Cardiac resynchronisation therapy (know indications for)
What CRT is
- Pacing both ventricles to correct electrical dyssynchrony in HFrEF
- Hardware: RA lead + RV lead + LV lead via the coronary sinus into a lateral/posterolateral epicardial vein
- CRT-P (pacemaker) vs CRT-D (with defibrillator)
- The only pacing therapy that improves mortality in heart failure
- Corrects the electrical substrate; it is not a substitute for GDMT
Who has dyssynchrony and who responds
- ~1/3 of HFrEF patients have QRS >120 ms; LBBB in ~25%
- ~60-70% respond; ~30% non-responders - the central clinical problem
- LBBB + QRS >150 ms + female + non-ischaemic = best responders
Why dyssynchrony matters
- LBBB -> septum activates early, lateral wall late
- -> Interventricular (RV before LV) and intraventricular (septum vs lateral wall) dyssynchrony
- -> septal "flash"/bounce: septum contracts against a relaxed lateral wall, wasted work
- -> effective dec stroke volume, inc mitral regurgitation (papillary muscle dyssynchrony), inc wall stress
- -> inc myocardial O2 consumption for less output
What CRT does
- Pre-excites the late-activating lateral wall -> coordinated contraction
- -> inc dP/dt, inc SV, dec functional MR, dec LVEDV
- -> reverse remodelling over 3-6 months (the substrate of the mortality benefit)
- Reverse remodelling, not acute haemodynamics, predicts outcome
Why RV pacing is harmful
- RV apical pacing creates an iatrogenic LBBB
- High RV pacing burden -> pacing-induced cardiomyopathy in ~10-20%
Selecting patients - an ECG decision, not an echo one
Selecting patients - it is an ECG decision, not an echo one
- QRS duration and morphology on 12-lead ECG
- LVEF by TTE (or CMR/MUGA if poor windows)
- NYHA class on optimal medical therapy for >=3 months
- Echo dyssynchrony indices (TDI, speckle tracking) were tested in PROSPECT and failed to predict response - do not select on them
Predictors of response
| Better response | Worse response |
|---|---|
| True LBBB morphology | RBBB or non-specific IVCD |
| QRS >=150 ms | QRS 130-149 ms, and no benefit <130 ms |
| Non-ischaemic aetiology | Extensive posterolateral scar (LV lead over scar) |
| Female sex | Male |
| Sinus rhythm | AF with incomplete biventricular capture |
| Less LV dilatation | Severe dilatation, advanced disease |
- Biventricular pacing percentage >98% is required - check it at every review
Indications
Indications - LVEF <=35%, symptomatic, on optimal GDMT >=3 months, expected survival >1 yr
| QRS | Morphology | Class |
|---|---|---|
| >=150 ms | LBBB | I |
| >=150 ms | non-LBBB | IIa |
| 130-149 ms | LBBB | I (IIa in some documents) |
| 130-149 ms | non-LBBB | IIb |
| <130 ms | any | III - harm (EchoCRT: inc mortality) |
Other Class I settings
- HFrEF requiring ventricular pacing for high-grade AV block -> CRT rather than RV pacing (includes AF)
- Upgrade from pacemaker/ICD to CRT if EF <=35% with high RV pacing burden and worsening HF
In atrial fibrillation
- CRT is reasonable if QRS criteria met, but requires near-100% biventricular capture
- If capture inadequate: AV node ablation ("pace and ablate")
- APAF-CRT: dec mortality vs rate control in narrow-QRS AF with HF
- Rate-control drugs alone rarely achieve adequate capture
CRT-P vs CRT-D
- CRT-D if the patient also meets primary-prevention ICD criteria (ischaemic aetiology, younger, good survival expectancy)
- CRT-P preferred in: non-ischaemic aetiology with good expected reverse remodelling, elderly, frailty, significant comorbidity
- DANISH: no mortality benefit from ICD in non-ischaemic cardiomyopathy
Conduction system pacing - where it sits in 2026
Conduction system pacing (CSP) - where it sits in 2026
- Left bundle branch area pacing (LBBAP) - lead screwed deep into interventricular septum to capture the conduction system
- LEFT-BUNDLE-CRT (2026): LBBAP did NOT meet non-inferiority vs biventricular CRT in typical LBBB, though both achieved high response rates
- LECART: fewer device-related reinterventions with LBBAP
- -> BiV pacing remains the default CRT strategy in LBBB; LBBAP is the preferred CSP option when the coronary sinus lead fails, the anatomy is unsuitable, or an AV-block patient needs physiological pacing
- His-bundle pacing: fallback if LBBAP fails (high thresholds, lead revision rate)
Optimisation of a non-responder
1. Confirm BiV pacing >98% - ectopy, AF, fusion, loss of capture
2. Reprogram AV/VV delays (echo-guided or device algorithm)
3. Reposition or add an LV lead (avoid apical position and scar)
4. Re-check GDMT is truly maximal
5. Consider CSP upgrade, or advanced HF referral (transplant/LVAD)
Trial evidence
- Trial evidence
- CARE-HF - CRT-P vs medical therapy: dec mortality
- COMPANION - CRT-P and CRT-D vs medical therapy: dec death/hospitalisation
- MADIT-CRT and REVERSE - mild symptoms (NYHA I-II): dec HF events, reverse remodelling; benefit confined to QRS >=150 ms and LBBB
- EchoCRT - narrow QRS with echo dyssynchrony: harm
- BLOCK-HF - CRT better than RV pacing when pacing needed with EF <=50%
- APAF-CRT - AV node ablation + CRT in AF: dec mortality
- ~60% gain a modest but sustained (6-12 months or longer) symptomatic improvement
- Association with AF: CRT reduces AF burden through reverse remodelling, but AF undermines capture
Response over time
- Reverse remodelling peaks at 3-6 months; reassess EF before deciding on non-response
- Super-responders (~10%): EF normalises
- This is HFimpEF - continue all GDMT and the device; withdrawal causes relapse (TRED-HF principle)
- Non-responders: worse prognosis than never having met criteria
Complications
- Implant: coronary sinus dissection/perforation, tamponade, pneumothorax, LV lead failure to place (~5%)
- Phrenic nerve stimulation - diaphragmatic twitching; reprogramme the pacing vector or reposition
- Lead dislodgement (LV lead the most prone), infection (device pocket infection -> full system extraction)
- Loss of biventricular capture - the commonest reversible cause of deterioration
- Inappropriate ICD shocks (CRT-D)
- Device-related: generator change every 5-8 yr, venous stenosis/occlusion
- Driving restrictions apply after implant and after any ICD therapy
🔒
15 more sections, plus exam facts
Premium unlocks every note across every specialty, and the full exam fact library behind it.
Get premium access