Hypertension

1. Is it effective?YesThere are two studies that looked at the treatment of chronic hypertension in the elderly. One was a trial,the other a meta-analysis. These trials were not about the acute treatment of malignant hypertension butfocused on lowering systolic blood pressure from about 160 to 150.Beckett NS, et al. N Engl J Med. 2008 May 1;358(18): 1887-98.Gueyffier F et al. Antihypertensive drugs in very old people: a subgroup meta-analysis of randomisedcontrolled trials. INDANA Group. Lancet. 1999 Mar 6;353(9155):793-6.

2. Is it effective for a clinical endpoint the patient cares about?Treating hypertension in the elderly mainly decreases strokes and heart failure exacerbations but notheart attacks.There was a mortality benefit in the only randomized trial of those older than 80. It is unclear if this is atrue benefit.

3. What is the time frame for efficacy?The time frame for efficacy is likely as short as 1.8 years. Possibly shorter.

(efficacy increases over time)

2 years30%

Death 141 15 21%Stroke 194 13.3 2 yearsCHF 153 11.5 2 years 64%

CV Event 130 13 2 years 34%

Death 116 15 5 years 21%Stroke 148 13.3 5 years 30%CHF 121 11.5 5 years 64%CVEvent 112 13 5 years 34%(CV event is a stroke, MI or CHF event-combined end point)Beckett NS, et al. N Engl J Med. 2008 May 1;358(18): 1887 -98.Treating a 1000 patients over 2 years would prevent 24 deaths, 10 strokes, 19 heart failure events and33 CV events. 96 deaths, 23 strokes, 28 heart failure events and 66 CV events will happen anyway.Treating a 1000 patients over 5 years would prevent 62 deaths, 21 strokes, 48 heart failure events, 83CV events overall.

Hypertension has never been studied for secondary prevention of CAD events. NNT would likely belower but that is speculation

Secondary stroke 168prevention

3 4 years 33%

Rashid P, et al. Stroke. 2003;34:2741-2748. (not in elderly)

Death from any cause36

Strokes

CHF

j!li~.;;

.111.;Q.QQ...•q;c..1/11:~

Ui11.~

1/1 7.•..s::¢I-,;::

6&Q :5Q....•.•..

:1P 4e,III•..c :1lP>-

W1.0.0 200Z 1

O·D

1/1i::liQ.;oQ.-ti!!lc~w

"0

oz

2Q P""O.02

10

C)o 2 3.

Follow-up (yr)1

1'=0.06

1 2

Follow-up (yrJ

7

P<O,QOl

6

s.4

3·

2""1

h!

opt', .. . 11 I

o 1 2 3follow-up (yr)

Hypertension

Plsoebegroup

Active·treatment

~rou:p

4

Placebogroup

Active·treatment J!!.

group ;2tI)

..t:';

~--t::<I>e&:

Placebogroup

..•..•t:::(I)

f:A '. (I).'ctive- Q.

treatmentgroLjP

·4

Death from any cause100

Treatment

75

- .•..•..t::(I)e(I)Q.

5;0

25 , _-_ __ _.-

'0' .;;.c.~o 32

Years

stroke100

!ill: Placebo Bi"I"~HTlem

.......

'nY'

1'5

50

25

o'3

Years

CHF100

III!Treatment

75 I·.. . ; .....................................•.............

50 ,- ..... _ ..... --_ .•--- .. _ ....

25 ,.

o ~I -~~~~;;;;;=::::2 3

Years

Hyperli pidemia

1. Is it effective?YesAlthough several major cholesterol trials included older adults (4S, CARE, LIPID, Heart Protection Study,TNT), there is only one placebo controlled study targeting adults older than 70-the PROSPER study.Sub group analysis of the LIPID study suggests that the relative risk reduction in events is the same forolder adults and because older adults have higher absolute risk for CV events, the absolute risk

reduction is therefore greater in older adults.

2. Is it effective for a clinical endpoint the patient cares about?Treating hyperlipidemia reduces heart attacks but not strokes or total mortalityWhile treating high cholesterol in younger adults prevents deaths, strokes and heart attacks, in thePROSPER study, the treatment of hyperlipidemia produced fewer fatal and non fatal myocardialinfarctions but not all cause mortality or strokes.

