Pregnancy in Genetic Cardiomyopathy
Definition
Pregnancy in genetic cardiomyopathy is the management of a woman who carries a pathogenic or likely pathogenic cardiomyopathy variant — or has the phenotype — through conception, pregnancy, delivery and the puerperium. Its defining problem is that the risk tools used to decide who is safe to become pregnant cannot see the genotype. CARPREG II and the modified WHO classification both score the phenotype in front of them; neither accepts a variant as an input. A 30-year-old with an LMNA variant, a normal ejection fraction and no arrhythmia is mWHO class I — indistinguishable, by either tool, from any other pregnant woman — while carrying 90–95% lifetime penetrance and the highest arrhythmic risk in cardiomyopathy (sources/cmp-pregnancy-jaccadv-2026, rating: medium).
Key Concepts
Why pregnancy is the stress test for a genotype
- Plasma volume, stroke volume and heart rate rise, systemic vascular resistance falls, and cardiac output increases 50–80% during labour, resolving over 48 hours to 2 weeks. Autotransfusion from the uteroplacental circulation and inferior vena caval decompression then raise central venous pressure again in the immediate postpartum period.
- Those are exactly the haemodynamic levers a compensated cardiomyopathy cannot afford. Pregnancy does not create cardiomyopathy; it advances the date at which an existing predisposition becomes a phenotype.
- Up to 40% of pregnancy-related maternal deaths from cardiovascular disease in the US are attributable to cardiomyopathies — the reason the field is being reorganised around genetics rather than around a single peripartum window.
The five phenotypes, by risk
| Phenotype |
mWHO |
Key risk figures |
The one thing that must happen before conception |
| DCM |
III (LVEF 30–45%), IV (<30% or NYHA III–IV) |
MACE in 39% of pregnancies; mortality up to 7%; pre-pregnancy LVEF <50% is the strongest predictor |
Convert GDMT off teratogens, accepting the risk of decompensation during conversion |
| PPCM |
IV if any residual dysfunction |
~15% carry a DCM gene, >10% TTNtvs; recurrence risk persists even at recovered EF |
Genetic counselling and testing — recommended for all PPCM, done in 11% |
| ACM |
II–III |
13% of pregnancies with ventricular arrhythmia, 5% with heart failure; postpartum VA HR 13.74 |
Secondary-prevention ICD before conception |
| HCM |
II–III |
Mortality 0.2%; <10% symptom exacerbation; no difference obstructive vs nonobstructive |
Myectomy if indicated; stop cardiac myosin inhibitors |
| GP/PN |
Not assigned |
10% significant cardiac complications in rare DCM variants; 34% of PKP2 relatives progress to definite ARVC |
Baseline ECG, Holter, echo — phenotype, not genotype, drives the decision |
DCM — the highest-risk phenotype, and the lowest-technology answer
- Pre-pregnancy LVEF <50% outranks every genetic variable. It predicts heart failure, further LVEF deterioration and death. Genotype tells you how likely that LVEF is to fall; the echocardiogram tells you where it is now.
- mWHO III (LVEF 30–45%) means pregnancy is discouraged, not contraindicated; mWHO IV (LVEF <30% or NYHA III–IV) means contraindicated, with Pregnancy Heart Team assessment at a tertiary centre and at least monthly follow-up if the pregnancy continues anyway.
- The therapeutic bind is that GDMT itself is the teratogen. ACE inhibitors, ARBs, MRAs, SGLT2 inhibitors, ivabradine and vericiguat must all be stopped and replaced with beta-blockers, hydralazine and nitrates — and the conversion itself can precipitate reflex hypertension or decompensation. Severe or symptomatic patients may need admission to make the switch.
- ESC 2023's nondilated LV cardiomyopathy has no pregnancy data at all; the paper extrapolates from DCM because there is nothing else to extrapolate from.
PPCM — a genetic disease that is only sometimes recognised as one
- ~15% of PPCM shares a genetic predisposition with DCM — truncating variants in TTN, FLNC, DSP, MYH6, MYH7, BAG3. Over 10% of 469 women in an international sequencing study carried TTNtvs, associated with lower ejection fraction but not with timing of presentation, preeclampsia, or likelihood of recovery. See entities/Peripartum-Cardiomyopathy.
- Guidelines now recommend genetic counselling and testing for all women with PPCM. The implementation gap is stark: 11% were referred in a recent study.
