Dosimetrically Correlating the Acute Effect of Bone Radiation on Anemia in Men with Intact Prostate Cancer

Abstract

<h3>Purpose/Objective(s)</h3> Published data evaluating the effect of radiation (RT) on pelvic bone marrow is mainly based on women with gynecological cancers receiving concurrent chemotherapy. The effect of RT on bone marrow in men with prostate cancer (PC) is less clear, where chemotherapy is not routinely delivered. This study aimed to determine if there is a dosimetric correlation between the volume of pelvic bone irradiated when treating PC and reduced hemoglobin (Hgb) values. Impaired marrow function affects patient quality of life and future systemic therapy. The a priori hypothesis was that volume of pelvic bone receiving at least 4 Gy (V4Gycc) would correlate significantly with marrow toxicity, given that 4Gy is the LD50 dose for a total body exposure. <h3>Materials/Methods</h3> A prospective database of PC patients treated at one institution was utilized. Non-metastatic patients treated between 2008 and 2021 with definitive RT were selected. Patients were included if they had at least one Hgb value within two years prior to their RT start date and at least one Hgb value within three years of RT completion. Patients with confounding diagnoses were excluded and were censored at the time of a second oncologic treatment course or confounding event known to lower Hgb levels. All patient charts were individually audited for accuracy. Pelvic bone (L3/L4 interface through ischial tuberosities) was contoured, and the dose to this structure was calculated. Doses were converted to 2 Gy equivalent doses (EQD2) using an α/β of 10. Exploratory analysis suggested dichotomizing into low volume exposures (LVE; V4Gycc <=1000cc) and high-volume exposures (HVE, V4Gycc >1000cc). Non-parametric kernel regressions evaluating the effects of time, V4Gycc, and androgen deprivation therapy (ADT) use on Hgb values was performed. Percent reduction in Hgb from baseline was evaluated. <h3>Results</h3> 222 patients were included in final analysis. Median starting Hgb was 15.0 g/dL (IQR 14.2-15.9). The mean V4Gycc was 1283cc (IQR 1740-2044). Hgb nadir occurred around day 90 post-RT start. Both HVE and LVE groups had significantly lower Hgb as a percentage of baseline at day 90 (HVE= 85.4%, p < 0.001; LVE=94.8%, p < 0.001). HVE resulted in significantly lower Hgb levels compared to LVE at day 90 (95% confidence intervals; 80.3% - 88.9% vs 90.9% - 98.3%, p < 0.001). ADT use had no effect on Hgb levels (p = 0.45). Kernel-weighted local polynomial smoothing implies HVE also have impaired marrow recovery compared to V4Gycc < 1000. <h3>Conclusion</h3> In prostate cancer treatment, a V4Gycc of bone marrow >1000 results in significantly lower Hgb nadir as a percentage of baseline than patients who received V4Gycc < 1000. Our model predicts that if a patient receives V4Gycc to pelvic bone marrow >1000, they will become mildly anemic with a baseline Hgb of 15.2, moderately anemic with a baseline Hgb of 12.8, and severely anemic with a baseline Hgb of 9.3. Limiting pelvic bone marrow to <1000cc receiving ≥4Gy EQD2 is a proposed dose constraint to limit acute marrow toxicity.