Baseline extracellular-vesicle microRNAs were associated with breast radiotherapy toxicity

Baseline miR-144-5p and miR-222-3p differed in patients who later developed greater radiation toxicity, but the signals require validation.

KEY POINTS

  • Pretreatment plasma was analyzed from 89 breast cancer patients enrolled in a prospective biospecimen study. Samples were drawn one day before radiotherapy, after surgery but before radiotherapy-related toxicity developed; no chemotherapy, endocrine or targeted therapy had been given before sampling.
  • Patients were subsequently classified into four toxicity groups: 9 with minimal toxicity, 45 with expected grade 1/mild grade 2 effects, 7 with acute grade 2–3 toxicity, and 28 with late grade 2–3 toxicity including fibrosis, fat necrosis or pigmentation changes.
  • Extracellular vesicles were isolated using size-exclusion chromatography plus precipitation and characterized with protein markers and particle analysis. Of 164 microRNAs screened, 22 showed large between-group expression differences and underwent individual quantitative polymerase-chain-reaction testing; four remained significantly different.
  • miR-144-5p was markedly enriched in both high-toxicity groups: median relative expression was 77.61 in the acute high-toxicity group and 87.64 in the late high-toxicity group, versus 1.14 with minimal toxicity and 2.28 with expected toxicity (p<0.0001).
  • miR-222-3p was specifically associated with acute toxicity: relative expression was 112.09 in the seven acute high-toxicity patients versus 17.63, 29.23 and 20.01 in the minimal, expected and late-toxicity groups, respectively (p<0.05).
  • Conversely, miR-200a-3p and miR-335-3p were enriched in patients with minimal toxicity. miR-335-3p expression was 49.94 in that group versus 18.95, 0.34 and 1.82 in the other groups (p<0.0001).
  • The acute high-toxicity group contained only seven patients, radiation techniques included both external-beam treatment and brachytherapy, and the study did not report a multivariable predictive model, discrimination metrics, prespecified biomarker threshold or independent validation cohort.

CLINICAL TAKEAWAY

A blood test performed before breast radiotherapy that genuinely predicts later normal-tissue sensitivity would be clinically valuable, and these extracellular-vesicle microRNAs are plausible candidates. At present they are discovery biomarkers, not tests that should influence dose, fractionation or prophylactic treatment.

SOURCE

Frontiers in Oncology