Supplemental Tables

Second International HHT Guidelines

Supplemental Table 1

Supplement Table 1. Curaçao Criteria for Clinical Diagnosis of HHT: Using these criteria, a diagnosis of HHT is considered ‘definite’ if three or more Curaçao criteria are present, ‘possible or suspected’ if two criteria are present, and ‘unlikely’ if 0 or 1 criterion is present. 

Criteria  Description
Epistaxis  Spontaneous and recurrent
Telangiectases  Multiple, at characteristic sites: lips, oral cavity, fingers, nose
Visceral lesions  Such as gastro-intestinal telangiectasia, pulmonary, hepatic, cerebral or spinal AVMs
Family history  A first degree relative with HHT according to these criteria

 

Supplemental Table 2

Supplement Table 2:  Randomized Controlled Trials for Treatment of Epistaxis in HHT.  All trials were performed in adults (Age 18+) and included only patients with definite clinical diagnosis of HHT.  
Study, Year (Reference) Participants Intervention Design and Methods Primary Outcome Measures Primary Outcome Results
Boyer H. et al. Int Forum Allergy Rhinol. 2015. (15) N=17 Sclerotherapy versus Control ("standard treatment", defined as  continuation of any treatment that the patient had previously undergone) RCT (crossover)        Treatment= 6 weeks, each

Washout period: None. 

ESS Improved ESS scores (0.95 difference, 1-sided p = 0.027), 

The standard deviation of the difference scores was 1.82. 

Treatment order was not statistically significant.

Dupuis-Girod S. et al. JAMA 2016 (27) N=80 Bevacizumab nasal spray (25mg, 50mg, or 75mg) for 4 weeks vs placebo nasal spray RCT Phase II-III (placebo controlled)                  Treatment:  doses 14 days apart for a total treatment duration of 4 weeks, resulting in a total dose of 75mg, 150mg, and 225mg in each treatment group. Mean monthly epistaxis duration  for 3 months AFTER end of treatment compared with  3 months BEFORE  beginning of treatment. No statistical difference was observed in mean monthly epistaxis duration among treatment groups and placebo (p = .57), with higher standard deviation than expected in trial design.  
Gaillard S. et al. J Thromb Haemost. 2014 (10)              N=135 Oral tranexamic acid (3g per day) versus placebo RCT (double-blind, placebo controlled crossover):    Treatment: 3 months, each.             Mean monthly epistaxis duration  for last 2 months of the treatment  compared with the last 2months on placebo. The mean duration of epistaxis per month was significantly shorter with tranexamic acid than placebo (0.19 on the log scale; SD = 0.07; p = 0.005). This difference corresponded to a decrease of 17.3% in the duration of epistaxis per month (95% CI, 5.5–27.6).
Geisthoff U.W. et al. Thromb Res.2014   (11) N=22 Oral tranexamic acid (3g per day) versus placebo RCT (double-blind, placebo controlled crossover):    Treatment: 3 months, each. 

Washout period: None.

Delta hemoglobin (final minus initial) for each treatment period. No significant difference in delta hemoglobin between tranexamic acid and placebo was detected (p=0.33, Mann–Whitney-U test).

Post-hoc analysis: Mean hemoglobin concentrations were significantly greater for tranexamic acid versus placebo (p=0.013, Mann– Whitney-U test).

