The CHAP Trial
/The CHAP Trial: Chronic Hypertension And Pregnancy
Formal Publication Title: Treatment for Mild Chronic Hypertension during Pregnancy
https://www.nejm.org/doi/full/10.1056/NEJMoa2201295
Some general background information
Who did the study and who published it?
The CHAP Consortium - a group of institutions in the USA, with protocol approved by the National Heart, Lung, and Blood Institute (NHLBI).
Recruiting took place across 61 institutions in the USA
Where was it published?
The New England Journal of Medicine in May 2022 - hot off the press!
Why was the study done?
Recall that after the CHIPS trial (we covered last week!), we still had some outstanding questions:
1) In the wake of CHIPS, there was renewed interest in the concern about antihypertensives and growth restriction.
Some smaller trials continued to substantiate this association; but a large meta-analysis referenced in the CHAP background found no association between SGA and cHTN treatment.
2) The CHIPS trial lumped together gHTN and cHTN – CHAP restricted care to true cHTN.
3) While CHIPS showed looser control of HTN didn’t result in major outcomes differences, there were some non-significant differences in rates of severe blood pressures and lab abnormalities.
With all of these things taken together, CHAP aimed to more narrowly answer the question of whether tight versus loose control of cHTN would result in fewer adverse pregnancy outcomes.
What was the research question?
Will a blood pressure goal of <140/90 (versus 160/105) result in a lower incidence of adverse maternal and perinatal outcomes in patients with chronic hypertension in pregnancy?
→ essentially the same question, but more narrowly targeted, than CHIPS.
Methods
Who participated and when?
Recruited
Eligibility:
Pregnant patients with known or new cHTN and singleton pregnancy prior to 23 weeks (33w6d in CHIPS)
New cHTN was diagnosed based on criteria of BP 140/90 on 2 occasions at least 4 hours apart prior to 20 weeks gestation without prior diagnosis.
Pre-existing cHTN was defined by documented elevations in BP and previous/current antihypertensive therapy, including lifestyle modifications alone.
Pregnancy dating needed to be confirmed according to ACOG criteria with ultrasound performed before randomization.
Exclusion criteria:
Severe HTN or BP requiring more than one antihypertensive treatment;
Secondary cause of hypertension (i.e., renal artery stenosis);
Multiple gestation;
“Pre-specified high risk illnesses or complications that may warrant treatment at a lower BP level” - severe cardiac or renal dz as examples
OB conditions that increased fetal risk;
Contraindications to first-line antihypertensive drugs used in pregnancy
How was the study done?
BP was measured with an automated cuff (same across sites) to screening/enrollment and to guide medication adjustments, with research staff performing measurements by a specified protocol.
Randomized to:
Tight control group: goal BP < 140/90
Less tight group: goal BP <160/105
Therapy, if ongoing, was stopped in the less tight group unless severe BP developed.
If a severe BP was seen, the target for acute treatment was <140/90.
Web-based variable block randomization program.
Treatment was supplied as 1st line with nifedipine XL or labetalol and prescribed by trial investigators
Amlodipine and methyldopa were also considered if preferred by patient
Meds were prescribed to maximal recommended dose that was not associated with poor side effects before iniiating a second medication
Control group received medications in a similar fashion only if severe HTN developed.
Pill counts were performed to assess adherence.
What outcomes were they looking for?
Primary outcome
A composite of:
Preeclampsia with severe features occurring up to 2 weeks after birth;
Medically-indicated preterm birth before 35 weeks because of maternal/fetal illness (i.e., not for PPROM/PTL)
Placental abruption
Fetal/neonatal death
ACOG criteria were used to define preeclampsia with severe features; however, a BP of 160/100 or greater in absence of signs and symptoms of preeclampsia/proteinuria/lab abnormalities was not considered sufficient to diagnose PEC with SF.
The primary outcome was assessed in five pre-specified subgroups as well:
cHTN treatment status at baseline:
New diagnosis of cHTN
Diagnosed and receiving meds
Diagnosed and not receiving meds
Race/ethnicity
Diabetes status
Gestational age at enrollment (<14 weeks or > 14 weeks)
BMI (<30, 30-40, and >40).
A primary safety outcome was also prespecified: poor fetal growth
Defined as birth weight less than 10%ile for gestational age and infant sex
Also assessed at <5%ile.
Secondary outcomes were numerous:
Maternal death and various serious complications
Exposure to severe hypertension
Cesarean delivery
Any preterm birth and any serious neonatal complications/NICU stay
Patients were followed to 6 weeks postpartum
A blinded outcome adjudication committee reviewed all patient charts suspected of having primary or secondary outcomes to assess and confirm
2404 patients was the intended sample size (1202 per group) to detect a reduction of 25% in the primary composite outcome, at a baseline incidence as low as 10%.
