Post-Chemotherapy Brain Irradiation in Primary CNS Lymphomas: Impact on Survival of Different Fields and Doses.

High-dose methotrexate (HD-MTX)-based chemotherapy (CHT) followed by whole-brain irradiation (WBRT), the conventional approach to primary CNS lymphomas (PCNSL), is associated with relevant risk of late neurotoxicity. To avoid or reduce irradiation volumes and doses have been proposed to minimize this complication, but no studies focused on RT parameters exist and the best RT schedule remains to be defined.

The impact on outcome of different RT fields and doses was assessed in a mono-institutional series of 85 HIV-negative patients (pts) with PCNSL treated with upfront CHT containing MTX 3.5 g/m² every 3 weeks, alone or in combination, followed by RT. All pts were irradiated to the WB with photons of 6 MeV and standard fractionation, including the first two cervical vertebras and the posterior two thirds of the orbits; a tumor bed (TB) boost (2 to 4 isocentric fields) to deliver higher doses in the tumor area but limiting the WBRT dose was indicated in 46 pts. In the case of multifocal lesions, the boost volume included each single lesion. Radiation doses varied in the years, oscillating between 30 and 45 Gy to the WB and between 36 and 54 Gy to the TB. Pts were stratified according to the response to CHT for analysis of WB and TB doses.

Response after CHT was complete (CR) in 37 (44%) pts and partial (PR) in 17 (20%); 24 pts experienced PD and 7 died of toxicity. The median f-up was 37 months (range 13–105).

Thirty-three of the 37 CRs were referred to consolidation RT: six (30%) of the 20 pts irradiated with 30 – 36 Gy to the WB and six (46%) of the 13 pts irradiated with a WB dose '40 Gy experienced relapse, with a 3-yr FFS of 64±13% and 64±14% (p= 0.31), respectively. Relapse rate was not significantly different between the 19 pts irradiated with a TB dose of 45 – 54 Gy and the 14 pts irradiated with a TB dose of 36 – 44 Gy, with a 3-yr FFS of 62±11% and 68±15% (p= 0.49), respectively. Twenty-one of the 33 pts treated with consolidation RT are alive, with a 3-yr OS of 71±9%. No survival differences were observed in subgroups of pts divided according to the WB or TB doses. Four (19%) survivors exhibited evident neurological impairment after treatment (bradipsychia, memory deterioration and dysphasia); these pts had a median age of 66 years (range 46–72) and were treated with a WBRT dose >40 Gy.

The 17 pts in PR after CHT were referred to complementary RT; ten of them (59%) achieved a CR, while the remaining pts had a residual enhanced image at the post-radiation MRI, with a 3-yr FFS of 48±16% and 0% (p= 0.004), respectively. Median WB and TB doses in pts who achieved CR after RT were 38 Gy and 45 Gy, while median doses in pts who did not achieve a CR were 36 Gy and 38 Gy, respectively. Correspondingly, pts treated with a TB dose of 45 – 54 Gy had a significantly better FFS (3-yr: 34±11% vs. 17±10%; p= 0.05) and OS (3-yr: 44±12% vs. 17±10%; p= 0.01) with respect to pts irradiated with a lower TB dose. Five of the 17 pts treated with complementary RT are alive, with a 3-yr OS of 34±12%. In the subgroup of PR pts, the number of lesions was independently associated with survival, with a 3-yr OS of 63±17% for pts with single lesions and 11±10% for pts with multiple lesions (p= 0.02).

Sixteen of the 24 pts with PD after CHT received immediate salvage RT (3 CR, 4 PR and 9 PD). All responders but one experienced relapse and died, with a 3-yr OS of 7±8%; WB and TB doses higher than 40 Gy and 45 Gy, respectively, had no impact on survival.

In pts with PCNSL, RT parameters should be chosen on the basis of response to primary CHT. Pts in CR after CHT should be treated with WBRT with 30–36 Gy, higher doses are not advisable since they do not change outcome and could increase the risk of severe neurotoxicity. Pts in PR with single lesions should be treated with 40 Gy to the WB with a TB boost of 9–14 Gy. Therapeutic results with high-dose brain irradiation in pts with PR and multiple lesions or PD after primary CHT are disappointing; in these patients the role of additional no cross-resistant CHT before RT should be investigated.

No relevant conflicts of interest to declare.