Three-week postmastectomy RT maintains durable locoregional control at 10 years

Ten-year CHN HYPOPMRT results show durable locoregional control and similar survival with 43.5 Gy in 15 fractions vs 50 Gy in 25 fractions after mastectomy, with no emerging severe late-toxicity signal.

Why this matters

Hypofractionation is well established after breast-conserving surgery, but the evidence base has historically been less mature for high-risk patients requiring postmastectomy radiotherapy with regional nodal treatment.

CHN HYPOPMRT previously showed that 43.5 Gy in 15 fractions over 3 weeks was noninferior to conventional 50 Gy in 25 fractions for 5-year locoregional recurrence.

The ASTRO 2026 update addresses the question that matters for a curative breast cancer regimen: does that result remain reassuring more than a decade later?

With a median follow-up of 11.5 years, locoregional control remains similar and no meaningful severe late-safety signal has emerged, although mild pulmonary fibrosis was more frequent after hypofractionation.

Study design

CHN HYPOPMRT was a single-center, open-label, randomized phase III noninferiority trial in women with high-risk breast cancer treated with mastectomy and axillary dissection.

Patients were randomized 1:1 to:

  • conventional PMRT: 50 Gy in 25 fractions over 5 weeks
  • hypofractionated PMRT: 43.5 Gy in 15 fractions over 3 weeks

Radiotherapy covered the chest wall and supraclavicular region.

The primary endpoint was 5-year cumulative locoregional recurrence.

Secondary endpoints included:

  • disease-free survival
  • overall survival
  • late radiation-related toxicity

A total of 820 patients were enrolled, with 810 included in the modified intention-to-treat analysis:

  • 409 received conventional fractionation
  • 401 received hypofractionation

The population was high risk. Approximately 94% had stage III disease, and all patients received chemotherapy.

Median follow-up was 11.5 years.

Key results

The previously reported 5-year locoregional recurrence rates were:

  • 8.6% with conventional fractionation
  • 9.6% with hypofractionation
  • HR 1.13
  • 90% CI 0.77-1.66

At 10 years, locoregional recurrence remained similar:

  • 10.3% with conventional fractionation
  • 12.0% with hypofractionation
  • HR 1.19
  • 95% CI 0.79-1.82

There were also no significant differences in long-term survival outcomes.

At 10 years:

  • DFS: 65.2% with conventional RT vs 69.9% with hypofractionated RT
  • OS: 72.9% vs 77.5%

Most locoregional failures were regional rather than local.

Ten-year local recurrence remained uncommon:

  • approximately 2.0% vs 3.1%

while regional recurrence accounted for most failures.

Late toxicity

Severe late toxicity remained uncommon.

There were:

  • no grade 4-5 late toxicities
  • no cases of brachial plexopathy

Late lymphedema rates were similar between groups.

The notable difference was pulmonary fibrosis.

Grade 1-2 pulmonary fibrosis occurred in:

  • 11.7% with conventional fractionation
  • 19.7% with hypofractionation
  • p=0.008

The reported pulmonary fibrosis events were predominantly low grade, and the presentation did not identify a corresponding increase in severe pulmonary toxicity.

Other late adverse events, including shoulder dysfunction, ischemic heart disease, and rib fracture, were uncommon and did not differ meaningfully between groups.

Interpretation

The key result is durability.

The original clinical question was whether postmastectomy radiotherapy could be shortened from 5 weeks to 3 weeks without compromising locoregional control.

At 5 years, the randomized trial established noninferiority.

More than a decade later, there is no evidence that an oncologic penalty is emerging with longer follow-up.

The 10-year locoregional recurrence rates remain close at 10.3% and 12.0%, while disease-free and overall survival are also similar.

This matters because late recurrences and late radiation toxicity are particularly relevant in breast cancer. A shorter regimen becomes much more convincing when its safety and disease-control profile remain stable beyond the initial 5-year analysis.

The toxicity findings require some nuance.

Hypofractionation was associated with more grade 1-2 pulmonary fibrosis. That difference is statistically significant and should not be ignored.

However, there were no grade 4-5 late events and no brachial plexopathy, suggesting that the shorter regimen did not generate an important severe late-toxicity signal during long-term follow-up.

The result also fits within a broader randomized evidence base supporting 3-week postmastectomy schedules with regional nodal irradiation.

For clinical practice, the main message is increasingly difficult to argue against: extending postmastectomy treatment to 5 weeks is not necessary simply because regional nodes are included.

Limitations

The 10-year analysis was descriptive. Noninferiority was formally established for the original 5-year primary endpoint, not re-tested as a separate 10-year noninferiority endpoint.

The trial was conducted at a single center in China.

Treatment also differs from some contemporary regional nodal irradiation approaches. Internal mammary nodes were not routinely included, and the trial predates widespread modern techniques used in many current breast RT programs.

The population was predominantly stage III and relatively young, which should be considered when extrapolating the results to lower-risk postmastectomy populations.

The increased rate of low-grade pulmonary fibrosis with hypofractionation also remains a relevant late-toxicity finding.

Bottom line

The CHN HYPOPMRT 10-year update reinforces the durability of 3-week postmastectomy radiotherapy for high-risk breast cancer. Locoregional recurrence remained similar to conventional 5-week treatment, survival outcomes were comparable, and no severe late-toxicity signal emerged. The main tradeoff was more low-grade pulmonary fibrosis, but the overall long-term data strongly support 15-fraction PMRT as a durable alternative to conventional fractionation.
Browse more research Suggest a correction