Riss D. et al. Head Neck. 2015 (28) N=15 Single dose of intranasal submucosal injection of bevacizumab  RCT (double-blind, placebo controlled, parallel group, stratified by age and epistaxis severity). Patients received a single intranasal submucosal injection of 100 mg of bevacizumab in 10 mL saline or placebo (10 mL saline).  5 mL were injected into each side of the nose. Average daily post treatment epistaxis VAS score (range, 0–100) compared to  average daily pretreatment score in the month before the intervention (R = VAS-post/VAS-pre), for days 11-84.Patients recorded in a diary their daily epistaxis VAS scores ranging from 0 (best situation) to 100 (worst case).  Average daily VAS scores dropped from 18.8 (±16.5 SD) pretreatment to 13.4 (±11.6 SD) post-treatment in the bevacizumab group and from 20.5 (±13.4 SD) to 19.7 (±12.6 SD) in the placebo group.  No significant difference between average daily post-treatment VAS score compared to the average daily pretreatment score (p = 0.57). 
Whitehead K. et al. JAMA 2016 (9) N=121 Topical therapy with bevacizumab 1% (4 mg/d) OR estriol 0.1% (0.4 mg/d) OR tranexamic acid 10% (40 mg/d) nasal sprays RCT Phase II (double-blind, placebo controlled, stratified by epistaxis frequency) 4 treatment groups  (bevacizumab 1% (4 mg/d), estriol 0.1% (0.4 mg/d), tranexamic acid 10% (40 mg/d), or placebo (0.9% saline)   for 12 weeks. Median weekly epistaxis frequency (weeks 5-12) for each patient. Epistaxis frequency was not significantly different between any of the active drug groups and the placebo group or between any of the therapeutic agents.
Yaniv E. et al. Laryngoscope 2009 (64)  N=25 Oral antiestrogen, Tamoxifen 20mg once daily RCT (double-blind, placebo controlled)                      Treatment period=6 months.    Washout period: None. Frequency of epistaxis, duration of epistaxis, hemoglobin level.  Epistaxis frequency was significantly less in the treatment groups (p = 0.01), as was epistaxis severity (p = 0.049) at 6 months.

There was no significant difference in hemoglobin between groups at 6 months.

Abbreviations: ESS= Epistaxis severity score, RCT=Randomized controlled trial, VAS=visual analogue scale

