KEY POINTS
- This prospective, single-center phase II randomized superiority trial (TCTR20210929004) evaluated hyaluronic acid rectal spacer injection during MRI-guided adaptive brachytherapy for cervical cancer. Between August 2023 and December 2024, 26 patients were randomized equally to spacer or standard treatment, compared with 40 originally planned. Recruitment stopped because research funding was exhausted, not because of a predefined safety signal.
- All patients received definitive chemoradiotherapy with 50.4 Gy in 28 fractions, weekly cisplatin 40 mg/m², and HDR MRI-guided adaptive brachytherapy delivering 8.7–9.5 Gy in three fractions to the high-risk CTV. 92.3% underwent combined intracavitary/interstitial brachytherapy. In the intervention group, cross-linked hyaluronate gel was injected into the rectocervical space under ultrasound guidance, aiming to achieve approximately 1 cm of separation.
- The primary endpoint was not met. Mean per-fraction rectal D2cm³ was 3.8 ± 0.7 Gy with spacer versus 4.1 ± 0.7 Gy without, an absolute difference of −0.3 Gy (95% CI −1.0 to 0.2; p=0.206). Cumulative rectal D2cm³ was identical at 65.9 Gy EQD2 in both groups (p=0.999), and cumulative bladder D2cm³ and CTVHR D90/D98 did not differ significantly.
- Cumulative sigmoid D2cm³ was numerically higher with spacer injection, measuring 66.8 ± 4.9 versus 62.8 ± 4.7 Gy EQD2 (p=0.044), although all values remained within the protocol's sigmoid constraint of 70 Gy EQD2. Given the small sample size, anatomical variation, and multiple comparisons, this exploratory difference does not establish that spacer placement increases sigmoid toxicity or dose.
- Spacer geometry varied between patients and fractions. Mean gel thickness was 1.3 ± 0.5 cm, and mean CTVHR-to-rectum distance during brachytherapy was 2.4 ± 1.1 cm in the intervention group. Additional gel injection was required in 11/13 patients (84.6%) during the second fraction and 9/13 (69.2%) during the third. Greater target–rectum distance was associated with lower rectal D2cm³ (β −0.30 Gy/cm; nominal p=0.045), but this relationship lost significance after false-discovery-rate adjustment.
- Acute toxicity was generally mild, with no grade ≥3 adverse events or serious spacer-related complications. Grade ≥2 acute rectal toxicity occurred in 30.8% of spacer patients versus 15.4% of controls (p=0.640), with no grade ≥2 bladder events. Gel absorption was visible in 8/11 evaluable patients at three months and all six evaluable patients at 12 months, although these small numbers cannot characterize long-term spacer persistence or rare complications.
- The trial's limited statistical power, unblinded planning, and high baseline use of MRI-based optimization and combined IC/IS techniques may have reduced the ability to detect an additional spacer effect. Late rectal toxicity, quality of life, and oncological outcomes were not reported, and anatomical spacer–high-dose region overlap was not systematically evaluated. The findings therefore cannot establish either clinical benefit or absence of benefit in selected patients with unfavorable anatomy.
CLINICAL TAKEAWAY
Routine hyaluronic acid rectal spacer injection did not produce a statistically significant rectal dose reduction during optimized MRI-guided cervical brachytherapy in this underpowered randomized study. The results favor further investigation of anatomy-based patient selection and spacer positioning rather than universal spacer insertion, while late toxicity and clinical effectiveness remain unproven.