Fluid retention-associated adverse events in patients treated with BCR::ABL1 inhibitors based on FDA Adverse Event Reporting System (FAERS): a retrospective pharmacovigilance study

Abstract

Objectives

This study aimed to conduct a thorough analysis of fluid retention-associated adverse events (AEs) associated with BCR::ABL inhibitors.

Design

A retrospective pharmacovigilance study.

Setting

Food and Drug Administration Adverse Event Reporting System (FAERS) database for BCR::ABL inhibitors was searched from 1 January 2004 to 30 September 2021.

Main outcome measures

Reporting OR (ROR) and 95% CI were used to detect the signals. ROR was calculated by dividing the odds of fluid retention event reporting for the target drug by the odds of fluid retention event reporting for all other drugs. The signal was considered positive if the lower limit of 95% CI of ROR was >1. The analysis was run only considering coupled fluid retention events/BCR::ABL inhibitors with at least three cases.

Results

A total of 97 823 reports were identified in FAERS. Imatinib had the most fluid retention signals, followed by dasatinib and nilotinib, while bosutinib and ponatinib had fewer signals. Periorbital oedema (ROR=24.931, 95% CI 22.404 to 27.743), chylothorax (ROR=161.427, 95% CI 125.835 to 207.085), nipple swelling (ROR=48.796, 95% CI 26.270 to 90.636), chylothorax (ROR=35.798, 95% CI 14.791 to 86.642) and gallbladder oedema (ROR=77.996, 95% CI 38.286 to 158.893) were the strongest signals detected for imatinib, dasatinib, nilotinib, bosutinib and ponatinib, respectively. Pleural effusion, pericardial effusion and pulmonary oedema were detected for all BCR::ABL inhibitors, with dasatinib having the highest RORs for pleural effusion (ROR=37.424, 95% CI 35.715 to 39.216), pericardial effusion (ROR=14.146, 95% CI 12.649 to 15.819) and pulmonary oedema (ROR=11.217, 95% CI 10.303 to 12.213). Patients aged ≥65 years using dasatinib, imatinib, nilotinib or bosutinib had higher RORs for pleural effusion, pericardial effusion and pulmonary oedema. Patients aged ≥65 years and females using imatinib had higher RORs for periorbital oedema, generalised oedema and face oedema.

Conclusions

This pharmacovigilance study serves as a clinical reminder to physicians to be more vigilant for fluid retention-associated AEs with BCR::ABL inhibitors.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The study is a retrospective analysis of adverse events from the real-world Food and Drug Administration Adverse Event Reporting System (FAERS) database, which provides relevant insights into fluid retention events associated with BCR::ABL inhibitors.

• Data mining was conducted by calculating the reporting OR (ROR) and 95% CI to detect signals.

• FAERS is a self-reporting system and there may exist problems with adverse event under-reporting, incomplete information and poor quality of reported data.

• Due to inherent database limitations, although a positive ROR value is a threshold used for statistical signal detection in FAERS, ROR is an association measure and does not necessarily imply causality.

Introduction

Chronic myeloid leukaemia (CML, chronic myelocytic or chronic myelogenous leukaemia) is a haematopoietic stem cell malignancy and is classified as a myeloproliferative disorder along with polycythaemia vera, essential thrombocythaemia and primary myelofibrosis.¹ CML accounts for 15% of adult leukaemia, with a global annual incidence ranging from 1.6 per 100 000 to 2 per 100 000.² Tyrosine kinase inhibitors (TKIs) can achieve long-term control of CML in the majority of patients, making them the first-line treatment option for almost all patients with newly diagnosed CML. Other treatment options are allogeneic haematopoietic cell transplantation and palliative treatment with cytotoxic agents, for patients who have failed TKI therapy or for those with advanced CML.³