Supplemental Table 3

Supplement Table 3: Lower Quality Uncontrolled Clinical Trials for Treatment of Epistaxis in HHT.  All trials were performed in adults (Age 18+) and included only patients with HHT diagnosis. 
Study, Year (Reference) Study Design Intervention Outcome of Interest Outcome Results
Reh  D.D. et al.  Laryngoscope 2013 Prospective study (N=20) topical lubricant ESS Mean ESS improved (p<0.0001) at 3mo.
Fernandez-L A.  et al. Thromb Haemost 2007 Prospective study  (N=14) oral tranexamic acid Epistaxis frequency&severity 100% patients improved.
Zaffar N.  et al.Ann Hematol.  2015 (12) Retrospective study (N=29) oral tranexamic acid ESS Mean ESS improved (p<0.001).
Jorgensen G. et al. Eur Arch Otorhinolaryngol 2011 Prospective study (N=30) laser Epistaxis duration Epistaxis duration reduced (p<0.05) at 1.5 mo.& 6 mo.
Kuan E.C. et al. Lasers Med Sci  2017 (13) Retrospective study (N=20) laser  SNOT-22 Mean SNOT-22 improved at 1.5mo.
Fiorella M.L.  et al. ACTA otorhinolaryngologica italica 2012 Retrospective study (N=24) laser (diode) Epistaxis frequency&severity Group improved.
Poje G. et al. ENT-Ear, Nose & Throat Journal 2017 Retrospective study (N=17) laser (diode) Epistaxis frequency&severity Group improved.
Papaspyrou G. et al. ORL 2016 Retrospective study (N=38) laser (Nd:YAG) Need for recurrent intervention Recurrent intervention in 18% at 3 years.
Papaspyrou G. et al.J Craniomaxillfac Surgery 2017 Prospective study (N=45) laser (Nd:YAG) +/- APC Need for recurrent intervention Recurrent intervention in 20-33%% at 3-10 years.
Abdelghany  A.M. et al. 2013 Prospective study (N=16) radiofrequency coblation Epistaxis frequency&intensity 100% patients improved.
Luk L. et al. 2014 Prospective study (N=11) radiofrequency coblation vs laser (KTP) ESS No significant difference in mean ESS, at 12mo.
Mortuaire G. et al. 2013 Prospective study  (N=16) radiofrequency coblation Epistaxis frequency&duration Reduced mean epistaxis frequency (p<0.05) at 6mo. 
Rotenberg B. et al. 2015 (17) Retrospective study (N=37) radiofrequency coblation ESS Mean ESS improved (p=0.02) at 6 mo.
Boyer H. et al. 2011 (14) Retrospective study (N=7) sclerotherapy Epistaxis frequency&severity 100% patients improved.
Morais  D.et al. 2012 Retrospective study (N=45) sclerotherapy Epistaxis frequency&severity 95% patients improved.
Pagella F.et al. 2013 Retrospective study (N=26) thermal coagulation (APC) Epistaxis score Mean score improved (p=0.005) at 12 mo.
Pagella F. et al.2006 Prospective study (N=36) thermal coagulation (APC) Reported  bleeding 100% reported reduction in bleeding at 6mo. 
Al-Samkari H. et al. 2018 (23) Retrosepctive study (N=13) IV bevacizumab Epistaxis control Epistaxis control (reduction in epistaxis grade to <2) was achieved in 85% of patients, from 0 patients at baseline (p < 0.001).
Dupuis-Girod S. et al.  2012 (18) Prospective study (N=25)  IV bevacizumab Reported bleeding duration  Mean duration of epistaxis, significantly decreased from 221 minutes per mo. at baseline to 43 minutes per mo. at 3 mo. (p= 0.008). 
Epperla et al. 2016 (22) Retrospective study (N=5) IV bevacizumab blood transfusions blood transfusions were reduced from baseline in 5/5 patients.
Iyer V. et al. 2018 (21) Retrospective study (N=34) IV bevacizumab ESS Significant reduction in ESS from baseline to 3mo (p<0.001).
Faughnan M.E. et al. 2019 (35) Prospective study (N=7) oral pazopanib Epistaxis duration 6/7 patients had >50% decrease, from baseline to during treatment.
Baysal M. et al. 2019 (34) Retrospective study (N=6) oral thalidomide ESS Mean ESS improved from pre-treatment (7.40+/-2.02) to post-treatment (3.10+/-1.79)  (p=0.028).
Fang J. et al. 2017 (31) Prospective study  (N=7) oral thalidomide ESS Mean ESS improved from pre-treatment (5.03 +/- 2.05), to  end treatment (0.90 +/- 0.84, p= 0.003) and to 3 mo. after end treatment (1.98 +/- 1.33,  p= 0.006), respectively.
Invernizzi R. et al. 2015 (32) Prospective, Phase II (N=31) oral thalidomide frequency, intensity, or duration of epistaxis. All patients responded to therapy with a significant decrease in all epistaxis parameters (p<0.0001 for frequency, intensity, and duration).
Lebrin F. et al. 2010 (30) Prospective study (N=7) oral thalidomide Epistaxis severity Self-reported severity of epistaxis improved in 5/7 (71%) of patients after treatment.
Peng H. et al. 2015 (29) Prospective study (N=5) oral thalidomide ESS Mean ESS improved from pre-treatment (6.966 +/- 3.093) to post-treatment (1.799 +/- 0.627) significantly (p = 0.009).
Ichimura K. et al. 2012 Prospective study (N=7) nasal closure Epistaxis cessation 57% had cessation of epistaxis.
Lund V. et al. 2017 (40) Retrospective study (N=100) nasal closure Epistaxis cessation 50% of patients responded. 94% had cessation of epistaxis.
Richer S. et al. 2012 (38) Retrospective study (N=43) nasal closure Epistaxis cessation 84% of patients responded. 83% had cessation of epistaxis.
Wirsching K.E.C. et al. 2017 Prospective study (N=20) temporary nasal occlusion with tape ESS ESS decreased from pre-treatment median of 3.59 to post-treatment (at 3 mo.) median of 2.43 (p = 0.01).
Harvey  R. et al. 2008 Retrospective study (N=33) septodermoplasty Frequency of KTP laser Number of  KTP laser treatments decreased from 1.83 (+/-1.99) pre-septodermoplasty to  0.78 (+/-0.85) post-septodermoplasty (p=0.012).
Ichimura K. et al. 2006 Retrospective study (N=15) septodermoplasty Patient satisfaction 100% of patients satisfied with procedure.
Lesnik G.T. et al. 2007 (37) Retrospective study (N=9, severe) septodermoplasty plus septectomy Epistaxis frequency, QOL and blood transfusions All patients had improved self-reported QOL. Blood transfusions were reduced from baseline 22.61/year to 9.57/year post-procedure (p < 0.05). 
Levine  C.G. et al. 2008 (36) Retrospective study (N=106) septodermoplasty QOL 62% of patients responded: 86% patients had improved QOL  at mean 3.75 years.
Rimmer J. et al. 2014 (41) Prospective study (N=7) septodermoplasty Epistaxis frequency&severity 100% of patients reported reduction in epistaxis frequency and severity.
Abbreviations: ESS= Epistaxis severity score, mo.=month, QOL=quality of life, SNOT=sinonasal outcome test-22

