KEY POINTS
- This structured narrative review examined literature through March 2026 on four recurring sources of uncertainty in cervical high-dose-rate brachytherapy: steep spatial gradients, temporal delivery effects, high-dose biological conversion and inter-fraction dose accumulation.
- Existing EQD2 targets remain clinically validated. In EMBRACE-I, high-risk clinical target volume D90 of at least 85 Gy EQD2 was associated with 95% three-year local control for squamous-cell carcinoma.
- Steep brachytherapy gradients amplify geometric uncertainty. A 1 mm applicator-reconstruction error may alter rectal D2cc by approximately 5% and high-risk clinical target volume D90 by around 2%, with additional nonlinear amplification after EQD2 conversion.
- Iridium-192 activity falls to approximately 43% after 90 days, increasing delivery time about 2.3-fold. A fraction requiring 8–10 minutes with a fresh source may require 18–23 minutes near the end of the exchange cycle.
- For a 20-minute fraction, the Lea-Catcheside factor is approximately 0.96 with a repair half-time of 1.5 hours but 0.86 with a 0.5-hour assumption, illustrating the model dependence of time-corrected biological dose.
- Uncertainty in cervical-tumour alpha/beta estimates produced deviations of up to 10 Gy EQD2 for commonly used schedules. Standard linear-quadratic conversion is therefore clinically useful but not biologically exact at high per-fraction doses.
- Deformable accumulation provides spatial information but remains method-dependent. Reported differences from direct dose-volume-histogram addition ranged from 0.4 ± 0.3 Gy for bladder D2cc in one study to 2.4% for rectum and 5.2% for bladder in another.
- Model-based algorithms can also shift reported dose: AcurosBV produced an average 4.3% lower high-risk clinical target volume D90 and 5–8% lower organ-at-risk D2cc than TG-43. Because current outcome thresholds were calibrated with TG-43, both values should be reported in parallel rather than applying old thresholds directly to new calculations.
CLINICAL TAKEAWAY
Conventional EQD2-based high-risk clinical target volume D90 and organ-at-risk D2cc should remain the primary clinical reference because they are linked to mature outcome data. Delivery time, source activity, applicator type, accumulation method and dose-calculation formalism should be documented when relevant; newer biological or spatial metrics should complement rather than replace established constraints.