CML is linked to the Philadelphia chromosome t(9;22)(q34;q11) creating a breakpoint cluster region and Abelson (BCR::ABL)1 fusion gene.⁴ This genetic abnormality leads to the formation of a distinct gene product (BCR::ABL1), a constitutively activated tyrosine kinase. This deregulated tyrosine kinase has been linked to the development of CML and has become a primary target for treatment.⁵ BCR::ABL inhibitors are multitargeted TKIs that act on the BCR::ABL fusion protein to inhibit signal transduction. This class of drugs includes imatinib, dasatinib, nilotinib, bosutinib and ponatinib.⁶ Imatinib is the first-generation TKI approved by the Food and Drug Administration (FDA) to treat various tumours, mostly in Philadelphia chromosome-positive CML (Ph+CML) and gastrointestinal mesenchymal tumours. Dasatinib, nilotinib and bosutinib are second-generation TKIs that were initially approved to treat patients with Ph+CML with resistance or intolerance to prior imatinib therapy. Ponatinib is a third-generation TKI used to treat adult patients with BCR::ABL1 T315I-positive CML (chronic phase, accelerated phase or blast phase) or T315I-positive Philadelphia chromosome-positive acute lymphoblastic leukaemia (Ph+ALL), or to treat the patients with CML or Ph+ALL for whom no other TKI therapy is indicated.⁷

These agents are generally well tolerated as they have a favourable toxicity profile compared with traditional chemotherapy regimens, but their potent and specific adverse events (AEs) should also be closely monitored.⁶ The safety profiles of BCR::ABL inhibitors as we know include fluid retention, pulmonary, cardiovascular, haematological and gastrointestinal-related AEs.^{8 9} Due to strict study entry criteria, relatively small sample sizes and limited follow-up time, clinical trial data may not reflect the full profile of AEs with BCR::ABL inhibitors. In particular, those with rare or late-onset AEs will emerge gradually as the drug is applied to a broader population. FDA Adverse Event Reporting System (FAERS) is the world’s best-known spontaneous reporting system for AEs. FAERS serves as an important pharmacovigilance tool, collecting case reports of suspected adverse drug reactions and reflecting the use of drugs in clinical practice.¹⁰ Data mining algorithms are often used to perform quantitative detection of signals of AEs.

Fluid retention-related AEs are common and clinically significant side effects during BCR::ABL inhibitor treatment. To better understand such AEs, we conducted a pharmacovigilance study using the real-world FAERS database to provide a reference for the clinical use of these drugs.

Methods

Patient and public involvement

Patients and the public were not involved in our study.

Data sources and extraction

The data for this study were obtained from the US openFDA public open project platform. FAERS is a critical source of post-marketing safety surveillance for all approved pharmaceutical and therapeutic biological products.¹⁰ The data are extracted using the OpenVigilFDA analysis tool, which is efficient and accurate via a docking application interface.¹¹ We searched the FAERS database for all BCR::ABL inhibitor AEs reported between 1 January 2004 and 30 September 2021. The reports obtained were used for subsequent AE data mining and analyses.

AE identification

Preferred terms (PTs) from the Medical Dictionary for Regulatory Activities were used to identify AEs.¹² FAERS database’s disproportionality revealed 1114 positive PTs for imatinib, 380 for dasatinib, 837 for nilotinib, 217 for bosutinib and 740 for ponatinib. PTs regarding fluid retention were screened and further joint searches were carried out for AEs with two or more PTs (online supplemental table 1). Some detected signals disappeared after combined retrieval by multiple similar PTs, such as face oedema of nilotinib, swelling of eyelid of dasatinib, brain oedema of ponatinib and hydrocephalus of dasatinib. PTs without obvious region manifestations, such as capillary leak syndrome, fluid retention, pleurisy, oedema, fluid overload and effusion, were removed.