3. What is the time frame for efficacy?6 months to 2 years for any efficacy.In the WOSCOPS study, the authors suggested efficacy was seen as early as ,6 months. However inboth Heart Protection Study that looked specifically at this question, statistically significant separation didnot occur until 2 years.

4. What is the likelihood of achieving benefit?

.©utt'omeDeath (1)

Not effective

Not effective

6

Primary MI prevention 1111 10

3.2 years 19%

CVA (1)

MI (fatal/non) (1) 148

3.2 years 10%

Secondary MI prevention* (2) 125 4 3.2 years 25%

Secondary CVA prevention* (3) 152 6.7 4.9 years 15%*Not in elderly> 75.1. Shepherd J, et al. Lancet. 2002 Nov 23;360(9346):1623-30.2. Sacks FM, et al. Circulation. 2000; 102: 1893-1900.3. Amarenco P, et al. N Engl J Med. 2006 Aug 10;355(6):549-59.

Treating 1000 adults over 3.2 years would prevent 21 MI's.

For primary prevention, treating 1000 adults over 3.2 years would prevent 9 MI's. 90 will have an MIanyway.For secondary prevention, treating 1000 adults over 3.2 years would prevent 40 MI's. 120 will have anMI anyway.

Hyperlipidemia

To prevent Ml's

B~ 15~'"~ 10

£>'30c 5

o ,? I8. I [ ,o O~- i ,6: j

p=O·,Q06/'~-.'

. ~j/"."", .,.-..._",,J<!""--'

To prevent cv»:«

c~ 7.... 6e§?q; 5E 4~c 3.Q 21 p"'O,811:::8. 1e 00...

!0

st:::>o/.£::

~,/ !",,' r::.>" Q)~cr.

1 .. "j t ... ~ f 1 I

1 1·5 2 2·5 3 3·5 4follow,up (years)

Hyperlipidemia and MI's100 ,...-_ - ~ .. ~.-- -..- ~--- ---- - - ..

/ ~ Placebc It' Treatment

~

~

75 I·---··~-_ -<- --"-~'- ..----;. .. '·-0" •• - •••••• -. ,.-.-- •••• -- •• -

50 I ................................ ~..•....... ~ ..•...•t:::Q)

~~ 25 I----··-·~···· .-,~•.<- .-..-~.--.,...............•.---~- ...~~.-.------ ...- ~..

~~<:H ~.....•.

o 32

Years

Prevention of CVA100 ,

.• ' Placebo ~ Treatment

'-''.i(."

4

50

25 I ..··..........

o I ~0 1 2 " 4"

Years

75

Shepherd J, et al. Lancet. 2002 Nov 23;360(9346):1623-30.

-,

Diabetes

1. Is it effective?It is unknown since there are no trials of glucose control in the elderly. The UKPDS study was done innewly diagnosed diabetics. The ACCORD study had an average age of 62 (40-79). Because ofmethodology issues, it is very difficult to know how this applies to a geriatric population. Would tightcontrol be more effective in patients with longer standing diabetes who are at higher risk ofmacrovascular events? Or less?

2. Is it effective for a clinical endpoint the patient cares about?Tight control reduces microvascular endpoints (retinopathy, neuropathy and nephropathy) but notmacrovascular endpoints (stroke, MI).In the UKPDS study, tight control of diabetes (A 1C of 7 vs 8) reduced microvascular outcomes-mainlyretinopathy-but not macrovascular outcomes (MI, CVA). The ACCORD study looked at an A1C of <6and found it to be harmful. For retinopathy, the outcome was not blindness but the need forphotocoagulation.No benefit for mortality, stroke or myocardial infarction. Myocardial infarction was almost statisticallysignificant in UKPDS at p=.052.

3. What is the time frame for efficacy?Unfortunately the studies do not attempt to answer this question. Looking at the graphs, it seems clearthat there is no separation of the curve by 9 years. And apparently there is by 10 years. So somewherebetween 9-10 years.

