Lung dosimetry is correlated with radiation pneumonitis (RP) in lymphoma patients treated with mediastinal intensity modulated radiation therapy (IMRT). Furthermore, salvage chemotherapy in the refractory setting and the inflammatory milieu may influence toxicity and outcomes. We hypothesized that implementation of strict dosimetric lung constraints (DLCs) would lower the RP rate compared to historical data. We also examined the relationship between pre-IMRT neutrophil-lymphocyte ratio (NLR) and symptomatic RP. We retrospectively reviewed 261 patients with mediastinal lymphoma treated with IMRT between 2009 and 2018; 190 and 71 patients were treated, respectively before and after implementation of DLCs- mean lung dose (MLD) ≤ 13.5 and total lung irradiated to > 5 Gy (V5) ≤ 55%. We compared symptomatic RP incidence before and after DLC implementation and analyzed the clinical, treatment, and dosimetric factors associated with RP. We compared pre-IMRT NLR in patients who did and did not develop RP. Overall, median age was 33 years, 54.4% were female and 82% had stage I/II lymphoma. Also, 75% of patients (n=195) received IMRT after frontline chemotherapy (median dose 30.6 Gy); 25% of patients (n=66) received IMRT after salvage chemotherapy for refractory disease (median dose 39.8 Gy). The RP incidence was lower in patients treated after DLC implementation (5/71) compared to those treated before (27/190) (7.0% vs 14.2%, p=0.12). In patients treated after DLC implementation, 97% (n=69) had MLD ≤ 13.5 Gy and 91% (n=64) had V5 ≤ 55 %. For patients who received IMRT after frontline chemotherapy, RP incidence was significantly lower after implementation of the DLCs (0/49) compared to those treated before (16/146) (0 vs 11%, p=0.02). While the lung V10, V15, V20, and V25 were significantly lower (p<0.05) for patients treated for refractory lymphoma after DLC implementation (n=22) compared to those treated before (n= 44), the incidence of RP was similar in refractory patients treated before (11/44, 25%) and after (5/22, 22.7%) DLC implementation (p>0.05). In the 66 patients who received prior salvage chemotherapy, mean pre-IMRT NLR was significantly lower in patients who developed RP than those who did not (2.04 vs 5.42, p = 0.02). After institutional implementation of MLD ≤ 13.5 Gy and V5 ≤ 55 %, RP incidence was reduced in patients receiving IMRT after frontline chemotherapy, with no patients experiencing RP. However, RP incidence was similar before and after employment of DLCs in patients treated with salvage chemotherapy prior to RT. Thus, stricter DLCs and inflammatory factors may influence RP risk in the salvage population despite lower lung doses after DLC implementation. NLR should be further evaluated as a potential factor associated with RP in lymphoma patients treated with IMRT for refractory disease.