Oral itraconazole for epistaxis in hereditary hemorrhagic telangiectasia: a proof of concept study

Abstract

The inhibiting effects of itraconazole, an antifungal drug on vascular endothelial growth factor (VEGF) have recently been discovered. By inhibiting VEGF, itraconazole has shown potential in clinical trials as anti-cancer treatment. In hereditary hemorrhagic telangiectasia (HHT) patients, VEGF levels are elevated and inhibition of VEGF can decrease bleeding. Itraconazole could potentially serve as anti-angiogenic therapy for HHT-related bleeding. We report a proof of concept study with HHT patients and severe epistaxis. Patients were treated with daily 200 mg orally administered itraconazole for sixteen weeks. Twenty-one HHT patients, 8 females (38%), 13 males (62%), median age of 59 years (interquartile range (IQR) 55–69) were enrolled. Of these patients, 13 (62%) were diagnosed with HHT type 1, seven (33%) with HHT type 2 and in one patient (5%), no pathognomonic HHT mutation was found. Four patients (19%) prematurely terminated the study (3 due to mild or moderate side-effects) resulting in 17 patients included in the analyses. The median epistaxis severity score significantly decreased during treatment from 6.0 (IQR 5.1–7.2) to 3.8 (IQR 3.1–5.2) (p = 0.006). The monthly epistaxis frequency decreased from 56 to 38 epistaxis episodes (p = 0.004) and the monthly duration from 407 to 278 minutes (p = 0.005). Hemoglobin levels did not significantly change. The quality of life showed a small but significant improvement. In conclusion, oral itraconazole significantly improved epistaxis in HHT patients. The potential benefit of itraconazole in HHT should be further investigated.

Abbreviations

ACVRL1

Activin A receptor like type 1

AE

Adverse event

APC

Argon plasma coagulation

ENG

Endoglin

ESS

Epistaxis severity score

HHT

Hereditary hemorrhagic telangiectasia

IQR

Interquartile range

MCS

Mental component summary

MFI-20

Multidimensional fatigue inventory 20

PCS

Physical component summary

RBC

Red blood cell

SAE

Severe adverse event

SF-36

Short Form Health Survey 36

VEGF

Vascular endothelial growth factor

Introduction

Hereditary hemorrhagic telangiectasia (HHT), an autosomal dominant inherited disease, is characterized by mucocutaneous telangiectasia, arteriovenous malformations and recurrent, spontaneous epistaxis. HHT is caused by a disruption of the transforming growth factor β signaling pathway in most of the cases due to mutations in the genes encoding for endoglin (ENG) or activin A receptor type II-like 1 (ACVRL1) which cause HHT type 1 and HHT type 2, respectively [1, 2]. Up to 97% of the HHT patients will eventually suffer from recurrent, spontaneous epistaxis [3]. Epistaxis treatment in HHT patients can be very challenging because standard epistaxis management such as cautery rarely has a durable effect [4]. In addition, the quality of life (QoL) of HHT patients is often negatively affected by the recurrent epistaxis [5]. In severe cases, invasive surgical measures and systemic medical treatment are used in an attempt to reduce the epistaxis severity. But, even with treatment, epistaxis may result in iron-deficient anemia sometimes requiring intravenous iron treatment or blood transfusions.

There is considerable interest in developing therapies that target the pathogenic pathways of HHT. Evidence suggests that the loss of ENG or ACVRL1 allele functioning results in a pro-angiogenic state. This is supported by the fact that elevated plasma levels of vascular endothelial growth factor (VEGF) have been found in HHT patients and by the fact that VEGF inhibition, with the VEGF inhibiting antibody bevacizumab, decreases the HHT-related symptoms remarkably [6]. However, due to the side-effects such as proteinuria, the intravenous administration and especially the high costs of bevacizumab, the search for alternative drugs is needed.

Itraconazole is a widely used antifungal agent primarily used for the treatment fungal infections [7]. It is also used in clinical trials as cancer therapy for various types of cancer including lung and ovarian cancer and these studies have shown promising results [8]. One of the mechanisms includes a strong inhibition of endothelial cell proliferation in response to VEGF [9]. Based on this evidence and the fact that VEGF inhibition has shown to decrease epistaxis severity in HHT, the anti-angiogenic effects of itraconazole could potentially reduce epistaxis severity in HHT patients. Furthermore, itraconazole is an orally administered, affordable, and generally well tolerated drug. Therefore, we designed a non-randomized, open label, pilot study to investigate therapeutic potential of itraconazole in HHT.

