Risk of subsequent primary cancers after carbon ion radiotherapy, photon radiotherapy, or surgery for localised prostate cancer: a propensity score-weighted, retrospective, cohort study.

BACKGROUND: The risk of subsequent primary cancers in patients with prostate cancer after treatment with photon radiotherapy is small in absolute numbers, but it is higher than that after surgical treatment. Carbon ion radiotherapy has a theoretically lower risk of inducing secondary malignancies than photon radiotherapy, but this risk has not been investigated in practice because of the low number of facilities offering such therapy worldwide and the limited data on long-term follow-up because the therapy has only been available since 1994. We aimed to analyse the risk of subsequent primary cancers after treatment with carbon ion radiotherapy in patients with localised prostate cancer and to compare it with that after photon radiotherapy or surgery in this setting.
METHODS: In this retrospective cohort study, we reviewed records of patients who received carbon ion radiotherapy for prostate cancer between June 27, 1995, and July 10, 2012, at the National Institute of Radiological Sciences (NIRS) in Japan. We also retrieved the records of patients diagnosed and treated for prostate cancer between Jan 1, 1994, and Dec 31, 2012, from the Osaka Cancer Registry. Eligible patients had histologically confirmed localised prostate cancer and a minimum follow-up of at least 3 months; no age restrictions were applied. We excluded patients with metastasis, node-positive disease, or locally invasive (T4 stage) prostate cancer, those with previous or synchronous malignancies, and those who received previous radiotherapy or chemotherapy. We did a multivariable analysis to estimate predictors of subsequent cancers after carbon ion radiotherapy treatment. We also used propensity score inverse probability weighting to retrospectively compare the incidence of subsequent cancers in patients with localised prostate cancer treated with carbon beams, photon radiotherapy, or surgery.
FINDINGS: Of 1580 patients who received carbon radiotherapy for prostate cancer at the NIRS, 1455 (92%) patients met the eligibility criteria. Of 38 594 patients with prostate cancer identified in the Osaka registry, 1983 (5%) patients treated with photon radiotherapy and 5948 (15%) treated with surgery were included. Median follow-up durations were 7·9 years (IQR 5·9-10·0) for patients who received carbon ion radiotherapy (after limiting the database to 10-year maximum follow-up), 5·7 years (4·5-6·4) for patients who received photon radiotherapy, and 6·0 years (5·0-8·6) for those who received surgery. 234 subsequent primary cancers were diagnosed in the carbon ion radiotherapy cohort; some patients developed several tumours. On multivariable analysis, age (p=0·0021 for 71-75 years vs ≤60 years; p=0·012 for >75 years vs ≤60 years) and smoking (p=0·0005) were associated with a higher risk of subsequent primary cancers in patients treated with carbon ion radiotherapy. In the propensity score-weighted analyses, carbon ion radiotherapy was associated with a lower risk of subsequent primary cancers than photon radiotherapy (hazard ratio [HR] 0·81 [95% CI 0·66-0·99]; p=0·038) or surgery (HR 0·80 [0·68-0·95]; p=0·0088), whereas photon radiotherapy was associated with a higher risk of subsequent primary cancers than surgery (HR 1·18 [1·02-1·36]; p=0·029).
INTERPRETATION: Our analysis suggests that patients with localised prostate cancer treated with carbon ion radiotherapy appear to have a lower risk of subsequent primary cancers than those treated with photon radiotherapy. Although prospective evaluation with longer follow-up is warranted to support these results, our data supports a wider adoption of carbon ion radiotherapy for patients with expected long-term overall survival or those with poor outcomes after receiving conventional treatments. FUNDING: Research Project for Heavy Ions at the National Institute of Radiological Sciences (Japan).