Methods for mining pharmacovigilance signals and statistical analysis

Fluid retention-associated AEs for BCR::ABL inhibitors were summarised using descriptive analysis. A case–non-case disproportionality analysis was performed by calculating the reporting OR (ROR) and 95% CI.¹³ The algorithm is based on a 2×2 contingency table for disproportionality analyses (table 1), where a, b, c and d represent the target drug AE, target drug other AEs, all other drugs target AE and all other drugs other AEs, respectively. The ROR and 95% CI were calculated as ${\displaystyle \mathrm{R}\mathrm{O}\mathrm{R}=\frac{a/b}{c/d}}$, ${\displaystyle 95\mathrm{\%}\mathrm{C}\mathrm{I}={\mathrm{e}}^{\mathrm{l}\mathrm{n}\left(\mathrm{R}\mathrm{O}\mathrm{R}\right)\pm 1.96\sqrt{\frac{1}{\mathrm{a}}+\frac{1}{\mathrm{b}}+\frac{1}{\mathrm{c}}+\frac{1}{\mathrm{d}}}}}$. The signal was considered positive if the lower limit of 95% CI of ROR was >1.^{13 14} The analysis was run only considering coupled fluid retention events/BCR::ABL inhibitors with at least three cases. Subgroup analyses were performed by age and gender. Data analysis was performed using Microsoft­ Excel V.2016.

Table 1.

 Two-by-two contingency table for disproportionality analyses

	Target adverse event	All other adverse events	Total
Target drug	a	b	a+b
All other drugs	c	d	c+d
Total	a+c	b+d	a+b+c+d

Results

Patient demographics of AE reports submitted for BCR::ABL inhibitors

A total of 97 823 AEs were reported concerning BCR::ABL inhibitors, and 47 031 of these cases were associated with imatinib, 23 227 with dasatinib, 21 132 with nilotinib, 4490 with bosutinib and 7257 with ponatinib. Table 2 shows patient demographics of AE reports related to BCR::ABL inhibitors from the FAERS database, excluding reports of unknown information.

Table 2.

Patient demographics of AE reports of BCR::ABL inhibitors from FAERS database

	Imatinib	Dasatinib	Nilotinib	Bosutinib	Ponatinib
Total number of reported AEs	47 031	23 227	21 132	4490	7257
Sex
Male	22 731 (48.33%)	10 412 (44.83%)	10 098 (47.79%)	2071 (46.12%)	3434 (47.32%)
Female	20 058 (42.65%)	10 640 (45.81%)	9049 (42.82%)	2165 (48.22%)	2823 (38.90%)
Unknown	4242 (9.02%)	2175 (9.36%)	1985 (9.39%)	254 (5.66%)	1000 (13.78%)
Age
65	12 673 (26.95%)	7719 (33.23%)	6039 (28.58%)	1700 (37.86%)	2016 (27.78%)
≥65	7970 (16.95%)	4199 (18.08%)	3612 (17.09%)	1836 (40.89%)	1126 (15.52%)
Unknown	26 402 (56.14%)	11 310 (48.69%)	11 485 (54.35%)	954 (21.25%)	4115 (56.70%)
Outcome
Congenital anomaly	105 (0.22%)	14 (0.06%)	31 (0.15%)	0 (0.00%)	4 (0.06%)
Death	16 759 (35.63%)	2379 (10.24%)	3935 (18.62%)	373 (8.31%)	1791 (24.68%)
Disability	614 (1.31%)	5243 (22.57%)	382 (1.81%)	15 (0.33%)	76 (1.05%)
Hospitalisation initial or prolonged	8877 (18.87%)	154 (0.66%)	4953 (23.44%)	932 (20.76%)	3127 (43.09%)
Life-threatening	1119 (2.38%)	625 (2.69%)	662 (3.13%)	58 (1.29%)	153 (2.11%)
Other	19 721 (41.93%)	9971 (42.93%)	8934 (42.28%)	1471 (32.76%)	4232 (58.32%)
Required intervention	44 (0.09%)	22 (0.09%)	6 (0.03%)	0 (0.00%)	1 (0.01%)

AE, adverse event; FAERS, Food and Drug Administration Adverse Event Reporting System.