Methods

Study design

This open label, prospective, non-randomized, phase 2, single center study (European Union Drug Regulating Authorities Clinical Trials (EudraCT) number 2017-003272-31) was performed in the St. Antonius Hospital, the Dutch HHT center. The study started in Augustus 2018 and ended in December 2019. The ethics committee of the St. Antonius Hospital (MEC-U) approved the study (NL62902.100.17 / R17.085). All patients signed informed consent before enrollment. This study performed is in accordance with the Declaration of Helsinki.

Patient selection

Patients were included if they were 18 years or older at the time of inclusion, suffered from at least four episodes of epistaxis per week and were diagnosed with HHT (genetically confirmed diagnosis or according to the Curacao criteria (≥ 3 positive criteria) [10]). Moreover, patients had to suffer from anemia in the last six months, and/or required the use of iron treatment or blood transfusions in the last twelve months. Exclusion criteria were a history of any ventricular cardiac dysfunction, elevated liver enzymes, any pre-existing liver disease, a history with known liver toxicity caused by any drug, a life-expectancy < 1 year, pregnancy, nursing, a pregnancy wish in the study period or inadequate use of anticonception, the use of chemotherapy or drugs that were contraindicated when using itraconazole. Finally, patients with a known hypersensitivity to or allergy for azole antifungal drugs, and those who could not give adequate informed consent or did not sufficiently understand the Dutch or English were excluded. We aimed to include 20 to 25 persons in this study based on previous studies by our group [11, 12]. We have not performed a sample size calculation because the effect size of itraconazole on epistaxis is unknown.

Treatment protocol

Patients visited the St. Antonius Hospital at the trial inclusion (screening visit), after four weeks (baseline visit), eight weeks and at the end of the trial (20 weeks). Following the screening visit, the patients were instructed to record their nosebleeds in an epistaxis diary for four weeks (from the screening visit to the baseline visit). Patients started with itraconazole following the baseline visit, with two capsules of 100 mg (in total 200 mg) once daily for 16 weeks. In the last four weeks of the trial, patients were instructed to record their epistaxis diary again. If indicated, patients received red blood cell (RBC) transfusions and intravenous iron infusions in their own local hospital and the patient’s independent treating physician determined the need for treatment. The physician was aware of the patient’s participation in this trial. All patients received standard care during the trial for the epistaxis similar as before enrollment. If patients used a proton-pump inhibitor, they were instructed to take the capsules together with an acidic drink to increase gastric acidity and thereby increase absorption of the drug. In Fig. 1, the flowchart with the treatment protocol of this study is depicted.

Fig. 1. Flowchart of the study. ESS, epistaxis severity score; Hb, hemoglobin; MFI-20, Multidimensional Fatigue Inventory 20; SF-36, Short Form Health Survey 36.

Outcomes

The primary outcome was the change in epistaxis severity score (ESS) measured at baseline compared to the end of the trial. The ESS is specifically developed for assessing epistaxis severity in HHT patients [13]. It consists of 6 questions which are used to calculate the severity on a continuous scale with a range from 0 (none) to 10 (severe epistaxis). The questions concerning typical epistaxis symptoms (the number, the duration and the intensity of epistaxis) of the ESS are usually an average of past three months. We have adjusted this period to the past four weeks. Secondary outcomes included safety and side-effects, and the change of hemoglobin and ferritin levels at the end of the trial as compared to baseline. Furthermore, at the end of the trial the efficacy of treatment was evaluated and patients reported their subjective change in epistaxis severity on a scale with 5 grades including: a strong decrease, a moderate decrease, no change, a moderate increase, and a strong increase in severity)

