hypertension: a personalized approach to meds

 hypertension: a personalized approach to meds



You forwarded this message on Mon 5/1/2023 6:33 AM

You forwarded this message on Mon 5/1/2023 6:33 AM

MG

Modest, Geoffrey (HMFP - Medicine)

To:

• 
  Modest, Geoffrey (HMFP - Medicine)

Mon 5/1/2023 6:32 AM

 

A very interesting Swedish study just came out suggesting that we should approach hypertension treatment in an individualized way (see htn variability of response to meds JAMA2022 in dropbox, or doi:10.1001/jama.2023.3322) 

 

Details: 

-- 270 participants in an outpatient research clinic in Uppsala Hospital who were 40-75yo and had systolic blood pressure (SBP) between 140 and 159 mmHg and diastolic < 109 mmHg within five years before the trial, were randomized in a double-blind repeated crossover trial. All were at low cardiovascular risk  

-- 46% female, median age 64, BMI 29, BP at randomization 154/89, ambulatory daytime BP 145/89

-- median of 3 years since hypertension diagnosis, 62% had been on a BP medication  

-- smoking: 6% at least occasional/35% previous/60% never smoked; smokeless tobacco: 6% 2-4 tins per wk/5% up to 2 tins per wk/87% never used; alcohol: 7% never used/ 14% once a month/ 42% 2-4 times per month/31% 2-3 times a week/7% >3x per week; physical activity: 14% regular training/35% regular leisure time physical exercise/46% moderate leisure time physical activity/6% sedentary 

-- women with preeclampsia 11%, men with  erectile problems 11%, cancer in 7%, sleep apnea in 6% 

-- exclusion criteria: possible secondary hypertension, other serious disease, gout, cardiovascular disease, kidney failure, and diabetes 

-- all were randomly assigned to one of four blood pressure medications: lisinopril 20 mg; candesartan 16 mg;  amlodipine 10 mg; and hydrochlorothiazide (HCTZ) 25 mg (they considered these to be equipotent) 

-- in addition, every participant repeated two of the treatment periods, selected at random (as a means to test the intraindividual variability between two trials of the same medications) 

-- each treatment period was 7-9 weeks, with half doses of medications for the first two weeks, then full doses, and with a one week washout between the medication cycles 

-- ambulatory daytime systolic pressure was measured at the end of each treatment. Measurements were recorded every 20 minutes during the day and every hour during the night, with successful registrations being at least 22 hours in duration with at least 2 measurements per hour and 14 measurements between 10 AM and 8 PM 

 

-- Main outcomes and measures: ambulatory daytime systolic blood pressure measured at the end of each treatment period, based on daytime ambulatory SBP 

    -- they also assessed the theoretical maximum mean gain from personalization of  the antihypertensive meds in the trial population 

 

Results:  

-- there were 1468 completed treatment periods for the 270 randomized participants, with the median length of 56 days 

-- the achieved SBPs on the meds: HCTZ 136.1 mmHg, amlodipine 130.9 mmHg, candesartan 131.8 mmHg and lisinopril 129.7 mmHg [ie, the meds were not equipotent on prolonged daytime ambulatory screening. See comments below] 

 

-- blood pressure response to different treatments varied considerably between individuals, p<0.001 

    -- observed crossover differences in SBP, between pairs of medications (ie taking the same meds twice): 

        -- hydrochlorothiazide versus lisinopril: 6.6 mmHg (lisinopril much better than HCTZ) 

        -- hydrochlorothiazide versus amlodipine: 5.6 mmHg (amlodipine much better than HCTZ) 

        -- hydrochlorothiazide versus candesartan: 4.4 mmHg (candesartan much better than HCTZ) 

        -- lisinopril versus candesartan: -2.0 mmHg (lisinopril somewhat better than candesartan) 

        -- amlodipine versus candesartan: -1.2 mmHg (amlodipine a little  better than candesartan) 

        -- amlodipine versus lisinopril: 0.9 mmHg (lisinopril a little bit better than amlodipine) 