4. What is the likelihood of achieving benefit?

Microvascularevents

4 10 years

13%

35

10 yearsMacrovascularevents*

46 7.5

25%

*P value = .052UKPDS Group. Lancet. 1998 Sep 12:352(9131):837-53.

Treating 1000 patients over 10 years would prevent 28 cases of retinopathy. 84 patients will developretinopathy anyway. After 10 years of tighter control, retinopathy would be delayed about 1.5 years.

5. Is it worth the risk of side effects?The risks of tight control and hypoglycemia are numerous including death, dementia, falls,institutionalization, hospitalization.

6G

~40

";II;.

Any dbrl>etes;.related endpoInt

2G

o

Diabetes

/.. C_on~-c~tional'/' IntenSitre

--t::~ 75<I>t::!U

;S 50.~-c::<I> 25~

r§:0

0

oI ill 1--'

:3 6' 9 12 :lS

nine from fl!!1!b:Hnh.athmJ jI<!!al$

CQcnventirlil.a!at ri!OO, 10l.0 $47 524 204 47Intl!nsWe <ft ri$\<.: 2447 2(1.67 U06 sse 110

3D'"! Microvascular •••.•od,polnts

..5et

20

l'

oj --r-1}

r6

Time from randomization, yeat"S

T9

;-----;12; U

100

"'-'eQ)::s.Q).c:~$.•...I;::Q)

E:Q)t:l

Int·en'ShlC

1:!iu:S.~1:<I>e~

Any Diabetes Endpoint100

2.5 5Years

7.5 10

Microvascular Endpoints100

75

50

25

oo 2.5 10

.~

5Years

7.5

Myocard ial Infarctlon"

Treatment

75

50

25

5 7.5 10o

Years*P value=.052

UKPDS Group. Lancet. 1998 Sep 12;352(9131):837-53.

o 2.5

1. Is it effective?Yes

Systolic Congestive Heart Failure

2. Is it effective for a clinical endpoint that the patient cares about? (goals of care)Treatment of heart failure both reduces death and hospitalizations for CHF exacerbations.

3. What is the time frame for efficacy (over long haul/short term)Trials show benefit as early as 3 month duration of treatment for a variety of interventions. Digoxin'sbenefit almost disappears when hospitalizations that are caused by digoxin toxicity are combined withthe benefits from reduced heart failure hospitalizations.

4. What is the likelihood of achieving benefit (i.e. NNT)?Betablockers

Death 138 13.1 7 months 32%CHF hospitalization 124 12.4

CHF or Death 115 12.77 months 41%7 months 37%

Lechat P, et al. Circulation 1998;98: 1184-91.Treating 1000 patients for 7 months would result in 26 fewer deaths, 42 CHF hospitalizations, 67combined events. 55 patients would die anyway, 58.8 would have a CHF hospitalization and 114combined events.

100

BB in CHF100

75

CHF and BB

r::~(II

~JS'a

........... ; .......................•.......•......... L. ~

'-esll;Cl

1---: --_ , __..

~'--.~~

50

25

75

50

25

e5234

Months

o5 234

Months

ACE inhibitors

Death 3.6 3-6 months17 23%

Death or 110 13.2 13-6 months 135%hospitalizationGarg R, et al. JAMA 1995;274:462.Treating 1000 patients for 3-6 months would result in 59 fewer deaths, 100 fewer deaths orhospitalizations. Despite treatment, 160 patients would die and 226 people would die or have ahospitalization.

ACE-i and CH F

~

I

Aldactone

100 ACE-i and CHF

t:.£Iii.~lii

~~....o:SIIIQ)

Q

100 !~ Placobo' <It Treatment

75

50

25

o2 3

Months4 5

25

o

2 years

2 years

Pitt B, et al. N Engl J Med 1999;341:709-17.Treating 1000 patient would result in 111 fewer deaths and 250 fewer hospitalizations.treatment, 333 patients would die and 575 would be hospitalized.