To assess pharmacodynamics of itraconazole, serum levels of itraconazole were recorded after 4 weeks of treatment (at week 8) and at the end of treatment (at week 20). In supplemental table 1 the questions and calculation of the ESS are shown. Epistaxis severity was further assessed with an epistaxis diary. This diary is frequently used in our hospital in a clinical setting to monitor epistaxis severity and efficacy of therapy in HHT patients. Patients recorded their epistaxis severity in two different periods: before itraconazole treatment (the first four weeks of the trial) and during the last four weeks of itraconazole treatment. In the diary the severity (pouring, gushing, severely gushing) and the duration of each nose bleed was recorded. The number, severity and duration of the epistaxis episodes as recorded in the diary in both periods (before and during treatment) were compared. Furthermore, the number of RBC transfusions and iron infusions during the study period (16 weeks) were compared with the number of transfusions and iron infusions required in the 16 weeks prior to treatment. Moreover, the QoL measured with the Short Form Health Survey 36 (SF-36) and the level of fatigue measured with Multidimensional Fatigue Inventory 20 (MFI-20) were assessed [14, 15]. The SF-36 consists of 36 questions concerning QoL that assesses eight health concepts. Both the physical component summary and mental component summary can be calculated and are aggregates of these eight subscale scores. These summary scores range from 0 to 100, with higher scores representing better self-reported health [14]. The MFI-20 is a questionnaire with 20 questions designed to measure fatigue [15]. It covers the 5 dimensions: general fatigue, physical fatigue, reduced activity, reduced motivation and mental fatigue. The MFI-20 subscales range from 0 to 20, with lower scores representing less self-reported fatigue complaints.

Statistical analysis

Quantitative data were presented as absolute frequencies or percentage. The median and interquartile range (IQR) were used and subsequently non-parametric testing was performed due to the small sample size and skewed distribution of data. We used the McNemar test (for dichotomous parameters) and Wilcoxon signed-rank test (for continuous and ordinal parameters) to assess the differences at baseline compared to the end of treatment. A p-value < 0.05 was regarded as statistically significant. Statistical analysis was performed with SPSS version 24.0 for Windows (IBM, Armonk, NY, USA) and GraphPad 5 (GraphPad Software, Inc., La Jolla, CA, USA).

Results

Baseline characteristics

In total, 21 patients were enrolled in this pilot study of which 8 (38%) were female. The median age was 58 years (IQR 55–69). All patients were clinically diagnosed with HHT according to the Curacao criteria. Thirteen patients (62%) carried a ENG mutation, in 7 patient (33%) a mutation was found in the ACVRL1 gene and in one patient (5%) no disease causing mutation could be identified in the ENG or ACVRL1 gene. All the baseline characteristics are shown in Table 1. During the study, 4 patients (19%) did not complete the entire trial due to side-effects (n = 3) or personal reasons (n = 1), leaving 17 (81%) patients for further analysis.

Outcomes

Epistaxis

In the 17 individuals that completed the trial, the median ESS decreased from 6.0 (IQR 5.2–7.2) at baseline to 3.8 (IQR 2.6–5.2) at the end of the trial (p = 0.001). The monthly epistaxis frequency decreased from 56 episodes at baseline to 38 episodes at the end of the trial, p < 0.001 and the monthly duration decreased from 407 at baseline to 278 minutes at the end of the trial, p = 0.001. The severity of epistaxis did not change during treatment. Before treatment severity of epistaxis was recorded as pouring in 68%, gushing in 24% and severely gushing in 8%. This changed during therapy to 61%, 29% and 10% for pouring, gushing and severely gushing, respectively while using itraconazole. During treatment, three individuals required surgical epistaxis treatment (two individual received APC treatment of nasal mucosa and another patient required trans catheter nasal embolization) despite itraconazole treatment to diminish epistaxis severity. The subjective change in epistaxis severity at the end of the trial included a strongly decrease in severity in 5 patients (29%), a moderate decrease in 8 patients (47%) and no change in 4 patients (24%). None of the patients experienced a moderate or severe aggravation of epistaxis during the trial. The outcomes are depicted in Fig. 2; Table 2.

Fig. 2. Depiction of change in epistaxis during the trial.

a ESS. b Monthly epistaxis number. c Monthly epistaxis duration. ESS, epistaxis severity score

Table 2. Outcomes.