-- blood pressure stability evaluating a treatment comparison that was repeated in individuals:  

     -- there was a quite small difference, largely between -0.08 to 1.0 mmHg between the first assessment and the second assessment of the same meds in the same individuals (the lisinopril-candesartan was an outlier at 2.1 mmHg) 

    -- on average, there was up to an estimated 4.4 mmHg difference if one were to choose a single drug that best fitted the patient (3.1 mmHg if compared to lisinopril, since it was the most efficacious) 

--amlodipine had the lowest within-participant SBP variation than other treatments 

  

Commentary:

-- as a background:  all classes of antihypertensive medications decrease the risk of stroke and coronary artery disease

-- This was a study of healthy people with low levels of hypertension, finding: 

    -- there were significant differences overall in response to the four different medications used, finding varied effectiveness between individuals 

    -- by assessing individuals and their response to a repeated study with the same medications and finding very similar results, it was clear that antihypertensive medication response was consistent for individuals (independent of the meds chosen). this finding opened the real possibility that we might be able to choose optimal medications differently for different individual patients 

-- the magnitude of the differences between individuals on average was similar to doubling the dose of the first blood pressure med, and more than half that of adding a second drug

    --there were a couple of studies of patients who were above BP goal on a single agent, and adding another low dose agent is far more effective: doubling the dose provided only 20% of the effect o adding another drug (see http://gmodestmedblogs.blogspot.com/2018/03/adding-yet-another-bp-med-helps-alot.html or for a geriatric assessment http://gmodestmedblogs.blogspot.com/2021/11/hypertension-add-another-med-vs.html 

-- they also did find that there were some medication differences that were quite large and more likely to provide a benefit from changing medications: especially for candesartan versus amlodipine or lisinopril versus amlodipine. Hydrochlorothiazide fared poorly overall 

-- several studies have assessed clinical outcomes, as well as the blood pressure readings, of different antihypertensive classes, with a few important differences between the medication classes:

 

    -- hydrochlorothiazide failed miserably overall in this study: the daytime blood pressure (the one we typically assess in the clinic) is often excellent (since patients typically take the med in the am), but the true 24-hour blood pressure reading in those on HCTZ 12.5-25mg/d is terrible (see htn HCTZ Meta-analysis messerli am j cardiol 2011 in dropbox, or doi:10.1016/j.jacc.2010.07.053), showing only 50% of the antihypertensive response vs the other groups:

 

 

     -- there was a comparative study between hydrochlorothiazide and chlorthalidone (which does have 24-hour blood pressure coverage) finding similar clinical outcomes, though to my reading this was a pretty terrible study: http://gmodestmedblogs.blogspot.com/2020/03/chlorthalidone-vs-hctz.html. this blog refers to a few other relevant blogs

          -- http://gmodestmedblogs.blogspot.com/2016/04/chlorthalidone-is-better-than-hctz-for.html a study showing that chlorthalidone 6.25 mg per day has about twice the blood pressure lowering as compared to hydrochlorothiazide 12.5 mg. This blog also references other studies noting benefit of chlorthalidone. One major point found was that clinic-based blood pressures tend to look good on hydrochlorothiazide 12.5mg, so the patient and clinician are happy with the med, but the 24-hour efficacy (the evaluation with the best correlation with future cardiovascular events) is far inferior to that of chlorthalidone. 

            --  http://gmodestmedblogs.blogspot.com/2017/11/new-aha-hypertension-guidelines.html has the 2017 AHA hypertension guidelines, which prefer chlorthalidone as the thiazide-like diuretic, and that if using hydrochlorothiazide, it should be at least in the 25-50 mg dose (the 50mg dose does have longer duration of action). 