- Subsequent pregnancy is decided on ejection fraction, and "recovered" has no agreed definition. ESC 2025 discourages further pregnancy if LV dysfunction persists (LVEF <50%) but notes that risk of recurrent dysfunction, life-threatening arrhythmia and rarely death continues even with recovered EF. Prospective European data suggest LVEF 40–50% may proceed under monitoring; US retrospective data (80% African American) show adverse outcomes in 53% versus 20% and 5-year mortality 13.3% versus 3.3% when LVEF did not recover.
ACM — the postpartum window is the danger
- Pregnancy is uneventful for most women with ACM; in the NORDIC ARVC Register (199 females) the number of prior pregnancies did not increase ventricular arrhythmia risk.
- But women who gave birth with previously diagnosed ACM had elevated postpartum ventricular arrhythmia risk, HR 13.74 (95% CI 2.9–63; P=0.001), and 13% of pregnancies in the Johns Hopkins/Dutch registry were complicated by ventricular arrhythmia. The third trimester and puerperium are the highest-risk periods.
- HRS 2023 gives a Class 1 recommendation for secondary-prevention ICD with prior ventricular arrhythmia or aborted SCD, and implantation before conception is preferred — device placement during pregnancy adds procedural and fetal risk to a decision that was already indicated.
- Beta-blockers first-line; flecainide, sotalol and mexiletine when beta-blockers are contraindicated; amiodarone only for life-threatening arrhythmia and only as a single dose; catheter ablation 2a if arrhythmia persists despite drugs.
HCM — the best-tolerated phenotype, with one new contraindication
- Maternal mortality 0.2%, no difference between obstructive and nonobstructive disease, <10% symptom exacerbation, and a 2025 propensity-matched analysis showing no difference in 10-year all-cause mortality between parous and nulliparous women. Pregnancy itself mildly dilates the LV and can lower the dynamic gradient.
- Risk concentrates in severe hypertrophy, a gradient >100 mmHg, and a personal or family history of severe arrhythmia.
- Cardiac myosin inhibitors are contraindicated in pregnancy for teratogenicity. The consequence is a pre-conception surgical decision: myectomy should be done before pregnancy where indicated, because the pharmacological alternative is unavailable. See entities/Mavacamten.
- Intrapartum: phenylephrine is the preferred vasopressor (pure alpha, no tachycardia), with volume to maintain preload. Assisted second stage — labouring down and delivering the head and body instrumentally — limits the haemodynamic cost of prolonged Valsalva.
Genotype-positive, phenotype-negative — the least evidenced group
- The fastest-growing population as cascade testing spreads, and the one with almost no data. Guidance is inferred from the variant, the family history and a baseline evaluation.
- Rare DCM variants: pregnancies mostly uneventful, but 10% developed clinically significant cardiac complications during or shortly after pregnancy — roughly the rate seen in women whose cardiac disease predated pregnancy.
- PKP2 families: 34% of GP/PN relatives progress to definite ARVC over time — a dynamic prephenotypic state that physiological stress may unmask.
- The operative principle, in the paper's own words: "near-term clinical outcomes are driven primarily by phenotype." Genotype sets the surveillance plan; the echocardiogram, ECG and Holter set the pregnancy decision.
- A VUS does not remove the need for evaluation. It changes the question from "what does the variant predict?" to "what does the phenotype show?" — ECG, Holter, echo, and in selected cases exercise testing and CMR.
- Aspirin from 12 weeks is recommended for preeclampsia prevention in cardiomyopathy generally, but its use in GP/PN women is explicitly admitted to be unstudied.
Reproductive genetics — three options and three gates
| Option |
Timing |
Gate that closes it |
| PGT-M with IVF |
Before conception |
Needs a diagnostic linkage protocol requiring multiple genotype-positive relatives — often impossible in de novo disease; never for a VUS; some variant classes (small deletions, duplications, complex rearrangements) are undetectable |
| Chorionic villus sampling |
10–13 weeks |
Procedural loss ~1/200; confined placental mosaicism can cause misdiagnosis because the cells are placental, not fetal |
| Amniocentesis |
After 15 weeks |
Procedural loss ~1/1,000; later result |
- PGT-M carries the risks of IVF itself on top of the genetics: embryo loss from sampling, false positive and false negative results from embryo mosaicism (the sampled cells may not represent the embryo), and higher maternal morbidity than spontaneous conception. Cost and emotional burden are substantial and are not the smallest considerations.