Parameter	Baseline	End of trial	p-value
Number of included patients, n	17	17	n/a
ESS (IQR)	6.0 (5.2–7.2)	3.8 (2.6–5.2)	0.001
Hemoglobin levels, mmol/L (IQR)	7.1 (6.3–7.8)	7.1 (5.8–8.1)	0.84
Ferritin levels, ug/L (IQR)	37 (23–183)	36 (21–62)	0.51
SF-36
PCS	38 (31–43)	40 (32–52)	0.015
MCS	45 (31–55)	47 (36–51)	0.91
MFI-20
General fatigue	16 (15–20)	15 (11–18)	0.060
Physical fatigue	16 (12–19)	15 (10–19)	0.25
Reduced activity	14 (8–16)	12 (9–16)	0.98
Reduced motivation	10 (8–13)	10 (7–13)	0.95
Mental fatigue	10 (5–16)	8 (4–13)	0.077
Parameter	Prior to trial	End of trial	P-value
Number of included patients, n	17	17	n/a
Monthly epistaxis number, n (IQR)*	56 (45–80)	38 (24–49)	<0.001
Monthly epistaxis duration, minutes (IQR)*	407(306–813)	278(156–406)	0.001
Epistaxis episodes (%)*
Pouring	68	61
Gushing	24	29
Severely gushing	8	10	n/a
Iron infusions, n (IQR; Range)¥	0 (IQR: 0–2; range: 0–4)	0 (IQR: 0–2; range: 0–4)	0.56
Individuals with iron infusions (%)¥	8 (47)	6 (35)	0.63
Blood transfusions, n (range)¥	0 (0–18)	0 (0–18)	0.66
Individuals with blood transfusions (%)¥	3 (18)	2 (12)	1.0

ESS epistaxis severity score, IQR interquartile range, MFI-20 Multidimensional Fatigue Inventory 20, MSC mental component summary, PCS physical component summary, SF-36 Short Form Health Survey 36

The asterisk (*) indicates a parameter was that measured in the one month period prior to itraconazole treatment and during the last month of the trial

The Yen sign (¥) indicates a parameter that was measured in the 4-month period prior up until enrollment and the in period with during the trial (with a similar duration of 4 months)

Blood panel parameters, iron infusions and blood transfusions

The median hemoglobin level was 7.1 mmol/L (IQR 6.3–7.8) at baseline and remained unchanged (7.1 mmol/L (IQR 5.8–8.1 mmol)) at the end of the trial (p = 0.84). The ferritin levels did not statistically change during this trial.

Similar, the number of iron infusions and blood transfusions did not change significantly. In the 4 months before enrollment 8 (47%) individuals received iron infusions and 3 (18%) individuals received blood transfusions out of the 17 that completed the trial; during the 4 months in trial 6 out of these 17 (35%) required iron infusions and 2 (12%) out of the 17 individuals required blood transfusions.

Quality of life and fatigue

The physical component summary of the SF-36 improved in 13 out of the 17 (76%) individuals and improved slightly from 38 (IQR 31–43) to 40 (IQR 32–52), p = 0.015. The mental component summary of the SF-36 and fatigue complaints measured with the MFI-20 showed a tendency of improvement.

Pharmacokinetic data

The median itraconazole serum level after 4 weeks of treatment was 0.53 mg/L (IQR 0.40–0.93) and 0.50 (0.34–0.92) mg/L at the end of the trial. The itraconazole levels were in 81% of the measurements below the therapeutic range for antifungal treatment (reference value: 1–2 mg/L). There was no correlation between the average itraconazole serum level and decrease in ESS. Proton-pump inhibitors, which can significantly decrease itraconazole uptake were used by 6 patients.

Safety and side-effects

The majority of the adverse events (AE) were mild and transient. The most frequently observed AE was stomach and abdominal cramps (n = 4). In four individuals either a moderate or severe AE occurred. The moderate AE’s included atrial fibrillation (without any signs of heart failure) requiring anticoagulant therapy, a skin rash and mild heart failure. Itraconazole was continued in the cases with atrial fibrillation and skin rash. The atrial fibrillation, probably unrelated to itraconazole, was still present at the end of the trial. The skin rash resolved after temporary application of topical corticosteroids. The heart failure resolved after the cessation of itraconazole and the temporary use of diuretics. One patient had a severe AE, which was probably not related to itraconazole: this patient suffered from a bacterial endocarditis that required surgical aortic valve replacement and the use of antiplatelet therapy. The cause of the endocarditis was most likely a period of septicemia that occurred just prior to enrollment in this clinical trial following prolonged nasal packing for a severe nasal bleeding episode. Of the four individuals who did not complete the trial, three individuals (14%) terminated the trial prematurely due to side-effects which included fatigue complaints, paresthesia at the extremities and congestive heart failure. These three side-effects completely resolved after the cessation of itraconazole. One individual (5%) did not complete the trial due to personal reasons. None of the participants suffered from elevated liver enzymes during therapy. The AE’s are depicted in supplemental Table 2.