Signal mining analysis of fluid retention-associated AEs

The positive signals of fluid retention events for BCR::ABL inhibitors are shown in table 3. Periorbital oedema (ROR=24.931, 95% CI 22.404 to 27.743) was the strongest signal detected for imatinib, and peripheral swelling was the most commonly reported. Dasatinib’s strongest signal was chylothorax (ROR=161.427, 95% CI 125.835 to 207.085), and the most commonly reported was pleural effusion. The strongest signal of nilotinib was nipple swelling (ROR=48.796, 95% CI 26.27 to 90.636), and the most reported AE was pleural effusion. Chylothorax (ROR=35.798, 95% CI 14.791 to 86.642) was the strongest signal detected for bosutinib, and pleural effusion was most commonly reported. Gallbladder oedema (ROR=77.996, 95% CI 38.286 to 158.893) was the strongest signal, and peripheral swelling was the most reported for ponatinib.

Table 3.

Positve signals of fluid retention-associated AEs for BCR::ABL inhibitors based on the ROR analysis

Preferred term	ROR (95% CI)	N	Preferred term	ROR (95% CI)	N	Preferred term	ROR (95% CI)	N
Imatinib			Mouth swelling	2.349 (1.571 to 3.512)	24	Pericardial effusion	7.295 (6.219 to 8.558)	154
Periorbital oedema*	24.931 (22.404 to 27.743)	376	Papilloedema	2.321 (1.577 to 3.417)	26	Chylothorax	6.303 (2.351 to 16.899)	4
Bronchial oedema	20.417 (7.336 to 56.824)	4	Laryngeal oedema	2.315 (1.625 to 3.298)	31	Mouth swelling	5.908 (4.042 to 8.637)	27
Haemorrhagic ascites	12.478 (6.377 to 24.417)	9	Pulmonary oedema*	2.159 (1.892 to 2.462)	225	Generalised oedema	5.708 (4.451 to 7.320)	63
Bone marrow oedema	10.353 (6.612 to 16.209)	20	Peripheral swelling*	2.058 (1.938 to 2.186)	1088	Scrotal swelling*	5.353 (2.667 to 10.745)	8
Chylothorax	9.450 (5.425 to 16.460)	13	Dasatinib			Gastrointestinal oedema	4.568 (2.170 to 9.615)	7
Skin oedema	8.620 (5.310 to 13.994)	17	Chylothorax	161.427 (125.835 to 207.085)	84	Periorbital oedema*	3.734 (2.574 to 5.416)	28
Generalised oedema	8.339 (7.239 to 9.606)	200	Nipple swelling*	39.993 (20.951 to 76.345)	10	Ascites*	3.335 (2.730 to 4.073)	97
Eyelid oedema*	7.969 (7.043 to 9.017)	262	Pleural effusion*	37.424 (35.715 to 39.216)	2061	Pulmonary oedema*	2.818 (2.374 to 3.346)	132
Pleural effusion*	6.993 (6.535 to 7.483)	878	Pericardial effusion	14.146 (12.649 to 15.819)	321	Bosutinib
Malignant ascites	6.438 (3.318 to 12.489)	9	Pulmonary oedema*	11.217 (10.303 to 12.213)	557	Chylothorax	35.798 (14.791 to 86.642)	5
Oedema neonatal	5.962 (1.894 to 18.765)	3	Scrotal swelling*	5.485 (2.842 to 10.585)	9	Pleural effusion*	15.008 (12.960 to 17.380)	187
Scrotal swelling*	5.783 (3.667 to 9.120)	19	Periorbital oedema*	5.116 (3.773 to 6.936)	42	Ear swelling	13.211 (4.942 to 35.316)	4
Nipple swelling*	5.568 (1.771 to 17.507)	3	Malignant pleural effusion	5.065 (2.523 to 10.168)	8	Pericardial effusion	9.787 (7.316 to 13.093)	46
Eye oedema*	5.181 (4.690 to 5.723)	400	Generalised oedema	4.941 (3.830 to 6.374)	60	Pulmonary oedema*	6.212 (4.853 to 7.953)	64
Pericardial effusion	5.170 (4.550 to 5.875)	242	Infectious pleural effusion	4.576 (1.896 to 11.044)	5	Periorbital oedema*	4.804 (2.399 to 9.618)	8
Hydrops fetalis	4.745 (2.452 to 9.183)	9	Testicular swelling*	4.469 (2.000 to 9.987)	6	Ponatinib
Ascites*	4.705 (4.293 to 5.386)	307	Eye oedema*	2.677 (2.207 to 3.248)	104	Gallbladder oedema	77.996 (38.286 to 158.893)	8
Oedema genital	4.665 (2.207 to 9.861)	7	Tongue oedema	2.278 (1.183 to 4.386)	9	Oedematous pancreatitis	26.696 (9.928 to 71.786)	4
Hydrocele	4.064 (2.238 to 7.377)	11	Ascites*	2.150 (1.697 to 2.725)	69	Pericardial effusion	8.919 (6.981 to 11.394)	65
Face oedema*	3.650 (3.365 to 3.958)	601	Nilotinib			Pleural effusion*	7.283 (6.173 to 8.593)	144
Localised oedema	3.134 (2.158 to 4.552)	28	Nipple swelling*	48.796 (26.270 to 90.636)	11	Nasal oedema	5.716 (1.839 to 17.766)	3
Conjunctival oedema	3.053 (1.448 to 6.436)	7	Nasal oedema	18.228 (12.413 to 26.768)	27	Pharyngeal swelling	3.492 (1.451 to 8.400)	5
Muscle swelling	2.942 (1.315 to 6.584)	6	Testicular swelling	11.620 (6.839 to 19.742)	14	Ascites*	3.193 (2.255 to 4.520)	32
Gastrointestinal oedema	2.639 (1.368 to 5.091)	9	Oedematous pancreatitis	11.494 (4.736 to 27.894)	5	Gingival swelling	3.085 (1.282 to 7.420)	5
Oedema mucosal	2.601 (1.163 to 5.817)	6	Oedema genital	10.410 (4.924 to 22.007)	7	Pulmonary oedema*	2.416 (1.763 to 3.311)	39
Palatal oedema	2.357 (1.119 to 4.963)	7	Pleural effusion*	9.090 (8.322 to 9.928)	515	Peripheral swelling*	2.114 (1.818 to 2.459)	173

