A randomised, double-blind, placebo-controlled trial of oral propranolol for hereditary haemorrhagic telangiectasia

Introduction

Epistaxis in hereditary haemorrhagic telangiectasia (HHT) has significant morbidity and impact on quality of life. Patients experience frequent nose bleeds, with many exhibiting chronic anemia requiring blood or iron replacement. As a genetic disorder, treatment aims to reduce frequency and severity of bleeding. Non-invasive therapies such as tranexamic acid, tamoxifen or hormonal therapy have varied efficacy and treatment resistance can develop (1). In severe cases, surgical interventions such as laser, septodermoplasty or nasal closure (Young’s procedure) may be indicated (1). Recent studies suggest intravenous bevacizumab can reduce HHT-related bleeding (2) by inhibiting vascular endothelial growth factor (VEGF); a key factor overexpressed in HHT with downstream effects of dysregulated neo-angiogenesis (3). Despite this, challenges of low accessibility, high costs, and patient resistance to frequent infusions are reported.

To date there is a paucity of data on effective therapies that are readily available for patients with HHT, especially using randomised trial designs. Beta-blockers have been proposed as a potential candidate in HHT following success in treating superficial haemangiomas in children. In-vitro observations in this group suggested Propranolol may cause dose-dependent inhibition of VEGF expression (4) and apoptosis of endothelial cells (5).

In HHT specifically, both timolol (commonly used in glaucoma and ocular hypertension) and propranolol have been trialed in adults (6-10). While a timolol 0.5% nasal spray did not show improvement (7), Peterson et al. [2020] (9) reported symptom reduction from a timolol 0.10% thermosensitive gel; though both placebo and timolol reported significant improvement and the lubricant effects of the gel was suggested as a major factor. A nasal propranolol spray reported improvement in epistaxis but also a high reported rate of intranasal burning sensation upon application of the spray (6). Clinical improvements were observed in an oral form of propranolol by Contis et al. [2017] with minimal side effects in a single-arm observational case series (8).

As such, oral propranolol is the most promising candidate for further exploration in a randomised control trial. A pilot study conducted at our site has also supported this. We hypothesize that oral propranolol in patients with HHT would decrease epistaxis severity symptoms and endoscopic evidence of telangiectasia compared to placebo, and thus may be a readily accessible and safe treatment option. We performed a randomised, placebo-controlled double-blind study examining the effects of oral propranolol in HHT.

Methods

The Royal Brisbane and Women’s Hospital (RBWH), Queensland, Australia provides a sub-specialist HHT rhinology clinic that accepts referrals across the state. Patients above 18 years of age with HHT diagnosed by the Curacao criteria were recruited from RBWH between 2021 and 2023. An Epistaxis Severity Score (ESS) of ≥2 was required to be able to observe clinical effects. Both newly diagnosed patients, and those had persistent symptoms despite conventional treatment were offered recruitment and written consent was obtained. Participants who were pregnant, breastfeeding, or had absolute contraindications (sinus bradycardia, severe respiratory airway disease, atrioventricular block) to propranolol were excluded. Patients with planned surgical intervention or bevacizumab therapy for epistaxis at time of screening were also excluded. A dosage of 40 mg twice-daily oral propranolol was determined based on dosage used for infantile haemangioma and in consultation with a cardiologist. This study is reported according to the CONSORT reporting checklist (available at https://www.theajo.com/article/view/10.21037/ajo-24-64/rc).

Randomisation

Randomisation was performed by a third-party (Optima Ovest, Western Australia, Australia) using a 2:1 intervention: placebo ratio. Optima Ovest also manufactured either placebo or propranolol tablets stored in sequentially numbered containers which were dispensed in order of participant recruitment. First dose was taken in the clinic under monitoring, with subsequent doses self-administered in the community. A staff member not involved in the trial held a randomisation schedule in case of emergency unblinding.

Outcomes

The primary outcome was the ESS, which is a widely used validated patient outcome measure specific for HHT-associated epistaxis. The overall score ranges from 0 (none) up to 10 (most severe) (11). A minimal clinically important difference (MCID) was set at of 0.71 (12). A general [Short Form-36 (SF-36)] (13) quality of life measure was also reported; key domains of interest of the SF-36 were social function, emotional, and mental health. An HHT-specific endoscopic staging score, the HHT Endoscopy Score (HES) (14) was assessed by an otolaryngologist who were blinded, as well as vital signs and biochemistry (full blood count, iron studies, electrolytes).