            -- http://gmodestmedblogs.blogspot.com/2018/04/ambulatory-blood-pressure-monitoring.html documents that the 24-hour ambulatory blood pressure is the one most predictive of future cardiovasc events 

            -- NICE (the National Institute for Health and Care Excellence of the UK) in 2011 strongly came out against using hydrochlorothiazide as an agent at all, promoting amlodipine or chlorthalidone as the best choices for most patients       

            -- though there was a recent (somewhat lame) VA study finding no difference between HCTZ and chlorthalidone: http://gmodestmedblogs.blogspot.com/2023/01/hypertension-hctz-vs-chlorthalidone.html         

-- And, the Swedish study above found that HCTZ was not equipotent with the other meds. Since so many studies before this one found that the clinic-based BP was okay on HCTZ, this suggests (to me) that HCTZ just does not have a long enough half-life, so the clinic reading in the afternoon might be great, but doing the longer recording from 10am til 8pm as in this study, might reflect the later decline 

 

--  ACE-inhibitors raise several concerns: 

    -- the ALLHAT study found: 

        -- diuretics (using chlorthalidone) were actually better than the amlodipine and lisinopril; doxazocin was used initially but dropped because of poor early outcomes.

        -- lisinopril was associated with 15% increased risk of stroke, but no difference between chlorthalidone and amlodipine (also there was the surprising finding that lisinopril was associated with more heart failure)  However, those on lisinopril did have a higher sustained blood pressure, but this was not felt to explain the increased stroke incidence (see https://jamanetwork.com/journals/jama/fullarticle/195626) 

    -- a Taiwanese real-world population study found a 26% lower incidence of stroke in those on an ARB vs ACE-I: https://www.sciencedirect.com/science/article/abs/pii/S016752731630794X 

    -- one concern about lisinopril (which is in the tier of those ACE-I's with longest plasma ½-life) is that the increased risk of stroke could be from its waning effectiveness in the early morning hours, when both the early AM cortisol surge/blood pressure increases and the peak incidence of strokes and cardiovascular disease in general occurs (see htn BP inc early am chronotherapy EuropReview1999 in dropbox, or  https://pubmed.ncbi.nlm.nih.gov/11261739/ ). This study, using ambulatory 24-hour BP monitoring (ABPM), found that optimal timing for taking lisinopril  to cover these times is at 10PM. [ the normal BP pattern is a slow increase in BP from 3am-7am, then more rapid increase until the peak at 10am]

        -- patients were randomized to take the drug at 8.00 AM, 4.00 PM or 10.00 PM, and they had repeated ABPM every two months, finding: 

            -- 8am dosing: baseline BP was 144/99 mmHg, decreasing to 135/90 from 6am-11am

            -- 4pm dosing: baseline BP 144/99 mmHg, decreasing to 133/91 mmHg from 6am-11am

            -- 10pm dosing: baseline BP 144/99 mmHg, decreasing to 122/80 mmHg from 6am-11am

 

-- ARBs seem to have several potential benefits over ACE-inhibitors: 

        -- there may well be decreased risk of cognitive impairment: http://gmodestmedblogs.blogspot.com/2023/03/hypertension-another-boost-for-arb-over.html . this blog also cites data from the SPRINT trial that ARBs provide 24 hour coverage (with the exception of losartan, which only provides 24 hour coverage at a dose of 100 mg)

            -- though a side issue: unlike the other ARBs, losartan actually lowers serum uric acid levels (see this blog for more details on this). 

            -- there is an intriguing comment that ARBs and dihydropyridine CCBs (such as amlodipine) stimulate angiotensin II receptors 2 and 4 (versus ACE-I and non-dihydropyridine CCBs that inhibit these receptors), and prior observational studies have suggested that stimulating AT receptors 2 and 4 lowers dementia risk, as confirmed by many animal studies (eg see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7884979/ )     

                -- a real word study found a better safety profile for ARBs than ACE inhibitors, and they are better tolerated: https://bucommunitymed.wpengine.com/hypertension-use-arbs-over-ace-i/  

-- Amlodipine:

         -- in a meta-analysis of 21 studies on within-individual visit-to-visit SBP variability, higher variability was a powerful predictor of stroke, independent of the achieved mean systolic blood pressure. 