 

Supplemental Table 4

Supplement Table 4: Guidance for Prescribing and Safety Monitoring of Systemic Therapies for in HHT patients (as per recommendations A2, A4, B5, D5)

 

Drug Suggested Dosing Safety Comments
Oral tranexamic acid Start at 500mg BID, gradually increasing up to 1000mg QID or 1500mg TID.  Can be co-administered with systemic anti-angiogenic therapy.

Contraindications: Recent VTE or arterial thrombosis.  

Relative contra-indications: Atrial fibrillation, thrombophilia or other procoagulant tendencies (e.g. elevated Factor VIII).

Intravenous bevacizumab (induction) 5 mg/kg every 2 weeks for 6 doses. Monitor for:
  • Hypertension
  • Proteinuria
  • Delayed wound healing (avoid major surgery)
  • Infection
  • VTE

Contraindicated in pregnancy.

Intravenous bevacizumab (maintenance) Variable, from none to 5mg/kg every 1-3 months for 1 year, followed by gradually longer intervals. As above. 

Additional risks of long-term maintenance therapy not known.

Abbreviations: VTE=venous thromboembolism

Supplemental Table 5

Supplement Table 5: Diagnostic yield of gastrointestinal diagnostic procedures in adults with definite HHT.
All studies were in adults (18 years+).
Study, Year (Reference) Population   Tests Diagnostic Yields
Canzonieri C. et al. 2014 (45) Definite HHT, consecutive adults, 22 (13 male), mean age 59 years (+/-9) Esophagogastroduodenoscopy

Capsule endoscopy

Colonoscopy

82%

91%

10%

Chamberlain S.M. et al. 2007 (49) Definite HHT, consecutive adults with suspected GI bleeding, 32/38 complete  (18 male), mean age 54 years (+/-13) Capsule endoscopy Any GI telangiectasia=81%

Gastric=28%

Proximal small bowel=56%

Mid small bowel=59%

Distal small bowel=63%

Chetcuti Zammit S.  et al. 2018 (60) Definite HHT, consecutive adults with suspected GI bleeding, 10 patients  (6 male), mean age 63 years (+/-14) Capsule endoscopy (N=7)                                                                                                      Proximal small bowel=86%

Mid small bowel=11%

Distal small bowel=33%

Greve E. et al. 2010 (44) Definite HHT, consecutive adults with anemia and suspected GI bleeding, 30 patients  (10 male), mean age 58 years (+/-11) Capsule endoscopy Gastric=47%

Small bowel=87%

van Tuyl S.A. et al. 2007 (43) Definite HHT, consecutive adults with anemia, 25 patients  (13 male), mean age 49 years (+/-15) Esophagogastroduodenoscopy

Capsule endoscopy

Colonoscopy

67%

76%

32%

 

Supplemental Table 6

 

Supplement Table 6: Lower Quality Uncontrolled Clinical Trials for Treatment of GI Bleeding in HHT. All trials were performed in adults (Age 18+) and included only patients with HHT diagnosis.
Study, Year (Reference) Study Design Intervention Outcome of Interest Outcome  Results
Zaffar N. et al.  2015 (12) Retrospective study (N=29, of whom 10 had GI bleeding) oral tranexamic acid Requirement for any GI-endoscopic intervention            Reduced from 80% pre-treatment to 40% on treatment (trend, p=0.07).
Al-Samkari H. et al. 2019 (23) Retrospective study (N=13, of whom 10 had GI bleeding) IV bevacizumab Change in hemoglobin.          Reduction in pRBCs Mean hemoglobin improved by 4g/dL or by 45% from the pre-treatment period to the maintenance period (P<0.001).

pRBC requirements decreased by 92% from the pretreatment period to the maintenance period.     