*Indicates secondary joint search with two or more preferred terms.

AEs, adverse events; FAERS, Food and Drug Administration Adverse Event Reporting System; N, number of events in the FAERS database; ROR, reporting OR.

Three fluid retention signals, pleural effusion, pericardial effusion and pulmonary oedema, were detected in all the BCR::ABL inhibitors. Pleural effusion (ROR=37.424, 95% CI 35.715 to 39.216), pericardial effusion (ROR=14.146, 95% CI 12.649 to 15.819) and pulmonary oedema (ROR=11.217, 95% CI 10.303 to 12.213) were disproportionately reported with dasatinib among all BCR::ABL inhibitors.

Subgroup analysis of pleural effusion, pericardial effusion and pulmonary oedema signals

Subgroup analyses of pleural effusion, pericardial effusion and pulmonary oedema signals by age and gender were performed and the results are shown in figure 1 (data shown in online supplemental table 2). Patients aged ≥65 years using dasatinib, imatinib, nilotinib or bosutinib had higher RORs for pleural effusion, pericardial effusion and pulmonary oedema. Patients aged ≥65 years using dasatinib were almost 42 times more likely to report pleural effusion than those using other drugs in the database (ROR=42.234, 95% CI 38.824 to 45.943), which is a huge signal. While the ROR (95% CI) of pleural effusion with dasatinib in patients aged <65 years was relatively lower, it also reached 29.772 (27.384 to 32.368). In the subgroup by gender, with imatinib, the RORs (95% CI) for pleural effusion signal and pulmonary oedema signal were higher in females than in males, with 8.013 (7.247 to 8.860) vs 5.488 (4.975 to 6.054) and 2.657 (2.205 to 3.202) vs 1.553 (1.273 to 1.896), respectively. With nilotinib, the ROR (95% CI) of the pulmonary oedema signal was also found to be higher in females than in males, with 3.417 (2.675 to 4.364) vs 1.927 (1.474 to 2.520).