Measurements of the above outcomes occurred at baseline and follow up visits at 3 and 6 months. At the conclusion of 6-month patients were trialed off intervention (placebo or propranolol) for a further 30 days to assess symptom change. Due to the potentially significant change in quality-of-life, patients were offered unblinding at the end of the trial with an option of continuing oral propranolol open-label if allocated.

Safety

All patients were screened with a baseline electrocardiogram (ECG), blood pressure (BP), heart rate and clinical history. If significant cardiac disease history was identified, a cardiology review was arranged to determine eligibility. The first dose of propranolol was given in the clinic followed by a period of observation. All patients had BP monitored at each visit and were provided contact information to a clinical nurse.

Power and statistical analyses

It was estimated a-priori that 20 participants would be required per group to demonstrate treatment effect with 80% power at significance level of 0.05 (two-sided) based on preliminary data from the pilot study. Intention-to-treat data analysis was performed using SPSS, version 27.0 (IBM, Armonk, NY, USA) (15). Normality of data was determined by Shapiro-Wilk tests. Patient characteristics were described as mean with standard deviation (SD), or median and interquartile range (IQR) depending on normality. Change in repeated measurements over time was analysed with a two-way repeated measures analysis of variance (ANOVA). Box’s test was used to assess equality of variance between groups. If univariant significance was suggested over the study duration, pairwise comparisons across time points without mathematical correction were conducted. Comparison between change in symptom score between placebo and propranolol was determined by an independent t-test or Mann-Whitney U test. A 95% confidence interval (CI) was used for reporting; a P value of <0.05 was considered statistically significant.

The trial was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Royal Brisbane and Women’s Hospital Research Governance Office (HREC/2019/QRBW/4855) and informed consent was taken from all individual participants.

Results

A total of 23 patients were screened for inclusion in this trial. There was one screening fail due to an ESS score of 0.28. Age range of inclusion was 28–69 years, with a median (IQR) of 55 years (43–61 years). Median (IQR) baseline ESS score of all patients was 5.10 (3.85–5.66). Seven patients pre-emptively discontinued after commencement of the study. Four patients assigned to propranolol discontinued; three due to side effects and one was lost-to-follow-up prior to their 3-month review. A further two patients in placebo group discontinued due to epistaxis requiring hospital admission and unblinding. One patient on placebo withdrew from placebo-related side effects.

Overall, there were 15 patients that completed the study, with 5 patients receiving placebo, and 10 patients allocated to oral propranolol. Their baseline characteristics are shown in Table 1.

ESS

There was a significant improvement in ESS scores in the oral propranolol group at 3 months compared to baseline (mean change: −1.68, 95% CI: −3.25 to −0.12, P=0.038). By 6 months, the observed mean change was −1.44 (95% CI: −2.98 to 0.10, P=0.064), but this did not reach statistical significance. No placebo effect was observed during the study by 6 months (mean change −0.21, 95% CI: −2.39 to 1.97, P=0.801). Individual changes in ESS score are shown in Figure 1. Despite the clinically observed difference between these groups, the overall mean change did not reach statistical significance in propranolol compared to placebo (mean change −1.23, 95% CI: −3.55 to 1.10, P=0.133) (refer to Figure 2).

Figure 1 ESS at each study visit including from baseline to 6 months (active treatment phase), and the observation period of 1 month after treatment completion (post-treatment phase). (A) Placebo; (B) propranolol. ESS, Epistaxis Severity Score.

Figure 2 Net change in ESS from baseline compared to 6-month visit (end-of-study). The figure shows the median net change of ESS scores in patients over 6 months treatment with propranolol and placebo. Comparison of means was conducted with an independent samples t-test, with a non-statistically significant P value of 0.13. CI, confidence interval; ESS, Epistaxis Severity Score.

After the end-of-treatment (6 months), and before the end-of-study visit (7 months), symptoms in the propranolol group regressed towards baseline severity scores as expected.

Furthermore, in the propranolol group (see Figure 1), it was clinically observed that patients demonstrated three main response patterns to the intervention. Type A response was ‘complete responders’ defined as a sustained/or continued improvement in ESS score at both 3 and 6 months. These patients also had the highest median (IQR) baseline ESS score of 5.33 (4.75–5.86). Type B response, or ‘partial responders’ were patients that exhibited improvement at 3 months, however reverted towards baseline by study end. There were two patients with no response at all; though one patient was receiving half dose propranolol and the other had documented medical compliance issues.

