ESC Debate | Should hypertension be redefined?ESC 2018 - debates
Chairpersons: Giuseppe Mancia (Milano, Italy) and Alta Schutte (Potchefstroom, South Africa)
Positioning of the topic - Defining hypertension
Alexandra Konradi (St. Petersburg, Russian Federation)
Hypertension is the number one driver of cardiovascular (CV) mortality. It is a big problem, but the easiest disease to diagnose. Moreover, Konradi noted, it is easy to treat, as there are plenty of efficient and safe drugs available. Nevertheless, millions of affected subjects remain underdiagnosed, and the majority are undertreated. Poor compliance of patients and physicians’ inertia are highly prevalent.
The definition of hypertension brings about discussion. How hypertension is defined is important for risk stratification, because it determines indications for treatment, because it affects prevention strategies and it has financial consequences, in light of reimbursement, lifestyle interventions, evaluation and frequency of doctor visits.
A challenge lies in that elevated blood pressure (BP) is easy to measure, but difficult to justify. Three major questions arise: who should be considered ill, whom to treat and how intensive to treat. The main concern for doctors is what level of BP needs to be pharmacologically treated. The pragmatic definition of hypertension is the BP level at which treatment is worthwhile, but this varies from patient to patient, and benefits of treatments and risks of not treating should be weighed against disadvantages of medication.
When trying to define hypertension, the fact that we do not define the disease by symptoms, diagnostic criteria, laboratory or instrumental findings, complicates matters. It is just based on BP. But measurement of BP is not a stable, objective procedure. Various approaches exist, namely office measurement (HBPM), home measurement and 24-hour ambulant BP monitoring (ABPM). All approaches have advantages and disadvantages.
It is important to realize that BP is variable in many ways: from beat to beat, from measurement to measurement, from day to day, from day to night, from visit to visit, from nurse to nurse, from nurse to doctor, from doctor to doctor, from winter to summer and from year to year. Thus, even within one patient defining hypertension can be difficult. We therefore need to confirm sustained hypertension, since high BP is not equivalent to hypertension, and normal BP does not equal normotension. In general it can be said that the more measurements are done, the better is diagnostic accuracy.
The 2018 ESC Guidelines for management of arterial hypertension presented at this year’s ESC congress, have unchanged classification of BP and definition of hypertension, as compared with previous European guidelines, meaning that hypertension is defined as office systolic BP (SBP) ≥140 mmHg and/or diastolic BP (DBP) ≥90 mmHg, which corresponds to a 24-hour ABPM average of ≥130/80 mmHg or HBPM average ≥135/85 mmHg.
Considering all these difficulties and sources of variability, Konradi postulated that maybe we need individualized BP thresholds, which take life expectancy, age and fragility into account. But first: the debate on whether BP >130/80 mmHg is hypertension: yes or no.
Everybody with a blood pressure above 130/80 is hypertensive – PRO
Prof. Paul K WHELTON (New Orleans, LA, USA)
Prof. Whelton chaired the ACC/AHA BP Guideline writing committee, thus he set out to defend the American viewpoint on how to define and diagnose hypertension. First, he emphasized that when comparing guidelines, we tend to focus on the differences. He prefers to look at commonalities. Yes; the ESH/ESC and ACC/AHA guidelines show differences, but they also have many common elements. For instance, they both emphasize the importance of accuracy in BP measurement, which he referred to as the elephant in the room. We have tolerated less accuracy for long, and this is a problem that needs solving.
Typically, the term hypertension is used at a BP level that has CV consequences, and at which treatment trials warrant antihypertensive drug therapy. As opposed to earlier US guidelines, it was now thought relevant to define hypertension also with a middle category: stage 1 hypertension, with SBP 130-139 mmHg. A distinction is made between stage 1 and stage 2 (SBP ≥140 mmHg), because treatment recommendations differ. These choices are based on observations in several studies that CV risk is continuous with increasing BP, without evidence of a threshold.
So, if hypertension is identified according to this definition, what should be done? Quite a number of approaches have been described to lower BP. The US guidelines recommend six lifestyle interventions with particularly good evidence, including weight loss, healthy diet (DASH diet), reduced dietary sodium, enhanced dietary potassium, physical activity and moderation in alcohol intake. With regard to antihypertensive drug therapy, a meta-analysis demonstrated that if a lower BP is achieved, the benefit for CVD reduction is larger. This progressive reduction in CVD risk is seen down to levels below current US and European recommendations.