- The structural unfairness: the reproductive option is least available exactly where the diagnosis is newest — a de novo variant in a young woman with no affected relatives cannot be worked up for PGT-M.
The surveillance schedule
| Patient group |
Rhythm and biomarker monitoring |
| All cardiomyopathy |
ECG, 24-hour Holter, NT-proBNP every 3 months |
| Reduced LVEF, NYHA III–IV, or any symptom change |
Monthly, with ambulatory ECG |
| mWHO II |
Bimonthly clinical reassessment |
| mWHO III |
Monthly |
| Low-risk (2025 European guideline) |
Every 3 months with NT-proBNP, ECG, echo and 24-hour Holter |
- Low-dose aspirin (81 or 162 mg in the US) from 12 weeks for preeclampsia prevention — preeclampsia superimposed on cardiomyopathy carries a disproportionate mortality.
- Delivery at ≥39 weeks for most, earlier only for worsening function, increasing arrhythmia burden or preeclampsia. Vaginal delivery preferred, with neuraxial anaesthesia to blunt catecholamine surges. Telemetry intrapartum and through the immediate postpartum period; electrolytes optimised intrapartum.
- Postpartum is the highest-risk window — most PPCM presents in the first postpartum month, with cardiogenic shock, arrhythmia and venous thromboembolism. Established BNP/NT-proBNP thresholds remain valid despite the haemodynamic shifts.
- Breastfeeding: beta-blockers, ACE inhibitors, most ARBs (except candesartan) and MRAs are compatible. ARNI has no documented human harm but animal data show milk transfer. SGLT2 inhibitors are not recommended — theoretical neonatal renal toxicity, no human safety data.
Contradictions / Open Questions
- The risk tools are structurally blind to the thing that matters most. CARPREG II and mWHO score phenotype only. A genotype-positive woman with normal function scores as low risk in both, despite gene-specific penetrance ranging from 55% (MYBPC3) to 90–95% by age 60 (LMNA). The paper identifies the problem clearly and does not solve it — no genotype-inclusive score exists. (sources/cmp-pregnancy-jaccadv-2026)
- The ACM postpartum hazard ratio is implausibly large and rests on 199 women. HR 13.74 with a confidence interval of 2.9 to 63 is either the most important number in the paper or an artefact of a small registry with few events — and the same registry reports that pregnancy was "uneventful for the majority." Both statements are in the same paragraph.
- "Recovered" PPCM has no definition, and the guidance turns on it. ESC 2025 uses LVEF <50% as the line for discouraging pregnancy; prospective European data suggest 40–50% can proceed; US data show a 2.7-fold adverse outcome rate in non-recovered women. A woman at 50% sits on all three rules at once.
- Myosin inhibitor teratogenicity converts a cardiac decision into a surgical one. Myectomy before pregnancy is now recommended partly because the drug alternative is contraindicated, not because the myectomy indication itself changed — an inversion of the usual evidence hierarchy.
- The GP/PN evidence base is two small studies wearing a guideline's clothes. The 10% complication rate comes from one DCM-variant cohort; the 34% progression rate from PKP2 family screening. Neither is a pregnancy outcome study, yet together they determine counselling for the largest and fastest-growing group.
- PGT-M's eligibility criteria exclude the patients who most need reproductive planning. It requires multiple genotype-positive relatives, excludes VUS, and cannot detect some variant classes — so the newest diagnoses, which are the ones that prompt the question, are the least eligible.
- Aspirin for preeclampsia prevention in GP/PN women is recommended and unstudied in the same sentence. The paper says the benefit "has not been rigorously evaluated and further research is needed" while listing it in the management table.
- Amiodarone's "single dose in life-threatening emergency" has no defined dose, duration or threshold — in a population where the alternative drugs are also imperfect and the fetus is at risk throughout.
- Surveillance intervals differ between cited sources without adjudication: the 2025 European guideline says every 3 months for low-risk patients, while the consensus document specifies bimonthly for mWHO II and monthly for mWHO III. No source explains which applies to a patient who is both.
- No cost, access or equity analysis of a care model requiring a Pregnancy Heart Team, genetic counsellors, PGT-M and serial CMR — in a field where PPCM incidence is highest in the populations least likely to reach any of them.
Connections
Sources