Spironolactone in CHFIIPlacebo ~ Treatment

Death 19 14

rtl O.BO>'s;; 0.75

V>

"0~ 0.70

~ 0.65.Do.t 0.60

75

50

Months

30%

35%Hospitalizations 14 13.3

1.00

0.55

0.95100

~ 75~-..m 50::.;;..~~

CI) 25

0a

0.90

0.85

0.50

0.45);

0.001 , , I I I I I I I I I I

o 3 6 9 12 15 18 21 24 27 30 33 36

Months

Despite

9 18Months

27 36

Hydralazine and nitrates. "';~>&_"" ."' ""'.... </"_. .,.:.,.;..:1...... .-, _'.0"0 -_.

Death (1) 114 3 1 year 38%23%Death 110 4.3 3 years

43%Death (2) 125 3 10 months

1st hospitalization(2) 113 10 months 133%3For (2), efficacy started at 6 months.Treating for 1000 patients over 1 year would prevent 71 deaths and 142 patients would die regardless.1.Cohn IN, et al. N Engl J Med. 1986 Jun 12;314(24):1547-52.2. Taylor AL et al. Combination of isosorbide dinitrate and hydralazine in blacks with heart failure.NEJM. 2004 Nov 11;351 (20):2049-57.

0.7

o--c Placebo (~73)

~Pralosin (183)

~Hy(Hso (186)0.6

.,a; 0.5a:>.

2! 0.4(;::;~ 0.3a;:;5 0.2o

o

0.1

6 12 18 24 30Interval (months)

36 42

Digoxin

J!:l«l0:::

.?:-'.-~~~~~:::sE:::so

Hydral/Nitrate in CHF100

Treatment

75

3 4

50

25

oo 2

Years

Hospitalization for 123 113 137months 113%CVevents

Any hospitalization 136 124.2 137months 18%The Digitalis Investigation Group. N Engl J Med. 1997 Feb 20;336(8):525-33.

50CHF and Digoxin

100 ,...., .

75 I--.- -: - + --.- ..~ - ..--:..-.--L.

25 f--_ .., -_.-!- ..---+:o

Ol~~~10 20 30

Months

oi*" 40o~ ~H so ----------- ~

U Precebc _/" .> ~~-----' ;s'" ,.' ~,_ .1\1o 0> 20 .>: ~---- N lJ...'" ---- '"00 -: .:r:" -: /' '" u 50'.• 0 -- /' "-,> io / -: ~ -E

/ / "'00»0 ;:z::/ :l§~~ ---

o 'l""""""'" m8 12 16 io"""" , a24 28' '312' , Id I , • I i I I , i

Months 6 40 44" de' i \'240 50

Osteoporosis

1. Is it effective?Yes

2. Is it effective for a clinical endpoint that the patient cares about?Osteoporosis care is focused on prevention of fractures. It also has a small reduction in mortality.Mark J. Bolland et al. Effect of Osteoporosis Treatment on Mortality. J Clin Endocrinol Metab. 2010

3. What is the time frame for efficacy?In trials, efficacy is seen somewhere between 1-4 years. Possibly sooner.Maclean C, et al. Ann Intern Med. 2008 Feb 5;148(3):197-213.

4. What is the likelihood of achieving benefit?Vertebral Fractures

1-4 years

2 1-4 yearsRisedronate 16

Alendronate 15

Ibandronate 37

Zoledronic Acid 47-59

Raloxifene 16

Teriparatide 5-6

2.6

2.5 1-4 years

2.2 1-4 years

3.3 1-4 years

1.5 1-4 years

Non-Vertebral Fractures~,~:~i,~~r~6,~[:. i 0- • "'t~:sjEV~nts,<:

Ibandronate Not effective

3 1-4 years

Risedronate 134 4.5 1-4 years

Alendronate 136

Zoledronic Acid 37 4 1-4 years

Raloxifene Not effective

Teriparatide 29 3 1-4 years

Hip Fractures~~~od]6:atton'·::'::':-':'i:';;"(:'{--,:L<i''''::;·:,;~·'-.r.'i:',~

Risedronate/ O"U(atibn

"_·i::-1.,°· .

Ibandronate

96

~81 2.7

Not studied

91*~

Not effective

Not studied

1-4 years

1-4 years

Alendronate

Zoledronic Acid 1-4 years

Raloxifene

Teriparatide