How can the right ventricle be the chamber that often decides whether a patient thrives when clinical trials still treat it like a service corridor behind the lobby?
That is the architectural tension in Surkova and colleagues' 2026 review in JAMA Cardiology: the right ventricle, or RV, is load-bearing, weirdly shaped, and clinically loud, yet trial design often keeps measuring the easier, shinier rooms of the heart. The left ventricle gets the atrium-view penthouse. The RV gets a clipboard, a shrug, and the phrase "technically difficult images," which is cardiology for "the plumbing is behind the wall and everyone is tired."
A Strange Little Room With Big Structural Duties
The RV is not just a smaller left ventricle wearing the wrong jacket. It pumps blood into the lungs, works in a lower-pressure system, and changes shape when pulmonary pressure, valve disease, congenital heart disease, or heart failure starts leaning on it. Think of it as a flexible pavilion roof: elegant under normal load, alarming when someone parks a truck on it.
That shape makes measurement tricky. Echocardiography is fast and widely available, but the RV's geometry laughs at simple assumptions. Cardiac MRI gives beautiful structural sight lines and is often treated as the more exacting surveyor, but it costs more, takes longer, and is not always trial-friendly. CT can help. Right heart catheterization gives pressure and flow data from inside the machinery itself, though nobody confuses a catheter lab with a casual Tuesday errand.
Surkova et al. argue that RV metrics deserve more serious use as clinical trial end points because they can reveal whether a drug or device is improving the actual mechanical stress pattern of disease, not just repainting the facade [Surkova 2026].
End Points Are the Building Inspectors
Clinical trials need end points: measurable outcomes that say whether an intervention worked. Death, hospitalization, and symptoms matter most, of course. But they can require large trials and long timelines. Imaging and hemodynamic end points can act like structural sensors: not the whole building, but a well-placed gauge that tells you whether the beam is bending.
For RV trials, those gauges might include RV size, ejection fraction, free-wall longitudinal strain, fractional area change, TAPSE, pressure-volume relationships, pulmonary vascular resistance, or RV-arterial coupling. Some are elegant. Some are fussy. Some have the aesthetic confidence of a basement utility panel, but they tell you things you need to know.
Recent evidence gives the review some scaffolding. A 2024 meta-analysis found that RV free-wall longitudinal strain predicted outcomes in pulmonary hypertension and often outperformed older RV measures such as TAPSE and fractional area change [Nabeshima 2024]. The 2025 American Society of Echocardiography right-heart guideline also pushes for more systematic RV assessment, which is basically the profession saying: please stop eyeballing the load-bearing wall [Mukherjee 2025].
The Blueprint Problem
The review's strongest point is not "measure the RV because RVs are neat," although, fair, they are. It is that trial end points need three things at once: clinical relevance, analytical validity, and operational feasibility.
That triangle is where many promising RV metrics get trapped like a sofa in a stairwell.
A metric may be clinically meaningful but hard to acquire across 80 trial sites. It may be reproducible in a core lab but useless if half the scans arrive with missing views. It may detect short-term improvement, yet still lack proof that the change predicts fewer hospitalizations or longer survival. In architecture terms, the facade rendering looks gorgeous, but the elevator shaft is not aligned.
This is why standardization matters. Same acquisition protocols. Same definitions. Centralized analysis where possible. Clear thresholds for meaningful change. Disease-specific choices rather than one majestic RV number dragged through heart failure, pulmonary hypertension, tricuspid valve disease, and congenital disease like a universal Allen key.
Enter AI, the Intern With a Laser Level
Artificial intelligence may help here, but not by sprinkling algorithm glitter on bad data. The useful version is more boring and more valuable: automated segmentation, quality control, faster contouring, consistent measurements, and maybe better prediction from images that humans find visually noisy.
Recent reviews and statements describe AI as increasingly useful across cardiac MRI, CT, and echocardiography, especially for acquisition, reconstruction, segmentation, and reporting [Morales 2024; Maurovich-Horvat 2025]. A 2025 systematic review of AI for right-heart function found promise, but also heterogeneity and evidence limits - the algorithmic equivalent of a very confident apprentice who still needs supervision [Eini 2025]. PanEcho, a multitask deep-learning system for echocardiography, shows how broad automated echo interpretation is moving from sketchbook to construction site [Holste 2025].
The catch: AI cannot rescue sloppy trial architecture. If images are inconsistent, labels are noisy, and outcomes are weakly linked, the model becomes a high-speed machine for polishing ambiguity. Very modern. Very expensive. Very "open floor plan with no outlets."
Why This Review Lands
Surkova and colleagues are not asking trialists to worship the RV. They are asking them to stop designing cardiovascular trials as if the RV were decorative negative space. In many diseases, the RV is where pressure, volume, valves, lungs, and muscle performance meet. It is the junction box. Ignore it, and you may miss whether a therapy actually changed the structure of illness.
The review is also refreshingly practical. It does not pretend every RV metric is ready for prime time. It says the next phase needs validation, multicenter feasibility work, consistent imaging, core-lab discipline, and links between short-term RV changes and outcomes patients can feel. That is less glamorous than a miracle biomarker, but so is rebar, and buildings fall down without it.
The future trial may use RV metrics as sharper, earlier signals: smaller studies, cleaner mechanistic insight, and better decisions about which therapies deserve the next level of investment. If that happens, the RV will not just be the odd-shaped chamber in the corner. It will be part of the main structural plan.
References
Surkova E, Lakatos B, Kempton H, et al. Right Ventricular Metrics as End Points in Clinical Trials: A Review. JAMA Cardiology. Published online June 24, 2026. PMID: 42340710. DOI: 10.1001/jamacardio.2026.2053
Nabeshima Y, Kitano T, Node K, Takeuchi M. Prognostic value of right ventricular free-wall longitudinal strain in patients with pulmonary hypertension: systematic review and meta-analyses. Open Heart. 2024;11(1):e002561. PMID: 38325907. PMCID: PMC10860115. DOI: 10.1136/openhrt-2023-002561
Mukherjee M, Rudski LG, Addetia K, et al. Guidelines for the Echocardiographic Assessment of the Right Heart in Adults and Special Considerations in Pulmonary Hypertension. Journal of the American Society of Echocardiography. 2025;38(3):141-186. PMID: 40044341. DOI: 10.1016/j.echo.2025.01.006
Morales MA, Manning WJ, Nezafat R. Present and Future Innovations in AI and Cardiac MRI. Radiology. 2024;310(1):e231269. PMID: 38193835. DOI: 10.1148/radiol.231269
Maurovich-Horvat P, et al. Use of AI in Cardiac CT and MRI: A Scientific Statement. Radiology. 2025. PMID: 39873607. DOI: 10.1148/radiol.240516
Eini P, Serpoush H, Rezayee M, Tremblay J. Automated assessment of right heart function by artificial intelligence: A systematic review and meta-analysis. European Journal of Radiology Open. 2025. PMID: 41458524. PMCID: PMC12741415. DOI: 10.1016/j.ejro.2025.100713
Holste G, Oikonomou EK, Tokodi M, Kovács A, Wang Z, Khera R. Complete AI-Enabled Echocardiography Interpretation With Multitask Deep Learning. JAMA. 2025;334(4):306-318. PMID: 40549400. DOI: 10.1001/jama.2025.8731
Disclaimer: This blog post is a simplified summary of published research for educational purposes. The accompanying illustration is artistic and does not depict actual model architectures, data, or experimental results. Always refer to the original paper for technical details.