Moreover, he shared analyses of the SPRINT-MIND investigators, who evaluated mild cognitive impairment in relation to standard or intensive antihypertensive treatment. A reduction of adjudicated SPRINT-MIND events was seen with the intensive strategy. This good signal for small vascular disease strengthened the beliefs of the guideline committee when they wrote their recommendations.
Nevertheless, not all stage 1 hypertension needs drug therapy, or would benefit from it, Whelton demonstrated with a flowchart from the 2017 ACC/AHA Guidelines. It is especially important to understand the underlying CV risk in stage 1: only if a patients has clinical CVD or an estimated 10-year ASCVD risk ≥10%, then BP-lowering medication is indicated. These individuals with CV risk are where the benefit lies. It should be noted that at any given level of BP, there is enormous variation in the risk of CV events. With higher BP, the risk of ASCVD increases. The increase in risk is, however, small in adults at low ASCVD, but much greater in those at high ASCVD risk. The relative risk reduction with drug treatment in RCTs is relatively stable across BP levels, while the reduction in absolute risk is much greater in those at higher risk of ASCVD.
Thus, in addition to stressing the importance of accurate BP measurement for diagnosis and management of high BP, the ACC/AHA Guidelines recommend lifestyle modification combined with antihypertensive medication in those with stage 1 hypertension if they have high ASCVD risk, and in those with stage 2 hypertension.
Everybody with a blood pressure above 130/80 is hypertensive – CON.
Prof. Bryan WILLIAMS (London, United Kingdom)
Prof. Williams led the development of the European Hypertension Guidelines. He acknowledges that risk is clearly continuous, so one could always find a level at which you start an intervention.
In defense of the European standpoint, Williams thinks that there are several issues with calling somebody hypertensive at the BP level considered stage 1 by the American colleagues. The first issue he brought up, was that insurance companies are very interested in classifying persons as hypertensive at BP >130/80 mmHg. Nevertheless, he also sees many similarities.
Terminology differs between guidelines, which should be fixed, said Williams. What is stage 1 hypertension in the US guidelines, is considered high-normal BP in all other international guidelines, and stage 2 according to the Americans, is grade 1 in all other international guidelines. Williams expressed his surprise about this, and notes contradicting recommendations in the US guidelines over time. In 2014, for instance, a treatment threshold of 140/90 mmHg would apply to persons aged <65 years old, and 150/90 mmHg was used for those of over 65 years old. Then in 2017, the BP threshold was lowered to 130/80 mmHg for all ages, which meant that most people of over 60 years old will get treated. Williams wondered out loud what the evidence is for this new, lower threshold, especially in those with 10% CVD risk? Or more clearly: what is the evidence that an intervention will help these people?
He defines hypertension as “the level of BP at which treatment results in more benefit than harm”. This is difficult to define. A threshold is chosen for pragmatic reasons. But looking at the evidence, we must conclude that there is no evidence that lifestyle intervention improves outcomes for persons with a high normal BP, nor that medication provides benefit for primary prevention in persons with a high normal BP. Thus, there is no evidence that individuals with high normal BP should be labeled as hypertensive.
Thus, he thinks that redefining hypertension at 130/80 mmHg leads to unnecessary disease labeling. This can divert workload from truly hypertensive patients at higher risk. If patients are at high risk at that level, it is not due to the BP! In those individuals, their risk is more likely to be due to factors other than BP that need attention. Therefore, it is probably more useful to give a statin and call them hypercholesterolemic. He illustrated his point with numbers: on November 11 2017, overnight 50% more persons in the United States turned hypertensive with the publication of the new guidelines. Nothing changed, however, in the diagnostic capability of medic America. Workload is an important factor in medicine. He noted that there are in fact a few people in the US with normal BP. The US Guidelines recommend reassessing these people after a year. ‘Why?’, Williams wondered with great surprise. In his country, such a recommendation for people with normal BP would overwhelm the health care system.
Was this influenced by the treatment goal? Because, if recommending a goal for all of BP <130/80 mmHg, then maybe we should also define hypertension as such. But, he repeated, the evidence that these people should be treated is lacking.