Iyer V. et al. 2018 (21) Retrospective study (N=34, of whom 19 had GI bleeding) IV bevacizumab Requirement for any GI-endoscopic intervention            Significant reduction in RBC transfusions (p=0.007) in the entire group (GI bleeders not reported separately).
Faughnan M.E. et al. 2019 (35) Prospective study (N=7) oral pazopanib Epistaxis duration 6/7 patients had >50% decrease, from baseline to during treatment.
Abbreviations: GI=gastrointestinal, pRBC=packed red-blood cells

 

Supplemental Table 7

 

Supplement Table 7: Diagnostic Accuracy of Testing for Liver VMs in Adults with Definite HHT.  All studies were in adults (18 years+) and reported measures of diagnostic accuracy or agreement.
Study, Year (Reference) Population Tests Operating Characteristics
Buonamico P. et al. 2008 (124) Definite HHT (N=153) Ultrasound Doppler "color spots".

Multiphase CT as reference standard

Sensitivity=95%

Specificity=68%

Diagnostic accuracy=92%

Buscarini E. et al. 2008 (125) Definite HHT (N=110) Ultrasound Doppler Sensitivity=97-99% 

Specificity=97-99% 

Moderate inter-observer agreement (Kendall's coefficient of concordance=0.26) for severity.

Cavel A. et al. 2016 (126) Confirmed or suspected HHT (N=62) Ultrasound Doppler versus Multiphase CT  Significant disagreement with kappa=0.376 and a Bhapkar critical probability of p=0.0053. Staging of liver involvement was significantly more severe with CT in cases of disagreement.
Milot L. et al. 2008 (128) Definite HHT (N=23) versus Controls (N=23) MRI liver Hepatic artery diameter: greater in HHT patients than in controls: 8.69+/-1.63 mm versus 5.17+/-0.44 mm (p<0.05).

Vascular abnormalities: 91% HHT vs 0% controls

Ischemic cholangitis: 39% HHT vs 0% controls

Good interobserver agreement for vascular abnormalities (0.62)

Moderate interobserver agreement (0.42) with biliary ischemia.

Scardapane A. et al.2012 (129) Definite HHT (N=52) Multiphase CT versus 4D-MRA CT Diagnostic Yield=69%

MRA Diagnostic Yield=69%

No significant difference accuracy

Kappa=0.9 (good) for type of shunt.

Wu J.S.et al. 2006 (127) Definite HHT and symptomatic liver VMs (N=24) Multiphase CT Diffuse telangiectasias: 100%. 

Dilated hepatic artery: 100%

Cardiomegaly: 48%   

Hepatic arteriovenous shunt: 54%

Arterioportal shunt: 25% 

Agreement between CT and clinical type: 54%

Abbreviations: CT=computed tomography, VM=vascular malformation

Supplemental Table 8

 

Supplement Table 8: Lower Quality Uncontrolled Clinical Trials for Treatment of Liver VMs in HHT. All trials were performed in adults (Age 18+) and included only patients with HHT diagnosis.
Study, Year (Reference) Study Design Intervention Outcome of Interest Outcome  Results
Dupuis-Girod A. et al. 2012 (18) Uncontrolled series (N=25, HHT with HOCF from liver VMs) IV bevacizumab Decrease in cardiac output (from high-output state) Cardiac output improved or normalized in 83%
Azzopardi N. et al.2015. Uncontrolled series (N=25, HHT with HOCF from liver VMs) IV bevacizumab (maintenance dosing) Maintenance of improved cardiac output with different length bevacizumab intervals, after induction Every 3 mo.:  Maintained in 41%

Every 2 mo.:  Maintained in 45%

Every 1 mo.:  Maintained in 50% 

(All at 24 mo.)

Chavan A. et al. 2017 (20) Uncontrolled series (N=21, HHT with symptomatic liver VMs) IV bevacizumab Clinical symptom improvement Abdominal pain grade improved from 3·0 ± 2·2 (95% CI 1·99–3·91) pretherapy to 0·9 ± 1·0 post‐therapy (95% CI 0·48–1·33) (p <0 .001).

Mean NYHA stage improving from 2·8 ± 0·7 (95% CI 2·49–3·13) pretherapy to 1·6 ± 0·9 (95% CI 1·25–1·99) (p <0 .0001) following therapy.

Mean epistaxis grade fell from 2·4 ± 0·8 (95% CI 2·04–2·73) to 0·9 ± 0·4 (95% CI 0·70–1·01) (p<0 .0001). 