Nasal endoscopy scores

Endoscopic scores were assessed by multiple otolaryngologists across the study duration.

There were no observed net changes within endoscopy scores for either propranolol (0.61, 95% CI: −1.50 to 2.72) or placebo (0.33, 95% CI: −4.27 to 4.94) groups during the treatment period. No difference was observed in endoscopic scores between the two groups (P=0.805).

Despite this, Figure 3 shows a right middle turbinate from a patient that received propranolol from baseline to 6 months. There is an observed reduction in size of telangiectasia and generalized erythema of the right middle turbinate. This patient also reported an ESS reduction of 1.5 points by end of study (baseline ESS =5).

Figure 3 Patient 14 endoscopic image of right middle turbinate from baseline compared to end-of-intervention at 6 months. This patient was part of the propranolol group. There is a reduction in size of telangiectasias and generalized erythema when comparing the baseline to end of treatment period.

Quality of life outcomes

Baseline mean SF-36 scores showed that both groups had lower quality of life than the general population average (16). Patients receiving propranolol had improved scores across all domains by end of treatment (Figure 4), however no statistically significant change was reported in either treatment arms.

Figure 4 Mean change (95% CI and P value) in SF-36 domain scores in the propranolol treatment group over the 6-month treatment period (n=10). CI, confidence interval; SF-36, Short Form-36.

On sub-analysis, patients that exhibited sustained improvement on propranolol (type A response) also had the greatest improvements in role limitations due to emotional problems (mean change 53.00, 95% CI: 27.34 to 134.01), social functioning (10.00, 95% CI: −27.05 to 47.05) and emotional well-being (4.00, 95% CI: −19.02 to 27.02).

Haematology tests

At 3 months, the propranolol group showed a statistically significant increase in hemoglobin (Hb) levels (mean increase: 7.7 g/L, 95% CI: 1.57 to 13.93, P=0.018). While statistically significant, this increase is unlikely to be clinically meaningful and by 6 months follow up these changes had normalised back towards baseline. No other significant changes were observed in platelet count, transferrin saturation, ferritin or iron.

Adverse events

Hypotension and/or postural dizziness was the main reported side effect in the propranolol group (n=3). This resulted in three patient dropouts that occurred within the first 2 weeks of enrolment. Of these patients, one had a measured systolic BP drop from 119 to 90 mmHg and was discontinued based on advice from a cardiologist. The other two patients did not have a measured postural BP drop, and their dizziness resolved upon cessation of the study. Limited risk analysis of these patients did not find an association with age, sex or baseline co-morbidities. On average, there was an observed drop in BP compared to placebo as expected in use of a beta-blocker over time (Figure 5). Overall, no unexpected side-effects were reported and no physical injury or hospital admission was sustained in the propranolol arm. In the placebo group, headache, leg swelling, fatigue and dizziness were reported in a patient; these symptoms resolved on cessation of placebo.

Figure 5 Mean change in blood pressure over time. (A) Systolic and (B) diastolic.

Discussion

Our results suggest early evidence that oral propranolol may have an effect on severity of epistaxis in patients with HHT. While reduction in epistaxis scores in subjects receiving oral propranolol did not translate into statistical significance compared to placebo, the observed clinical benefit was above the MCID of ESS of 0.71 (12). In particular, those with the highest baseline ESS scores received a sustained improvement with corresponding improvement in general quality of life domains. Almost all patients in the intervention group (except one) elected to continue oral propranolol on completion of the study, further supporting the potential uptake of oral propranolol in clinical practice.

The benefit of oral propranolol is the ease of administration in a community setting, accessibility, relatively low cost, and well-established safety profile. Patients experiencing a mild-moderate disease burden that are not yet willing to commit to surgical options and associated risks, or intravenous bevacizumab may benefit most. Other demographics anecdotally identified in this study that reported unique benefit were those living in rural or remote settings, with limited access to specialist rhinology service, and patients with full-time employment.

