Morris J Brown

William Harvey Research InstituteQueen Mary University of London

Research and Trials within the BHS

Topics

• PATHWAYs– Recent (summary of P1, P2, P3)

– Present (mechanisms sub-studies, mainly P2)

– Future (GWAS, Mendelian Randomisation, P4?)

• Studies in Primary Aldosteronism– 11C-metomidate PET CT vs Adrenal Vein Sampling (‘MATCH’)

– Endoscopic radiofrequency ablation of adrenal adenomas(‘FABULAS’)

PATHWAY Steering Committee

Morris J Brown –Chairman

Gordon McInnes,

Thomas MacDonald Peter Sever

Bryan Williams Isla MacKenzie

David J Webb Sandosh Padmanabhan

Mark Caulfield Jackie Salsbury – Co-ordinator

J Kennedy Cruickshank Steve Morant - Statistician

Ian Ford

PATHWAY Executive Committee

Morris J Brown (Chairman): University of Cambridge

Thomas MacDonald: University of Dundee

Bryan Williams: University College London

PATHWAY Study Sites and Investigators

Cambridge: Anne Schumann, Jo Helmy, Carmela Maniero, Timothy J Burton, Ursula Quinn, Lorraine Hobbs, Jo Palmer,

Ixworth: John Cannon, Sue Hood

Birmingham: (2 sites) Una Martin, Richard Hobbs, Rachel Iles Kings College London: Krzysztof Rutkowski

Dundee: Alison R McGinnis, JG Houston, Evekyn Findlay , Caroline Patterson,

Imperial College London: Judith Mackay, Simon A McG Thom, Candida Coghlan

Leicester: Adrian G Stanley, Christobelle White, Peter Lacy, Pankaj Gupta, Sheraz A Nazir, Caroline J. Gardiner-Hill

Manchester: Handrean Soran, See Kwok, KarthiraniBalakrishnan

Edinburgh: Vanessa Melville, Iain M MacIntyre Norwich: Khin Swe Myint, Judith Gowlett

St Barts London: David Collier, Nirmala Markandu, Manish Saxena, Anne Zak, Enamuna Enobakhare

Glasgow: Scott Muir, Linsay McCallum

Data Centre and Monitor

Robertson Centre for Biostatistics, University of GlasgowSharon Kean, Richard Papworth, Robbie Wilson, Ian FordMonitor: Elizabeth Sprunt

BHF Research Programme

£2.6M award, Nov 2007, to BHS for 3 studies

investigating rational treatment algorithms:

‘PATHWAY’ =

Prevention And Treatment of Hypertension

With Algorithm based therapY

Common theme: Should renin

measurement be routine in hypertension?

Summary of Questions

Pathway 1

Could aggressive early treatment of raised blood

pressure prevent subsequent treatment resistance?

Pathway 2

Is resistant hypertension usually due to excessive

Na+ retention? Is spironolactone superior to other

potential add on drugs?

Pathway 3

Are K+ sparing diuretics neutral or beneficial in

their effect on glucose tolerance?

Combination versus sequential monotherapy for initial treatment of hypertension (PATHWAY-1)

N=605

Results: Home SBP

Unadjusted mean home SBP (95% CI) at each visit

Ho

me

sys

tolic

BP

(m

mH

g)

120

125

130

135

140

145

150

155

Weeks from baseline

0 4 8 12 16 24 32 38 44 52

Phase 1 Phase 2 Phase 3

Combination therapy

Monotherapy, HCTZ first

Monotherapy, losartan first

4.9mmHg p < 0.001

+1.2mmHg p=0.13

8mmHg 2.9mmHg

0

20

40

60

80

100

Phase 1 Phase 2 Phase 3

%

Controlled BPHBP< 135/85mmHg or Clinic BP <

140/90mmHg

Combination MonoRx

Proof of AB/CD But combination trumps personalisation:

‘initial combination trumps initialledmonotherapy’

Hom

e sy

stol

ic B

P (m

mH

g)

125

130

135

140

145

150

Renin tertile

Bottom Middle Top

HCTZ

Losartan

Combination

Randomised initial treatment

Combination HCTZ Losartan

Difference (95%

CI) p-value

Difference (95% CI)

p-value Difference (95%

CI) p-value

Top vs Bottom renin tertile(1) -1·41 (-3·52,0·71) 0·193 4·31 (-2·26,6·35) <·001 -3·71 (-5·70,-1·71) <·001

Aged over 55 vs 55 and under(1)

1·45 (-0·29, 3·19) 0·103 -2·94 (-4·73,-1·15) 0·001 -1·89 (-3·62,-0·16) 0·032

Renin (per 10 fold increase) -1·80 (-4·75,1·16) 0·235 4·96 (2·12,7·80) <·001 -3·70 (-6·43,-0·97) 0·008

Age (per 10 years) 0·13 (-0·85,1·12) 0·787 -0·97 (-1·98,0·04) 0·062 -0·20 (-1·18,0·77) 0·682