Discussion

This is the first trial that investigated the clinical effects of itraconazole on HHT-related epistaxis attributed to its anti-angiogenic effects. We observed a significantly decrease in ESS and the epistaxis frequency and duration measured with the diary following itraconazole treatment. We did not observe significant differences in hemoglobin levels, transfusion needs or numbers of iron infusions.

The drug itraconazole was approved in 1992 by the Food and Drug Administration. It is a widely used, affordable, antifungal agent with a well-known safety profile [7]. The costs of oral itraconazole capsules in the Netherlands with a dose regimen of 200 mg once daily, is approximately € 14 per month [16]. Itraconazole inhibits the synthesis of ergosterol, the major sterol component of fungal plasma membranes. In a wide drug screen with existing and already approved drugs to discover novel anti-angiogenic inhibitors, the surprising potential of itraconazole was discovered [17]. Itraconazole selectively inhibited endothelial cells by targeting multiple pathways including the VEGF, hedgehog, and mTOR [18]. Another study showed that itraconazole significantly inhibited the binding of VEGF-A to VEGF receptor 2 (VEGFR2) and that both VEGFR2 and the downstream pathway failed to become activated after VEGF stimulation [19]. VEGFR2 is a key component in stimulating variety of signaling pathways involved in angiogenesis. In addition, the signal transduction with VEGF-A and VEGFR2 is the most prominent ligand-receptor complex in the VEGF system and activation leads to endothelial cell proliferation and new vessel formation [20]. In the study by Nacev et al. the inhibition of VEGFR2 in vitro occurred at a very low itraconazole concentration, lower than the plasma concentration achieved on a standard 200 mg daily oral dosing regimen [19]. In cancer therapy, the safety and efficacy of itraconazole as anti-cancer treatment have been assessed in several forms of cancer [8]. In a randomized trial with non-small-cell lung cancer, the combination of itraconazole and chemotherapy significantly increased the overall survival compared to the overall survival in the monotherapy group with solely chemotherapy [21]. There is also evidence suggesting that the anti-angiogenic activity of itraconazole can be used for indications other than oncological diseases. In an animal model with laser induced choroidal neovascularization, intraocular itraconazole injections showed a strong vascular angiogenesis inhibition compared to effect of placebo injections [22]. VEGF is also believed to play an important role in the disease mechanism of HHT and severity of epistaxis. Elevated plasma levels of VEGF have been found in HHT patients [23] and VEGF inhibition with bevacizumab has shown to reduce epistaxis and gastrointestinal bleeding in HHT patients [6]. More recently, a clinical trial in HHT patients investigating pazopanib, a tyrosine kinase inhibitor that blocks VEGF receptors has shown that pazopanib significantly decreased HHT-related bleeding [24]. The VEGF inhibiting mechanism is most likely the cause of the epistaxis decrease that we observed during this study. The severity per nosebleed, however, did not change and not all participants experienced a decrease in bleeding: in three cases still invasive measurements were required to treat the epistaxis despite itraconazole. Furthermore, we did not observe a significant change in hemoglobin levels. This may have been caused by the slight decrease in transfusion and iron infusion needs: 6 patients required intravenous iron supplementation during the trial versus 8 before; and 2 required blood transfusions versus 3 before the trial. The serum concentration of itraconazole was below the reference value of itraconazole treatment for antifungal indications in the majority of the measurements. It could be possible that the efficacy of itraconazole is underestimated in this study because of the low serum levels, but this was not supported by the finding that there was no correlation between average serum concentration and decrease in ESS. The reason for the low serum values may have been the concomitant use of proton-pump inhibitors. Concomitant use of proton-pomp inhibitors and itraconazole can result in a substantial decrease of itraconazole serum levels because low gastric pH is required for an adequate uptake. The uptake can be significantly increased but not normalized with the concomitant use of acidic beverages such as cola [25]. Furthermore, 2 individuals required anticoagulant therapy during itraconazole therapy due to AE’s (atrial fibrillation and aortic valve replacement) both probably not related to itraconazole therapy. The use of anti-coagulants may have aggravated the epistaxis in these cases and thereby decreased the hemoglobin levels. Finally, the decrease in epistaxis frequency and duration that we observed may have been partly caused by a placebo effect which has been observed in two randomized trials with HHT patients and epistaxis [26, 27].