In terms of disease-modifying mechanisms of propranolol in HHT, this study is unable to provide specific evidence for existing theories. However, we do hypothesise that in addition to the potential reduction in VEGF expression and endothelial cell regulation, that systemic reduction in BP observed in this study due to propranolol may also contribute to reduced epistaxis.

Hypotension and postural dizziness were the most frequently encountered adverse events and symptoms manifested within the first 3 weeks of starting therapy. Regular BP and pulse monitoring is recommended as a key aspect of this intervention. The 40 mg twice daily dosing was based on the 1–2 mg/kg dosing used in infantile haemangioma and in discussion with a cardiologist. It is recognised that conversion from child to adult dose equivalence has several limitations and future dose adjustments may be of particular focus. An initial lower dose up-titrated to desired effect as used in other propranolol-related indications may be better tolerated (17,18). Other side effects not observed in this study, but in broader literature include fatigue, sleep disturbance, and erectile dysfunction (19). Specific avoidance in patients with underlying lung pathologies such as chronic obstructive pulmonary disease, asthma or emphysema is advised due to the risk of blockage of beta-2 receptors causing vasoconstriction of smooth muscle and worsening respiratory function (19).

Nasal endoscopy was performed in study groups against the HHT Endoscopic System (HES); scoring objective nasal disease burden (telangiectasia size, distribution, arterio-venous malformations, crusting, etc.) No trend in the HES scores were reported despite some cases which showed clear objective changes in the propranolol group on retrospective review of videos (Figure 3). Endoscopy scoring was performed by multiple otolaryngologists throughout the study, in contrast to the original design of the HES tool which relied on a single surgeon. As such, this likely impacted consistency of reporting, and future study designs should include cross-examination and agreement. Use of the HES system in evaluation of treatment response in HHT is also relatively novel, and the tool may also require further validation.

Performing randomised, placebo controlled clinical trials for rare conditions such as HHT is challenging. This trial had several limitations particularly with patient recruitment. Patients were reluctant to enroll due to the risk of receiving placebo treatment. This was particularly evident in transfusion-dependent patients, potentially contributing to selection bias. Many patients in the pilot study continued oral propranolol as open label after study closure and did not want to re-enroll in the randomized controlled trial (RCT). This study faced recruitment delay due to coronavirus disease 2019 (COVID-19) clinic restrictions. It is possible that if the study had met 20 subjects in each group that was initially estimated to demonstrate treatment effect with 80% power at significance level of 0.05 (two-sided), greater statistical significance could have been achieved. The small sample size may have also impacted randomisation, leading to chance variations in potentially important baseline characteristics. For instance, ferritin levels differed between the two groups despite no major difference in other iron parameters or Hb. The clinical significance remains uncertain.

Future studies may benefit from a multi-centre study design with the aim to recruit higher patient numbers. Another randomised-control trial, or an open-label trial that aims to define parameters to target optimal therapy and patient selection may be of particular benefit.

Other beta-blockers such as atenolol, which has evidence of less frequent adverse events with similar response rates to propranolol in infantile haemangiomas (10,20) could be trialled, as well as combination with other HHT treatments.

Conclusions

Our study suggests oral propranolol may be a potential candidate for reducing epistaxis severity in patients with HHT. While statistical significance compared to placebo was not reached in this small cohort, the observed clinical benefit in the intervention arm was above the MCID ESS score. Larger studies are required before clinical recommendations can be proposed, with additional focus on dose adjustments and combination therapy. Propranolol’s ease of administration and well-defined safety profile make it an appealing potential addition to current medical treatment options for HHT.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the CONSORT reporting checklist. Available at https://www.theajo.com/article/view/10.21037/10.21037/ajo-24-64/rc

Trial Protocol: Available at https://theajo.com/article/view/10.21037/ajo-24-64/tp

Data Sharing Statement: Available at https://www.theajo.com/article/view/10.21037/ajo-24-64/dss

Peer Review File: Available at https://www.theajo.com/article/view/10.21037/ajo-24-64/prf