Baseline HSBP 0·29 (0·22,0·36) <·001 0·48 (0·42,0·54) <·001 0·55 (0·48,0·61) <·001

Never vs previously treated 1·83 (-0·41,4·08) 0·111 -3·01 (-5·26,-0·77) 0·009 -2·85 (-4·96,-0·73) 0·009

Randomised initial treatment

Combination HCTZ Losartan

Difference (95%

CI) p-value

Difference (95% CI)

p-value Difference (95%

CI) p-value

Top vs Bottom renin tertile(1) -1·41 (-3·52,0·71) 0·193 4·31 (-2·26,6·35) <·001 -3·71 (-5·70,-1·71) <·001

Aged over 55 vs 55 and under(1)

1·45 (-0·29, 3·19) 0·103 -2·94 (-4·73,-1·15) 0·001 -1·89 (-3·62,-0·16) 0·032

Renin (per 10 fold increase) -1·80 (-4·75,1·16) 0·235 4·96 (2·12,7·80) <·001 -3·70 (-6·43,-0·97) 0·008

Age (per 10 years) 0·13 (-0·85,1·12) 0·787 -0·97 (-1·98,0·04) 0·062 -0·20 (-1·18,0·77) 0·682

Baseline HSBP 0·29 (0·22,0·36) <·001 0·48 (0·42,0·54) <·001 0·55 (0·48,0·61) <·001

Never vs previously treated 1·83 (-0·41,4·08) 0·111 -3·01 (-5·26,-0·77) 0·009 -2·85 (-4·96,-0·73) 0·009

PATHWAY-2 Study of Resistant Hypertension

Spironolactone25 – 50mg o.d.

Doxazosin MR4 – 8mg o.d.

Bisoprolol5 – 10mg o.d.

Placebo

Screening forResistant Hypertension• Rx A + C + D• DOT* to exclude non-

compliance• Home BP to exclude

white coat hypertension

• Secondary hypertension excluded

4 weekSingle blind placebo run in

Treated with A+C+D

Randomisation

*DOT = Directly Observed Therapy

Double blind, Randomised, Placebo-Controlled, Cross-over Study

• 12 weeks per treatment cycle• Forced titration; lower to higher dose at 6 weeks• No washout period between cycles

Home Systolic BP

measured at6 and 12 weeks

Williams B, et al. BMJ Open, 2015

AmilorideOpen-Label

Run-out10 -20mg

o.d.

Plasma Renin

Primary Outcome

PATHWAY-3 study of amiloride vs HCTZ

• Amiloride will have the opposite effect to hydrochlorothiazide (HCTZ) on K+ and glucose, but same effect on blood pressure.

• Combination of diuretics with different sites of action in the nephron will be synergistic for Na+ loss and hence BP reduction

• Consequently, the combination of half-maximal doses of amiloride and HCTZ will:

– Neutralise the undesired effects of HCTZ, on glucose and K+

– Potentiate the desired effect of HCTZ, on blood pressure

Adjusted means (95% CI) for change from baseline in 2 hr glucose during OGTT. Doses were doubled at 12 weeks. **=p<0.01 vs HCTZ

Hierarchical primary endpointsDifference in change from baseline in OGTT 2 hr glucose

for [i] amiloride vs HCTZ, [ii] combination vs HCTZ

2 h

r g

lucose:

ch

an

ge f

rom

baselin

e

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

Baseline 12 weeks 24 weeks

**

Hydrochlorothiazide (HCTZ) 25-50 mgAmiloride 10-20 mg

Amiloriden=132

Amiloride/HCTZ n=133

-0.55 (-0.96,-0.14)

P=0.009

Average difference from HCTZ (mmol/L) (12 & 24 weeks)

Hierarchical primary endpointsDifference in change from baseline in OGTT 2 hr glucose

for [i] amiloride vs HCTZ, [ii] combination vs HCTZ

Adjusted means (95% CI) for change from baseline in 2 hr glucose during OGTT. Doses were doubled at 12 weeks. **=p<0.01 vs HCTZ; *=p<0.05 vs HCTZ

2 h

r g

lucose:

ch

an

ge f

rom

baselin

e

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

Baseline 12 weeks 24 weeks

***

Hydrochlorothiazide (HCTZ) 25-50 mgAmiloride 10-20 mgAmiloride/HCTZ combination 5/12.5 -10/25 mg

High-dose difference from HCTZ (mmol/L) (24 weeks)

Amiloriden=132

Amiloride/HCTZ n=133

-0.73 (-1.20,-0.25)-0.50 (-0.98, -

0.025)

P=0.005 P=0.024

Secondary endpointsBlood Pressure reduction

Hom

e S

BP

(m

mH

g)

125

130

135

140

145

150

Weeks from baseline

0 12 24

HCTZAmilorideCombination

Hom

e S

BP

(m

mH

g)

125

130

135

140

145

150

Weeks from baseline

0 12 24

*

HCTZAmilorideCombination

Home SBP (mean, 95% CI) adjusting for baseline covariates

* p=0.02 for combination vs HCTZ at week 24.