Figure 1.

Age and gender subgroup analysis of pleural effusion, pericardial effusion and pulmonary oedema signals of BCR::ABL inhibitors. ROR, reporting OR.

Subgroup analysis of peripheral fluid retention-related signals with imatinib

We also performed a subgroup analysis of peripheral fluid retention-related signals with imatinib, as shown in figure 2 (data shown in online supplemental table 3). Patients aged ≥65 years using imatinib were almost 47 times more likely to report periorbital oedema than those using other drugs in the database (ROR=47.152, 95% CI 38.559 to 57.660), whereas patients aged <65 years were much lower at 17.897 (14.812 to 21.625). The ROR (95% CI) for periorbital oedema in female patients using imatinib was 30.542 (26.507 to 35.192), higher than in males at 18.331 (15.291 to 21.976). Also, patients aged ≥65 years and females using imatinib had higher RORs of generalised oedema and face oedema.

Figure 2.

Age and gender subgroup analysis of peripheral fluid retention-related signals with imatinib. ROR, reporting OR.

Discussion

Fluid retention is a palpable swelling caused by an increase in the volume of interstitial fluid in tissues. Fluid retention can be localised or generalised; generalised oedema or anasarca occurs when excess fluid accumulates massively and extensively. According to the region where fluid retention occurs, fluid retention can be classified as peripheral oedema, pulmonary oedema, pleural effusion, ascites and other forms, including lymphoedema, myxoedema, periorbital and scrotal oedema. Generally, superficial oedema is usually mild and manageable. However, pleural effusion, pulmonary oedema, pericardial effusion, ascites and generalised oedema can occasionally be severe and lead to significant complications. Many drugs can induce fluid retention events, and BCR::ABL inhibitors are one such class of drugs.¹⁵ However, the mechanism underlying the development of fluid retention events during BCR::ABL inhibitors therapy is unclear. It may be related to increased vascular permeability via endothelial ABL kinases, as well as off-target inhibition of platelet-derived growth factor receptor β and Src kinases.¹⁶

This pharmacovigilance study used the real-world FAERS database to present fluid retention events associated with BCR::ABL inhibitors. The signals of fluid retention consistent with the known safety profiles of BCR::ABL inhibitors as well as signals for less well-described AEs were detected. Imatinib was discovered to have the most fluid retention signals. Our findings agreed with those reported in phase II and III trials and post-marketing studies.^17–21 The eyelids and periocular skin were among the most common sites we found where local oedema could occur with imatinib treatment, and periorbital oedema demonstrated the strongest signal for imatinib. Imatinib-associated fluid retention was also detected to occur elsewhere, including central oedema, such as pleural effusion, pulmonary oedema, pericardial effusion, ascites and generalised oedema, as well as local oedema, such as peripheral, face and scrotal oedema. Dasatinib was detected to have the strongest signals regarding fluid retention in pulmonary and cardiac, such as chylothorax, pleural effusion, pericardial effusion and pulmonary oedema. Patients taking dasatinib may also experience ascites, generalised oedema and other forms of oedema, such as nipple swelling, scrotal swelling, periorbital oedema, testicular swelling and tongue oedema. Fluid retention events were reported relatively infrequently in nilotinib-treated patients.²² Nipple swelling, nasal oedema, testicular swelling, oedematous pancreatitis and oedema genital were found the strongest signals for nilotinib, and pleural effusion, pericardial effusion, pulmonary oedema, ascites and generalised oedema were the most commonly reported AEs. Fluid retention signals detected with bosutinib were mainly associated with the lung and heart. Chylothorax was the strongest signal detected and pleural effusion was the most reported AE for bosutinib. Although few positive signals were detected, the incidence of fluid retention due to bosutinib was not low, with a 13.3% incidence of fluid effusion-related AEs in patients with CML treated with bosutinib in a study with long-term follow-up (≥7 years).²³ While using ponatinib, we found fluid retention signals such as gallbladder oedema, oedematous pancreatitis, pericardial effusion, pleural effusion, nasal oedema, pharyngeal swelling, gingival swelling, ascites and peripheral swelling.

