Why this matters
Total neoadjuvant therapy has become an important strategy for locally advanced rectal cancer, but the optimal radiotherapy backbone remains debated.
Short-course radiotherapy followed by consolidation chemotherapy offers major logistical advantages and increases systemic therapy exposure before surgery. Long-course chemoradiotherapy, however, remains a well-established approach with strong local-control data.
This randomized phase III study directly compared these strategies in a predominantly stage III population.
The result is more nuanced than the abstract initially suggested. Short-course TNT produced a numerically higher pathological complete response rate, but the difference was not statistically significant, and no disease-free, locoregional, or overall survival advantage was demonstrated.
Study design
This was a single-institution randomized phase III trial enrolling patients with cT3-4 or node-positive rectal adenocarcinoma and ECOG performance status 0-1.
Patients were randomized between:
- long-course chemoradiotherapy followed by total mesorectal excision
- short-course radiotherapy followed by consolidation CAPOX and total mesorectal excision
The long-course regimen consisted of:
- 45-50.4 Gy in 25-28 fractions
- concurrent capecitabine
- subsequent surgery
The short-course TNT regimen consisted of:
- 25 Gy in 5 fractions over 1 week
- 6 cycles of CAPOX
- subsequent surgery
The primary endpoint was pathological complete response.
Secondary endpoints included:
- disease-free survival
- locoregional failure-free survival
- overall survival
- resection outcomes
- treatment toxicity
The updated presentation included 113 patients:
- 59 in the short-course TNT arm
- 54 in the long-course CRT arm
A total of 74 patients underwent resection:
- 39 after short-course TNT
- 35 after long-course CRT
Median follow-up was 28 months.
Pathological response
Among patients who underwent surgery, pathological complete response was:
- 30.8% with short-course TNT
- 20.0% with long-course CRT
- p=0.290
The absolute difference was therefore approximately 11 percentage points, but the trial did not demonstrate a statistically significant improvement in pCR.
The short-course TNT result is consistent with pCR rates reported in several international TNT studies, but it should not be interpreted as proof that this regimen produces superior pathological response.
Disease-free survival
Disease-free survival showed no statistically significant difference between treatment strategies.
At 3 years:
- DFS: 59.0% with short-course TNT
- DFS: 72.1% with long-course CRT
The estimated treatment effect was:
- HR 1.67
- 95% CI 0.89-3.15
- p=0.102
Median DFS was approximately:
- 40 months with short-course TNT
- not reached with long-course CRT
The difference was not statistically significant, but numerically favored long-course chemoradiotherapy.
Distant metastasis was the predominant pattern of recurrence.
Locoregional control
Three-year locoregional failure-free survival was:
- 82.8% with short-course TNT
- 92.2% with long-course CRT
The estimated HR was:
- HR 2.08
- 95% CI 0.64-6.77
- p=0.213
Again, this difference was not statistically significant.
However, an exploratory analysis restricted to the 74 patients who underwent surgery showed a concerning difference in locoregional failure:
- 23.1% after short-course TNT
- 2.9% after long-course CRT
- p=0.0157
Because this was a subgroup analysis in a small cohort, it should not be treated as definitive evidence that short-course TNT compromises local control.
But the magnitude of the difference is large enough that it should not be ignored and deserves confirmation with longer follow-up.
Overall survival
There was no statistically significant overall survival difference.
At 3 years:
- OS: 67.4% with short-course TNT
- OS: 74.8% with long-course CRT
The estimated HR was:
- HR 1.50
- 95% CI 0.74-3.03
- p=0.253
Median OS was approximately:
- 42 months with short-course TNT
- not reached with long-course CRT
As with DFS and locoregional control, the numerical direction favored long-course treatment, but the confidence interval was wide and the difference was not statistically significant.
Interpretation
The most important point from this trial is that a higher pCR rate does not necessarily translate into better clinical outcomes.
Short-course TNT produced a numerically higher pCR rate of 30.8% compared with 20.0% after long-course chemoradiotherapy.
But pCR was not significantly different.
More importantly, the survival endpoints did not favor short-course TNT.
Three-year DFS, locoregional failure-free survival, and overall survival were all numerically lower with short-course TNT, although none of these differences reached statistical significance.
That creates an important contrast between pathological response and clinical outcomes.
A regimen can produce more complete pathological responses without necessarily improving recurrence or survival across the entire randomized population.
The locoregional failure signal in the resected subgroup deserves particular attention. A 23.1% failure rate after short-course TNT compared with 2.9% after long-course CRT is difficult to dismiss completely, even though the analysis is exploratory and based on a small number of patients.
This should not be interpreted as evidence that short-course TNT is inferior.
The study is small, follow-up remains relatively short, and confidence intervals around the survival estimates are wide.
But the data also do not support claiming a survival advantage for short-course TNT.
The most defensible interpretation is that short-course TNT achieved a numerically higher pCR rate while long-course CRT maintained at least comparable, and numerically favorable, disease-control outcomes in this trial.
Limitations
This was a single-institution study with a relatively small sample size.
Only 113 patients were enrolled, and the pathological response analysis included 74 resected patients.
Median follow-up was only 28 months, which limits interpretation of long-term rectal cancer outcomes.
The pCR difference was not statistically significant.
None of the DFS, LRFS, or OS comparisons reached statistical significance, and confidence intervals were wide.
The significant locoregional failure difference in the resected population arose from a subgroup analysis and requires independent validation.
Finally, the comparison was not simply between two radiotherapy schedules. The short-course arm incorporated substantial consolidation chemotherapy, while the long-course arm used concurrent capecitabine without the same TNT structure.