Lerut J. et al. 2006 (119) Uncontrolled case series (N=40) Liver transplant Survival post-transplant 1-, 5- and 10-year survival rates= 82.5%. 
Dupuis-Girod S. et al.2010 (134) Uncontrolled case series (N=13) Liver transplant Survival post-transplant Mean follow-up 109mo., Survival 92.3%
Liu Z.C. et al. 2016 Uncontrolled series (N=13, HHT with symptomatic liver VMs) Double banding/ligation of hepatic arteries Clinical effectiveness measures Cardiac function improved: NYHA III-IV vs. NYHA I-II

Pulmonary arterial pressure decreased in all patients (48 +/- 8 mmHg vs. 24 +/- 4 mmHg; p <0 .001).

Gamma-glutamyl transpeptidase and alkaline phosphatase decreased in 11 patients (144 +/- 94 U/L vs. 71 +/- 34 U/L; p=0 .003) and 10 patients (207 +/- 71 U/L vs. 105 +/- 32 U/L; p =0 .001), respectively.

Abbreviations: HOCF=high-output cardiac failure, mo.=month, IV=intravenous, NYHA=New York Heart Association, VMs=vascular malformations

 

Supplemental Table 9

 

Supplement Table 9: Diagnostic Accuracy of Testing for Pulmonary AVMs in Children with Definite HHT.  All studies were in children (<18 years) with reported measures of diagnostic accuracy or agreement for pulmonary AVMs.
Study, Year (Reference) Population Tests Operating Characteristics
Soysal N. et al. 2017 Definite HHT ( N=59) High-resolution CT chest Yield: pulmonary AVMs 25%
Al-Saleh S. et al. 2012 (148) Definite HHT (N=75) TTCE screening 

CT chest (reference standard)

Intraobserver and interobserver agreement for interpreting TTCE results were excellent (kappa = 0.97 and 0.92, respectively)

Sensitivity=100% , Specificity=82%

PPV=39% ,   NPV=100%

Karam C. et al. 2015 (149) Definite HHT (N=93) TTCE screening CT chest (reference standard) Yield: Pulmonary AVMs 52%.

Sensitivity=100%, Specificity= 95%

PPV=96%,   NPV=100% 

Fernandopulle N. et al. 2018 (150) Possible HHT (N=293) TTCE screening CT chest (reference standard) TTCE positive: 26%.

Bubble timing was associated with need for treatment (p=0.008).

Shunt intensity  was associated with presence of CT-detectable pulmonary AVMs (p<0.001) and need for intervention (p=0.005).

Westermann C.J.J. et al. 2003 (171) Definite HHT (N=112) Screening with pulse oximetry and chest x-ray Yield: Pulmonary AVMs 22%, of whom 48% had had serious complication.
Hosman A.E. et al. 2017 (147) Definite HHT (N=175) Screening with pulse oximetry and chest x-ray Yield: Pulmonary AVMs 22%, of whom 85% required embolization.
Abbreviations: AVM=arteriovenous malformation, CT=computed tomography, TTCE= transthoracic contrast echocardiography

Supplemental Table 10

 

Supplement Table 10: Lower Quality Uncontrolled Clinical Trials for Treatment of Pulmonary AVMs and Brain VMs in HHT. All trials were performed in children (<18 years) and included only patients with HHT diagnosis.
Study, Year (Reference) Study Design Intervention Outcome of Interest Outcome  Results
Faughnan M.E. et al. 2004 (146) Definite HHT and pulmonary AVMs (N=42) Transcatheter embolization of pulmonary AVMs Reperfusion rate and safety Reperfusion in 15% of embolized pulmonary AVMs . No serious or long-term procedural complications.
Meybodi A.T. et al. 2018 (166) Definite HHT and brain VMs (N=6 children treated) Surgical management of brain VMs Neurological outcomes 5/6 children: improved or stable mRS post-op and 1/6 had temporarily worsened mRS post-op.
Krings T. et al. 2005 (172) Definite HHT and brain VMs (N=25 children treated, including 14 with brain AVFs) Embolization Clinical outcomes 87% patients had stabilization of the disease, ameliorating the symptoms or even complete resolution.
Abbreviations: AVM=arteriovenous malformation, VM=vascular malformations, mRS=modified Rankin score