A Normal Ejection Fraction in a Childhood Cancer Survivor
A 10-year-old girl (body surface area 1.05 m squared) is three years out from therapy for acute lymphoblastic leukemia, with a cumulative doxorubicin dose of 240 mg per m squared. She reports fatigue and cannot keep up in gym. Heart rate 100 bpm, blood pressure 96/60 mmHg, hemoglobin normal. The echocardiogram report reads: ejection fraction 52% by Teichholz, grossly normal systolic function. Left ventricular end-diastolic dimension 4.6 cm (z-score +2.1) and end-systolic dimension 3.35 cm, giving a fractional shortening of 27%. Biplane volumes are 98 mL in diastole and 47 mL in systole. The oncologist asks whether the fraction is fine and whether she can begin a new anthracycline-containing regimen.
1. Critique the reported ejection fraction, and state what a defensible measurement would look like given these numbers.
Read the worked answer
What the numbers say:
- Fractional shortening = (4.6 - 3.35) / 4.6 = 27%, which sits at or below the lower limit of normal for a child (commonly quoted around 28-30%).
- Biplane ejection fraction = (98 - 47) / 98 = 52% - low-normal at best, not normal.
- The ventricle is already dilated (z-score +2.1), and indexed volumes are at the upper limit or above for body size on most pediatric nomograms.
Why Teichholz is the wrong tool here: it assumes a normally shaped ventricle and derives volume from a single cubed dimension, so it systematically misestimates a dilated or distorted ventricle. It also cannot detect regional dysfunction.
What to do instead:
- Biplane method of disks (Simpson) from a true apical four- and two-chamber view, tracing the blood-tissue interface consistently, excluding trabeculations per guideline convention.
- Report indexed end-diastolic and end-systolic volumes with z-scores, not just a percentage.
- Avoid apical foreshortening (the most common error), average over three to five beats, and use the same method and vendor on every follow-up.
- If endocardial definition is poor, consider contrast enhancement or three-dimensional measurement rather than guessing.
Conclusion: 'normal' is not a defensible label for a dilated ventricle with a fractional shortening of 27% and an ejection fraction of 52%. At minimum this is borderline systolic dysfunction and needs a proper measurement before any drug decision.
2. What would you add to detect cardiotoxicity earlier, and how do you interpret each measurement?
Read the worked answer
Global longitudinal strain. Normal is roughly -18% to -22% depending on vendor and software. A relative reduction of more than about 15% from her own baseline is the accepted early signal of cardiotoxicity, and typically precedes an ejection fraction fall. Serial comparisons must use the same vendor and software version; absolute thresholds across platforms are not interchangeable.
Diastolic indices. Mitral E/A, septal and lateral annular e' velocities, E/e' ratio, and left atrial volume index. Diastolic abnormality frequently precedes systolic decline after anthracyclines.
Right ventricular function. Tricuspid annular plane systolic excursion z-score and fractional area change - the right ventricle is exposed too, and is often ignored.
Laboratory and clinical context. Troponin and natriuretic peptide, plus her end-of-therapy study as the comparator.
Load dependence is the trap: tachycardia, fever, anemia, and dehydration all shift these numbers. A single study in a fatigued child with a heart rate of 100 should be repeated once she is envolemic and rested before a durable conclusion is recorded.
3. The oncologist needs a recommendation, not a number. How do you frame the answer, and what specific caveats do you attach?
Read the worked answer
Framing: present the findings as a pattern, not a single index - dilated ventricle, fractional shortening at the lower limit, ejection fraction low-normal, and symptoms of exercise intolerance. That pattern is not 'fine'.
Recommendation, stated plainly:
- This is at least mild cancer-therapy-related cardiac dysfunction and warrants cardiology co-management and a proper quantitative study (biplane volumes, strain, diastolic indices, right ventricular assessment) before the next anthracycline exposure.
- Escalate the discussion if strain is reduced, if indexed volumes are confirmed enlarged, or if symptoms progress - the decision to continue or modify therapy belongs to oncology, but the echocardiographer's job is to make sure the risk is visible.
Caveats to attach:
- Ejection fraction thresholds for 'normal' differ between adult and pediatric practice and between guidelines for cardiotoxicity; do not import an adult cutoff, and do not call 52% normal merely because it is above 50%.
- Measurement method matters more than the number: repeat with the same modality for comparability.
- Preload and heart rate effects mean one study is a snapshot; document the conditions (heart rate, hydration, hemoglobin) in the report.
- Absence of a wall motion abnormality or of strain reduction does not exclude risk if the volumetric data are abnormal.
The judgment tested here is refusing to let a single reassuringly formatted number end the conversation when the supporting measurements point the other way.