The most frequent AE’s of itraconazole treatment are of gastrointestinal origin (20%). Elevated liver function tests (5%) and skin rash (3%) are also frequently observed; and neurological symptoms such as headaches and peripheral neuropathy can arise during long-term treatment. In approximately 20% of the patients that are taking itraconazole, treatment is terminated due to adverse reaction [28]. In a minority moderate or serious side-effects can occur especially drug-induced liver injury and congestive heart failure [7]. Several case reports have been published with itraconazole treatment and heart failure. Itraconazole has a negative inotropic effect, but the exact mechanism how heart failure is induced, is unknown. The risk of developing congestive heart failure is especially high in patients with known heart disease and itraconazole treatment is contraindicated in such cases [29]. The side effect that we observed in this clinical trial are quite similar to the reported AE’s in the literature. Gastrointestinal symptoms were most frequently reported. There was one case with skin rash and one with mild congestive heart failure but without any previous history of heart disease. In our study, 3 individuals (14%) prematurely terminated the trial due to side-effects.

The strength of our study was the prospective design with a standardized protocol and predetermined outcomes. We also assessed patient reported outcomes such as change in QoL and fatigue symptoms. Our study is limited by the absence of a control group and a small sample size. We aimed to include 20–25 patients with severe epistaxis in this study. The small sample size could have introduced selection bias, we believe that this was limited by the fact that the St. Antonius Hospital is the only HHT center in the Netherlands. Furthermore, the studied population only reflects the HHT population suffering from severe bleeding caused by HHT. The median age of our population was 58 years, comparable with other studies in which treatment of severe HHT-related bleeding is investigated [26, 27]. The genetic HHT distribution of the studied subjects (62% HHT type 1; 33% HHT type 2 and 5% unknown HHT type) is quite comparable to the genetic distribution of the HHT population seen in our country: approximately 53% with HHT type 1, 40% with HHT type 2 and 7% with an unknown HHT type [30].

In conclusion, treatment with oral itraconazole significantly decreased epistaxis frequency and duration in our cohort of HHT patients. However, moderate side-effects have occurred during treatment. A larger, randomized controlled clinical trial is needed to confirm the therapeutic benefit of oral itraconazole as treatment for HHT. Itraconazole could potentially be a good and affordable treatment option for HHT patients suffering from severe epistaxis.

Table 1. Baseline characteristics.

Parameter	Value
Number of patients	21
Median age, years (IQR)	58 (55–69)
Median body mass index, kg/m2 (IQR)	26 (24–28)
Race (%)
Caucasian	20 (95)
African American	1 (5)
Smoker, n
Never	6 (27)
Former	14 (64)
Active	2 (9)
Gender, n (%)
Female	8 (38)
Male	13 (62)
Clinical diagnosis (≥ 3 Curacao criteria), n (%)	21 (100)
HHT type, n (%)
Type 1	13 (62)
Type 2	7 (33)
Type unknown	1 (5)
Visceral localization, n (%)
PAVM	8 (38)
CAVMa	0 (0)
Digestive tract telangiectasis’b	5 (24)
Anticoagulant therapy before enrolment, n (%)	0 (0)
Use of proton-pump inhibitors, n (%)	6 (29)
Epistaxis therapies in past, n (%)
Nasal coagulation	21 (100)
Septodermoplasty	13 (62)
Nasal embolization	6 (29)
Oral tranexamic acid	11 (52)
Thalidomide	4 (19)
Intravenous bevacizumab	2 (10)
In 4 months prior to inclusion, n (%)
Nasal coagulation and/or embolization	8 (38)
Iron infusion(s)	10 (48)
Blood transfusion(s)	4 (19)

CAVM cerebral arteriovenous malformation, HHT hereditary hemorrhagic telangiectasia, PAVM pulmonary arteriovenous malformation, SD standard deviation

^aCAVM screening has been performed in 17 (81%) individuals

^bDigestive tract has been evaluated in 11 (52%) individuals