Funding: This study was funded as an in-house Royal Brisbane and Women’s Hospital study and did not receive any funding from third-party sources.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://www.theajo.com/article/view/10.21037/ajo-24-64/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The trial was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Royal Brisbane and Women’s Hospital Research Governance Office (HREC/2019/QRBW/4855) and informed consent was taken from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Rimmer J, Lund VJ. Hereditary haemorrhagic telangiectasia. Rhinology 2015;53:195-203. [PubMed]
2. Al-Samkari H, Kasthuri RS, Parambil JG, et al. An international, multicenter study of intravenous bevacizumab for bleeding in hereditary hemorrhagic telangiectasia: the InHIBIT-Bleed study. Haematologica 2021;106:2161-9. [Crossref] [PubMed]
3. Ardelean DS, Letarte M. Anti-angiogenic therapeutic strategies in hereditary hemorrhagic telangiectasia. Front Genet 2015;6:35. [Crossref] [PubMed]
4. Chim H, Armijo BS, Miller E, et al. Propranolol induces regression of hemangioma cells through HIF-1α-mediated inhibition of VEGF-A. Ann Surg 2012;256:146-56. [Crossref] [PubMed]
5. Kum JJ, Khan ZA. Mechanisms of propranolol action in infantile hemangioma. Dermatoendocrinol 2015;6:e979699. [Crossref] [PubMed]
6. Mei-Zahav M, Gendler Y, Bruckheimer E, et al. Topical Propranolol Improves Epistaxis Control in Hereditary Hemorrhagic Telangiectasia (HHT): A Randomized Double-Blind Placebo-Controlled Trial. J Clin Med 2020;9:3130. [Crossref] [PubMed]
7. Andorfer KEC, Zeman F, Koller M, et al. TIMolol Nasal Spray as a Treatment for Epistaxis in Hereditary Hemorrhagic Telangiectasia (TIM-HHT)-A Prospective, Randomized, Double-Blind, Controlled, Cross-Over Trial. Pharmaceutics 2022;14:2335. [Crossref] [PubMed]
8. Contis A, Gensous N, Viallard JF, et al. Efficacy and safety of propranolol for epistaxis in hereditary haemorrhagic telangiectasia: retrospective, then prospective study, in a total of 21 patients. Clin Otolaryngol 2017;42:911-7. [Crossref] [PubMed]
9. Peterson AM, Lee JJ, Kallogjeri D, et al. Efficacy of Timolol in a Novel Intranasal Thermosensitive Gel for Hereditary Hemorrhagic Telangiectasia-Associated Epistaxis: A Randomized Clinical Trial. JAMA Otolaryngol Head Neck Surg 2020;146:1006-14. [Crossref] [PubMed]
10. Albarki H, Rimmer J. The Use of Beta-Blockers in Hereditary Hemorrhagic Telangiectasia-Related Epistaxis: A Systematic Review. Am J Rhinol Allergy 2022;36:890-6. [Crossref] [PubMed]
11. Hoag JB, Terry P, Mitchell S, et al. An epistaxis severity score for hereditary hemorrhagic telangiectasia. Laryngoscope 2010;120:838-43. [Crossref] [PubMed]
12. Yin LX, Reh DD, Hoag JB, et al. The minimal important difference of the epistaxis severity score in hereditary hemorrhagic telangiectasia. Laryngoscope 2016;126:1029-32. [Crossref] [PubMed]
13. Lins L, Carvalho FM. SF-36 total score as a single measure of health-related quality of life: Scoping review. SAGE Open Med 2016;4:2050312116671725. [Crossref] [PubMed]
14. Reh DD, Yin LX, Laaeq K, et al. A new endoscopic staging system for hereditary hemorrhagic telangiectasia. Int Forum Allergy Rhinol 2014;4:635-9. [Crossref] [PubMed]
15. IBM SPSS statistics for Windows. Version 27. IBM Corp.; 2020.
16. McCallum J. The SF-36 in an Australian sample: validating a new, generic health status measure. Aust J Public Health 1995;19:160-6. [Crossref] [PubMed]
17. Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): The Task Force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) Developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J 2021;42:373-498. [Crossref] [PubMed]
18. Akbar S, Alorainy MS. The current status of beta blockers› use in the management of hypertension. Saudi Med J 2014;35:1307-17. [PubMed]
19. Tucker WD, Sankar P, Theetha Kariyanna P. Selective Beta-1 Blockers. StatPearls. 1st ed. Florida: StatPearls Publishing; 2023.
20. Ji Y, Chen S, Yang K, et al. Efficacy and Safety of Propranolol vs Atenolol in Infants With Problematic Infantile Hemangiomas: A Randomized Clinical Trial. JAMA Otolaryngol Head Neck Surg 2021;147:599-607. [Crossref] [PubMed]