Across weeks 12 (low-dose) and 24 (high-dose), BP fall on combination of amiloride and HCTZ was 3·4 (0·9, 5·8) mmHg greater than on HCTZ (p=0·007)

n-f

old

in

cre

ase in

ren

in

1

2

4

8

16

Weeks from baseline

0 12 24

***

Hydrochlorothiazide (HCTZ) 25-50 mg

Amiloride 10-20 mg

Combination (Amiloride/HCTZ 5/12.5-10/25 mg)

The Prevention And Treatment of Hypertension With Algorithm based therapYPATHWAY

Professor Bryan Williams FESC Chair of Medicine | University College London

Tom MacDonald FESC, Steve Morant and Morris Brown FESC

on behalf of the PATHWAY Investigators

Mechanisms for benefit of spironolactone in resistant hypertension in the PATHWAY-2 study

PATHWAY-2 Mechanisms study

Spironolactone25 – 50mg o.d.

Doxazosin MR4 – 8mg o.d.

Bisoprolol5 – 10mg o.d.

Placebo

Screening forResistant Hypertension

• Treatment A + C + D• DOT* to exclude non-

compliance• Home BP to exclude

white coat hypertension• Secondary hypertension

excluded

4 weekSingle blind placebo run in

Treated with A+C+D

Randomisation

*DOT = Directly Observed Therapy

12 weeks per treatment cycleForced titration; lower to higher dose at 6 weeks

No washout period between cycles

Home Systolic BP measured at

6 and 12 weeks

Plasma ReninAldosteroneAldosterone/

Renin ratio

Haemodynamicstudies

Haemodynamicstudies

Haemodynamicstudies

Haemodynamicstudies

Haemodynamicstudies

Baseline

AmilorideOpen-Label

12 week Run-out10 -20mg o.d.

Clinic Systolic BP measured at 6 and 12 weeks

Relationship between renin and aldosterone levels in resistant hypertension

p=0.340 for the linear term p=0.0215 for the quadratic term*

Very few patients with low renin and low aldosterone

Many more patients with a relative increasein aldosterone despite a low renin

*Quadratic equation:aldosterone=2.365-0.0309*renin+0.0806*renin*renin

Impact of treatment of resistant hypertension on haemodynamics

*P<0.002

Measurements made at baseline and at the end of each treatment cycle - Cardiodynamics BioZ

Placebo Spironolactone Doxazosin Bisoprolol

*P<0.001

Stroke index Cardiac index

P<0.066 for overall

treatment differences

Vascular Resistance index

*P<0.002

Thoracic Fluid index

Effects of amiloride versus spironolactone on clinic systolic BP in resistant hypertension

P<0.001

Baseline Placebo Amiloride10 – 20mg

Spironolactone25 – 50mg

Doxazosin4 – 8mg

Bisoprolol5-10mg

Clin

ic B

loo

d P

ress

ure

(m

mH

g)

r =0.64 p<0.0001.

Correlation of BP reduction with amiloride vs spironolactone

Change in clinic systolic BP from baseline on spironolactone

Ch

ange

in c

linic

sys

tolic

BP

fro

m b

asel

ine

on

am

ilori

de

Nomura et al. Circ Res. 2017;121:81-88 Reveal Study: DOI: 10.1056/NEJMoa1706444

Mendelian randomisation predicts morbidity-mortality outcomes

ObjectiveTo determine whether CV morbidity is reduced by better BP control or by choice of diuretic (K+-losing, sparing or neutral)

EligibilityAged > 60, and home SBP > 130 mmHg, and risk factor/markerSources of patientsRegistries for acute myocardial ischaemia, PCI, arrhythmiaPoints of recruitmentDuring acute admission, or via GP Research Database

Nutshell summary of PATHWAY-4

Is Metomidate PET CT superior to Adrenal venous sampling in predicting ouTCome from adrenalectomy in patients with primary Hyperaldosteronism (MATCH): a multi -centre, randomised, within-patient comparison of diagnostic techniques

PATHWAY-2 (Resistant Hypertensive) patients cured by diagnosis/treatment of Conns

B.S. d.o.b. 5/4/1965

Oct 2013

Laparoscopic adrenalectomy

6 mm adenoma

Feb 2014

BP 121/88 mmHg

Untreated

Renin 27 mU/L, aldosterone 143 pmol/L

Aldosterone

synthase

(CYP11B2)

‘Normal’ CT or MRI scan of adrenals does not exclude Conn’s adenoma

CT Overlay

Renin <2.0 mU/L

Aldosterone 522 pmol/L

BP 189/114 mmHg

Rx: sotalol, irbesartan, doxazosin

Topics

•PATHWAYs• Recent (summary of P1, P2, P3)

• Present (mechanisms sub-studies, mainly P2)

• Future (GWAS, Mendelian Randomisation, P4?)

•Studies in Primary Aldosteronism• 11C-metomidate PET CT vs Adrenal Vein Sampling (‘MATCH’)

• Endoscopic radiofrequency ablation of adrenal adenomas(‘FABULAS’)