Another difference between the two guidelines lies in that Europe applies a different threshold for treatment in older patients (≥160 mmHg). 10% risk is not so high if one gets older. In Europe, in those at high normal BP, drug treatment is only advised in patients at high risk, particularly in case of coronary artery disease, where there is starting to be some evidence. When interpreting outcome analyses according to baseline BP, for instance in SPRINT OUTCOMES, it should be noted that most participants were already treated, thus the baseline is not a true baseline BP. In the HOPE-3, on the other hand, 80% of included patients were not treated for hypertension at baseline. In this post-hoc analysis, with a true baseline, the benefit of treatment was directly related to baseline BP. And only SBP >140 mmHg at baseline showed a significant treatment benefit. A large meta-analysis of RCTs also showed that in primary prevention, the association of BP-lowering treatment with MACE was dependent on baseline SBP, and that therapy is only associated with a lower risk for death and CVD if baseline SBP is ≥140 mmHg. At lower BP levels, treatment was not associated with benefit in primary prevention.
A last study he thought relevant to show, is the TROPHY study in prehypertensive individuals. Those in the placebo group were given lifestyle intervention, which, in trial setting is more effective than in clinical practice. Williams concludes from the TROPHY study that lifestyle intervention does not work. Finally, he shared insights from a more amusing study that surveyed people on how they viewed the burden of various types of intervention in relation to the calculated life-expectancy gain. The study showed that of the surveyed US adults, when imagining a scenario in which they had hypertension, 79% said they would take a pill for an extra month of life, 90% to gain an extra year, and 96% to gain 5 extra years. Similar numbers were seen when they were asked whether they would be willing to drink a cup of tea (the placebo scenario) for these gains. When their willingness to exercise was questioned, the respective proportions were just 63%, 84% and 93%. Thus, concluded Williams: ‘You can label people, but it won’t necessarily change their actions.’ In his opinion, it is not justified to label those at high normal as hypertensive, if we do not have anything to offer them.
In his rebuttal, Whelton noted that obviously, what is usual, is not necessarily normal. In certain societies without the exposure to bad influences like the US, people have perfectly normal BP. We however, live in an adverse environment: we consume much packaged food, are less physically active than is good for us, and we smoke and drink too much. We have, unfortunately, a sick population. One can be mad with the Guideline writers, but these are the facts. Epidemics are cultural in nature, and they are dealt with when the culture improves. That is not sufficiently the case for now.
Labelling is certainly an issue, Whelton agreed, if it is not connected to management, if there is no solution. But when we label and we give a solution, he is not aware of major adverse events.
Considering workload, that is certainly a realistic point. They tried to write guidelines that are pretty specific and easy to follow. But they are guidelines, and they may differ from clinical practice. They could have chosen a better risk prediction tool, but that would have complicated matters for doctors. In the US, most adults see their doctor about three times a year, so coming back for BP reassessment is not a big change, nor issue.
About what was seen in HOPE-3, he noted that the level of BP reduction was low, and he doubted that that is what we are dealing with in clinical practice. Furthermore, he noted that ‘We can pick the meta-analysis that we like.’ The US Guidelines committee had a separate meta-analysis review committee, different from the writing committee. The review committee said: the evidence that treatment is beneficial is there. Finally, about TROPHY he says that it was carried out pharmacotherapists. Was the lifestyle intervention really their focus? Whelton remembers that the two largest and longest lifestyle trials demonstrated that when you do it properly, you can reduce BP.
Williams was not taken aback by this stream of arguments. In his rebuttal, he noted that the thing that still stands out is that we do not have a single trial that has set out to treat patients at a certain threshold (with lifestyle intervention or drugs), to show that it is beneficial. He repeated the importance of separating the treatment target and the threshold for treatment. Epidemiological evidence suggests a lot; but there is simply no evidence that justifies medicating so many people. They agree that treating prehypertension may delay the onset of hypertension. But the sustainability of that effect of intervention has been a bit disappointing. Lowering elevated BP turns out to be hard to maintain.
Williams shared that in the European Guidelines committee, there was even discussion about whether or not treatment should automatically start in all with 140 mmHg, depending on risk! Thus, Europe has already shifted towards earlier treatment.
As a reply to Wheltons comment on picking his ‘favorite’ meta-analysis; he chose it because it was based on individual patient data. And he could have chosen another one with an identical conclusion. They appreciated the insights from that network meta-analysis, but in the guideline it is stated that these types of analyses are not used for recommendations, because it takes away the effect of randomization.
During a discussion after this heated debate, as a critique to the European viewpoint, the question was raised ‘How do you justify antihypertensive medication in patients you don’t call hypertensive?’ The answer is that it should be seen as part of risk lowering. According to the European Guidelines, treatment should be initiated at SBP >140 mmHg. The target to achieve then is <130 mmHg, which is a bit of a strange gap. This gap is the consequence of a gap in the evidence. The Guideline committee wanted to give a goal, because otherwise doctors and patients may stop at 139 mmHg, but the evidence says that going lower is better.