All the BCR::ABL inhibitors showed three fluid retention signals: pleural effusion, pericardial effusion and pulmonary oedema. Dasatinib had the highest RORs for pleural effusion, pericardial effusion and pulmonary oedema among all BCR::ABL inhibitors. Pleural effusion was a more prominent side effect in clinical practice among dasatinib AEs reported in the literature, occurring in approximately 15–35% of patients treated with dasatinib.^24–26 Pericardial effusion sometimes occurred concomitantly, with 26.9% of patients taking dasatinib reported to have pleural effusion and also pericardial effusion.²⁷ Less frequent occurrences have been observed in patients treated with bosutinib, but if used after dasatinib treatment, the proportion of pleural effusion during bosutinib treatment increases.^28–30 Pleural effusions had been observed in 5–15% of patients after prolonged exposure to bosutinib.³¹ Fluid retention events associated with nilotinib, such as pleural effusion, pericardial effusion and pulmonary oedema, were reported uncommon (<4% of patients) in the ENESTnd 10-year analysis.²² However, it cannot be ignored that these events have nevertheless been reported in real-life clinical practice, including severe cases.^{32 33} Compared with dasatinib, pleural effusion was less common with imatinib, according to studies.^{25 34} As for ponatinib, pleural effusion, pericardial effusion and pulmonary oedema events were rarely been reported.³⁵

Furthermore, subgroup analyses were performed to investigate the association between the three fluid retention-associated AEs by gender and age in BCR::ABL inhibitor-treated patients. Prior reports suggest that males and advanced age (≥65 years) were risk factors for the development of pleural effusion under dasatinib treatment.³⁶ Our subgroup analysis revealed that patients aged ≥65 years using dasatinib, imatinib, nilotinib or bosutinib had higher RORs for pleural effusion, pericardial effusion and pulmonary oedema. For ponatinib, the three fluid retention signals were not found to be very significantly different by age or gender. Previous reports have shown that females and patients aged ≥65 years are more likely to develop fluid retention associated with imatinib.¹⁷ Our subgroup analysis revealed that patients aged ≥65 years and female patients had stronger signals of periorbital oedema, generalised oedema and face oedema.

However, there are some limitations to this study. FAERS is a self-reporting system and there may exist problems with AE under-reporting, incomplete information and poor quality of reported data. Significant bias may occur given the spontaneous and voluntary reporting of AEs. The method of ROR, like other disproportionality measures, may suffer from specificity issues for false-positive or false-negative signals. Due to inherent database limitations, although a positive ROR value is a threshold used for statistical signal detection in FAERS, the ROR is an association measure and does not necessarily imply causality.

Conclusions

This retrospective pharmacovigilance study provides a more precise profile of fluid retention-associated AEs with BCR::ABL inhibitors in clinical practice. Pleural effusion, pulmonary oedema and pericardial effusion should be of concern for all BCR::ABL inhibitors. Monitoring of cardiac-related and pulmonary-related signs and symptoms, body weight, and periorbital and peripheral tissue tone can aid in the early detection of possible fluid retention in patients treated with BCR::ABL inhibitors.

Data availability statement

Data are available upon reasonable request. Not applicable.

Ethics statements

Patient consent for publication

Not required.

Ethics approval

This study used anonymised information from the database, which is open to the public; therefore, in accordance with the Declaration of Helsinki, institutional ethics approval and informed consent to participate were not required.