The discussion in the methods of how this sample size was agreed upon was very interesting and worth a look through for any of our statistics friends out there! – initially wanted to have 4700 patients but after IRB review settled upon this smaller size.
Results
Who did they recruit?
29,772 patients underwent screening, and 2419 subsequently underwent randomization; the final sample size for analysis was 2408, with 1208 in the active treatment arm (tighter control) and 1200 in the control arm (loose control).
83 patients were lost to follow up for the complete study; 38 in the active group and 45 in the control group.
Baseline characteristics were similar:
cHTN status:
56% in each arm had known cHTN and were receiving medication
22% had known cHTN but were not on medication
22% had newly diagnosed cHTN
BMI: both around 37.5
DM: 15.8% in each arm
44.7% in each arm on aspirin therapy
Labetalol was most common medication used (61.7%) followed by nifedipine (35.6%), and only 2.7% received other meds.
Active treatment group had more patients taking meds (88.9% for active, 24.4% for control).
BP also was predictably lower in the active treatment group after randomization:
129.5 mmHg vs 132.6 mmHg (-3.1) systolic
79.1 mmHg vs 81.5 mmHg (-2.3) diastolic
Outcomes
Primary: composite of severe preeclampsia, abruption, medically-indicated PTB < 35wk, fetal/neo death
30.2% of active treatment group
37.0% of control group
aRR 0.82, CI 0.74 - 0.92 – p<0.001
Number of patients needed to treat to prevent one primary outcome event: 14.7 (95% CI 9.4 - 33.7)
By event:
PEC + SF: 23.3% active vs 29.1% control
PTB < 35 wks: 12.2% active vs 16.7% control
By pre-specified subgroups:
The benefit seemed to be present for all pre-specified subgroups, except:
Newly diagnosed cHTN (RR 1.00)
BMI > 40 (RR 0.98)
Primary safety outcome: birth weight < 10%ile
11.2% in active group vs 10.4% in control group
aRR 1.04, 95% CI 0.82 - 1.31; p=0.76. → not statistically different!
For <5%ile: 5.1% vs 5.5% – also not different!
I.e, more aggressive treatment didn’t seem to impact rates of birth weight <10% or <5% as potentially feared.
Secondary outcomes:
Maternal: no substantial differences, except:
Severe-range HTN in 36.1% of active and 44.2% of control
Preeclampsia in 24.4% of active and 31.1% of control (RR 0.79, CI 0.69-0.89)
Fetal: no substantial differences, except:
PTB before 37 weeks: 27.5% active and 31.4% control (RR 0.87, CI 0.77-0.99)
Low birth weight <2500g: 19.2% active and 23.1% control (RR 0.83, CI 0.71-0.97)
Interesting as well that aspirin use did not seem to demonstrate a difference in development of any primary or secondary outcome…
Conclusions and What We Do Now / What Should We Take Away
The authors conclude from this paper that having a target BP of 140/90 or lower was associated with better pregnancy outcomes than a target of 160/105, without any significant differences in safety outcomes for neonates.
Strengths:
A diverse, nationwide cohort with lots of patients
Strictly looking at chronic hypertension with early pregnancy enrollment (prior to 23 weeks)
Modern definitions of preeclampsia and other hypertensive disorders of pregnancy
Overall results consistent with CHIPS – i.e., ~50% reduction in rates of severe-range BP, no difference in birth weight/growth restriction
Weaknesses:
High ratio of patients screened : patients enrolled – over 29k were screened for a trial size of ~2400!
This probably reflects a lot of vigorous selection which is a strength of this study, and importantly the demographics of those screened versus selected did not significantly differ.
Left out a lot of higher risk patients: cardiac/renal disease patients, secondary hypertension, etc.
Additionally, in the prespecified subgroup analyses, the treatment effect was not seen in patients with BMI > 40 or patients with newly diagnosed cHTN – the study was not powered to assess these independently but may need to be seen if other strategies are better in these groups.
The definition of cHTN changed! – ACC/AHA in 2017 (mid-recruitment) lowered the target to 130/80. We don’t know if that might be better, or worse, as a target.
Only short term follow up – longer term follow up will help inform if there are any benefits ultimately with maternal or neonatal risks.
Interesting points:
NNT of 14.7 to reduce primary outcome is really excellent, especially given the other safety data provided in this trial (short-term).
Aspirin use was equal between groups – post hoc analysis demonstrated it did not influence primary outcome measure!
This study probably lends some support to the aspirin skeptics out there, but wouldn’t necessarily throw aspirin away based on this trial alone.
SMFM CHAP Statement: overall supportive of a target to goal BP of <140/90 based on this trial, mentioning the limitations we just went through.