30-Second Takeaway
- Adjuvant breast radiotherapy did not increase overall skin cancer incidence in a large Korean cohort.
- Antibiotic exposure around ICI therapy in NSCLC is linked to worse survival in observational data but not clearly causal.
Week ending June 13, 2026
Five recent papers with immediate relevance to radiation oncologists
Adjuvant RT after breast surgery was not associated with higher skin cancer incidence in Korea.
In a nationwide retrospective cohort of 37,957 breast cancer patients with 1:1 propensity matching (9,928 per group), overall skin cancer rates were similar after adjuvant RT (0.97% vs 1.02%, p=0.720). No significant association was observed between RT and malignant melanoma (p=0.094) or non-melanoma skin cancer (p=0.196). Older age, mole presence (HR 5.254), pre-malignant lesions (HR 12.905), and lymphedema (HR 1.978) predicted higher skin cancer risk. Clinical implication: focus surveillance on dermatologic risk factors and lymphedema rather than attributing skin cancer risk to modern adjuvant RT.
Multidisciplinary lifestyle program during radiotherapy improved body composition, fitness, and symptoms.
In a single-center RCT of 95 women with stage II/III breast cancer, the CompLife program reduced visceral adipose tissue at 3 months (118 vs 141 cm2; p=0.044). CompLife improved VO2 max at 3 months (23.4 vs 18.8 mL/kg/min; p=0.025) and 6 months, and produced better physical well-being scores over 12 months. Participants reported fewer symptoms at 12 months (MDASI 1.3 vs 2.5; p=0.014) and higher fiber intake throughout follow-up. Interpretation: consider integrating structured nutrition and exercise during RT to improve fitness and symptoms, noting multiple secondary outcomes.
On-premises LLMs accurately extracted prognostic data from head and neck reports with clinician adjudication.
Locally deployed Llama3.3-70B achieved 98.6% F1 extracting 30 prognostic fields from 1,360 reports, with processing time ~53 s/report in zero-shot prompting. LLM-extracted HPV status, smoking, and Charlson score added prognostic value to a Cox model (DFS p=0.014; ΔC-index +0.071) after bootstrap validation. This clinician-AI collaborative workflow enables privacy-preserving tumor board support and longitudinal data curation. Caveat: performance and integration require local validation and clinician review before routine clinical deployment.
References
Numbered in order of appearance. Click any reference to view details.
Additional Reads
Optional additional studies from this edition.