         -- a subsequent study confirmed this, finding that calcium channel blockers (amlodipine being the most commonly studied) had the least blood pressure variability (as was seen in the current Swedish study above), and that ARBs, ACE inhibitors, and beta blockers had the most variability (diuretics being in-between). And, there was a significantly decreased risk of stroke in those on calcium blockers. ARBs were the only drug class that did not show this pattern for stroke risk that was expected by blood pressure variability (see  htn variability and stroke lancet 2010 in dropbox, orLancet 2010; 375: 906–15). For a more recent study on blood pressure variability, see htn bp variability amlodipine htn2017 in dropbox, or DOI: 10.1161/HYPERTENSIONAHA.117.10087) 

         -- also, amlodipine and nifedipine also lower uric acid levels; HCTZ increases them (as well as clinical gout)

Limitations:  

-- though they actually measured BP at night, there results were not reported. Would have been useful to see their night-time data, especially for HCTZ and lisinopril. And we do not know when the patients in the study took their assigned meds

-- this was a healthy group of patients with mild hypertension, so results may not be generalizable to less healthy people or those with more significant hypertension 

-- the researchers chose the med strengths to be “equipotent”, which clearly was not the case as noted above. They did control for this mathematically, but it would have been a cleaner study if the participants did achieve the same on-med blood pressure 

-- there is no longer-term analysis of clinical outcomes 

    -- for example, there may well be differing effects of the different meds on other responses than blood pressure (as suggested above for ARBs) that might affect clinical events: sort of like the profound pleiotropic effects of statins beyond their direct effects on lipids 

    -- and, lowering blood pressure is an important endpoint, but it is only a surrogate marker for the important endpoint of clinical outcomes

 

So, what does this all mean: 

-- there are overall different responses to the different medications used in this study, perhaps reflecting the fact that the doses chosen were not equipotent (and perhaps different doses might have performed differently)

-- there was a pretty consistent pattern of response when the same individual was on the same treatment twice, allowing for considering differential responses to different meds for the individual. this would support the possibility of personalized BP med prescribing

    -- perhaps effecting a personalized approach, as per this study, would be to consider changing from one medicine to another in the case where the blood pressure was within 5 mmHg of goal 

        -- and that assessment would best be by ambulatory blood pressure monitoring (though home-based monitoring is reasonable for many patients, but without clear evidence this is true) 

    -- if >5mm Hg, probably best to add another med (combo pills are often available and have 1 less co-pay for the patient). And a second agent should be considered in those who were switched around on different single agents without hitting the target BP.

-- given what was mentioned above about the meds, I still feel that amlodipine is the single best BP med unless one is treating 2 conditions in a patient, such as blood pressure and also heart failure or diabetes with microalbuminuria, etc. 

    -- chlorthalidone is also a reasonable first line therapy, though hypokalemia is so common 

    -- ARBs are a second choice to me, given the above arguments and the data showing all but losartan have a full 24-hour effectiveness (losartan does at the 100mg dose) 

    -- lisinopril, in the top tier of the ACE inhibitors because of its longer life, to me is a lower choice because of the potential adverse effects on cognition as well as the decreased effectiveness and potential of increased strokes in the early AM 

    -- and HCTZ should probably not be used as a single agent (though seems to be okay for 24 hours in combo with an ACE-I or ARB, likely from creating a higher renin state). the good news here, i think, is that for decades HCTZ was the most prescribed BP med in the US. now it is number 3.  Lisinopril is number 1...... (then amlodipine)

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

  

to get access to all of the blogs (2 options):

1. go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

  -- click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

  -- or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them

2. go to https://bucommunitymed.wpengine.com/, a website from the Community Medicine section at Boston Medical Center.  This site does have a very searchable and accessible list of my blogs and is the easiest to view blogs and displays more at a time.

please feel free to circulate this to others. also, if you send me their emails (gmodest@bidmc.harvard.edu), i can add them to the list