., 1.9% [23]. Consistent with other reports, the most common fungal infection was due to the Aspergillus fumigatus [24][25][26][27][28]. It has been argued that the inhibitory effects of ibrutinib on the activity of T-cells, natural killer cells, and macrophages may play a role in the development of these infections [36][37][38][39][40][41]. Given the relatively low incidence, routine prophylaxis of fungal infections has not been recommended. Similarly, the low incidence of PJP in this study, as well as in other studies [28,29], makes one also question the need for specific prophylaxis in patients receiving ibrutinib. However, the presence of an opportunistic infection should be kept in mind, in patients who develop clinical signs of a severe infection.
Treatment discontinuations due to infections were more frequently observed than those due to other treatment-related adverse events, such as atrial fibrillation and bleeding events. Patients who experienced pneumonia or a severe infection showed a significantly inferior survival than those who were infection-free. Moreover, patients who developed pneumonia or a severe infection showed a significantly inferior survival than those who progressed. These findings underline the unfavorable impact of infections and the need to identify factors associated with a higher risk of infectious events. Three baseline factors revealed a significant and independent impact on the infection risk. The first one, an infectious event in the year preceding the start of ibrutinib, emphasizes the impact of a pre-existing immunodeficiency in predisposing further infections. The second, the presence of a concomitant chronic obstructive pulmonary disease, is a well-known risk factor for respiratory infections. Finally, the third factor, heavy pretreatment, is also a well-known clinical condition associated with an increased risk of infections. In this study, pretreated patients showed double the incidence of infections compared to treatment-naive patients. Earlier use of ibrutinib can result in a lower rate of infectious events.
Concerns about infectious events led several scientific groups to recommend more attention to the infection risk of patients treated with BTK inhibitors and other targeted molecules [12][13][14][15]53]. Other BTK inhibitors, acalabrutinib, zanubrutinib, show a more favorable safety profile [54,55]. Interestingly, in contrast to ibrutinib, tirabrutinib does not show inhibition of T lymphocyte function and this could result in less susceptibility to infections [56].
The scoring system we designed requires an external validation cohort to confirm its predictive value. However, it identified about 30% of patients with a two-to threefold increase in the rate of infectious events. These more vulnerable patients could benefit from measures to prevent and mitigate the harmful impact of infections.

Conclusions
The results of this study show that infections are still a critical issue in the treatment management of CLL patients who receive ibrutinib and highlight the adverse impact of infectious events on treatment discontinuation and survival.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/cancers13133240/s1, Table S1. Infections recorded in patients treated with ibrutinib ± rituximab, Table S2. Events leading to permanent treatment discontinuation in 2% of patients treated with ibrutinib ± rituximab, Table S3. Cause of death in patients treated with ibrutinib ± rituximab.  Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the Università Sapienza (Rif. 5278-07/02/19)." Informed consent was obtained from patients involved in the study.
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: Data may be made available request for consideration to the study corresponding author and establishment of the data transfer agreement.
Conflicts of Interest: F.R.M. received research funding from Gilead, advisory board participation fees from AbbVie, Gilead, Janssen, AstraZeneca, Takeda, Roche; speaker bureau fees from Gilead, Janssen, and Abbvie. A.V. received advisory board participation fees from Gilead, Janssen, Italfarmaco, speaker bureau fees from Janssen and Abbvie. G.R. received advisory board participation fees from Gilead, Janssen, Abbvie, and AstraZeneca; speaker bureau fees from Gilead, Janssen, and Abbvie. P.S. received research funding from Gilead, advisory board participation fees from AbbVie, Janssen. S.C. received an honorarium for an advisory board or lecturing for AbbVie, Janssen, AstraZeneca; speaker bureau fees from Janssen and Abbvie. C.V. received consulting fees and advisory board participation fees from Janssen. L.L. (Luca Laurenti) aurenti received research funding from Abbvie and Roche; advisory board participation fees from AbbVie, Janssen, AstraZeneca; speaker bureau fees from Gilead, Janssen, Abbvie, AstraZeneca, Roche. L.D.P received advisory board participation fees from Amgen, Gilead, Janssen, Takeda, Celgene-Bristol, and consulting fees from Amgen. R.M. received grants from AbbVie, Gilead, Janssen. M.G. received advisory board participation fees from