The study population consisted of 21 patients with oligometastatic cancer (1–5 sites of metastatic cancer measuring ≤6 cm) from various primary tumors enrolled on a phase I/II trial of sunitinib and radiotherapy between May 2007 and September 2010 with clinical follow-up through July 2012 (NCT00463060). The institutional review board of Mount Sinai School of Medicine (New York, NY, USA) approved this study and all participants provided written informed consent. For all patients, chemotherapy was discontinued ≥21 days prior to initiating protocol therapy. Sunitinib (25–50 mg qd) was administered on days 1–28. Radiotherapy was delivered to all clinically apparent sites of disease with a margin of 5–10 mm using an image-guided technique described elsewhere (10). The radiation doses were 40 or 50 Gy in 10 fractions from days 8–19. Further systemic therapy consisting of either maintenance sunitinib or chemotherapy was administered starting on day 42 at the discretion of the treating medical oncologist.

Stratifying for the location and volume of bone marrow irradiated, a matched-pair cohort analysis was performed on a contemporary cohort of 21 patients with metastatic cancer who were treated with radiation alone. These patients underwent complete blood counts immediately prior to the start and following the completion of radiotherapy. Baseline patient and treatment characteristics for the two groups are described in Table I.

Complete blood count, differential and platelet count were obtained prior to treatment and on days 8 and 19. For each patient, pretreatment values served as controls for blood levels on days 8 and 19. In the control group, complete blood counts were obtained immediately prior to the start and following the completion of radiotherapy. All hematological measurements were taken at the Mount Sinai School of Medicine laboratories using the Coulter LH 750 Hematology Analyzer (Beckman Coulter, Inc., Brea, CA, USA), which generates a five-part differential using flow cytometry technology based on volumetric impedance (using direct current) and conductivity (using high frequency electromagnetic energy and laser light scatter). This methodology has been demonstrated to correlate well with manual slide review (14).

As patients in the clinical trial required radiation to treatment sites throughout the body, the volume of bone marrow irradiated was quantified. To standardize measurement of bone marrow, volumes were calculated using the Eclipse Treatment Planning System 8.0 (Varian Medical Systems, Inc., Palo Alto, CA, USA) using a commercially available automatic segmentation algorithm. To account for an uneven distribution of active bone marrow throughout the body, the volume of bone marrow irradiated was estimated by applying a correction factor that was derived by cross referencing volumes of bone mass against anatomically measured volumes of marrow contained within specific bones (15). Specifically, the correction factors were 1.0 for lumbosacral spine; 0.65 for thoracic spine; 0.5 for cervical spine, rib, sternum, scapula and iliac bone; and 0.3 for humerus, mandible, femoral head and femoral neck.

A two-sided paired t-test was used to make comparisons between pre- and post-treatment hematological measurements. Microsoft Excel was used to plot the differences in changes in blood counts over time. Progression-free survival and overall survival rates were calculated using the Kaplan-Meier method on Stata software version 13.1 (StataCorp LP, College Station, TX, USA). Univariate comparison between groups was performed using the log-rank test.

The primary tumor types for patients treated with sunitinib included lung (24%), kidney (14%), liver (14%), head and neck squamous cell carcinoma (14%), prostate (14%), pancreas (5%), breast (5%), melanoma (5%) and colorectal (5%). Using pretreatment values as controls, significant changes in hematological measurements were observed within 7 days of initiating sunitinib (Fig. 1 and Table II). On average, there was a 13% reduction in leukocytes (6.72×10⁹ vs. 5.88×10⁹/l, P=0.02), largely due to a 30% decrease in monocytes (0.50×10⁹ vs. 0.35×10⁹/l, P<0.01) and an 18% decrease neutrophils (4.91×10⁹ vs. 4.01×10⁹/l, P<0.01). Sunitinib resulted in a marginal increase in mean hemoglobin levels (13.2 vs. 13.9 g/dl, P<0.01). No patients had experienced hematological toxicity of grade 3 or above after 7 days of sunitinib treatment.

There were no statistically significant differences in baseline blood counts when comparing the radiation alone and combined sunitinib and radiation cohorts. Following radiation therapy alone, the only detectable difference was a 39% decrease in the mean lymphocyte count compared with baseline (1.15×10⁹ vs. 0.70×10⁹/l, P<0.01). Grade 2–4 lymphopenia was observed in 76% of patients; this comprised 19% with grade 2, 52% with grade 3 and 5% with grade 4.

Hematological measurements on day 19 demonstrated significant changes compared to pretreatment levels (Fig. 1 and Table II). On average, there was a 40% reduction in total leukocytes (6.72×10⁹ vs. 4.03×10⁹/l, P<0.01), which included a 38% decrease in monocytes (0.50×10⁹ vs. 0.31×10⁹/l, P<0.01), 43% reduction in lymphocytes (1.15×10⁹ vs. 0.65×10⁹/l, P<0.01) and 39% decline in neutrophils (4.91×10⁹ vs. 2.99×10⁹/l, P<0.01). There was also a 26% reduction in the mean platelet count (196.2×10⁹ vs. 145.9×10⁹/l, P=0.006).

Compared to sunitinib alone, combined sunitinib and radiation demonstrated a further decrease in mean leukocyte (P<0.01), platelet (P<0.01), neutrophil (P=0.04) and lymphocyte counts (P<0.01), but no significant change in mean hemoglobin level (P=0.07) or monocyte count (P=0.09). The incidence rates of grade 3–4 leukopenia, neutropenia, thrombocytopenia and lymphopenia were 19, 5, 10 and 57%, respectively.

Compared to radiation alone, patients treated with concurrent sunitinib and radiation exhibited a 40% decrease in mean leukocyte counts (P=0.01) and a 26% reduction in mean platelet counts (P<0.01). The effect on leukocytes was largely due to a 39% decline in mean neutrophil counts (P=0.03) and a 39% reduction in mean monocyte counts (P<0.01). On average, there was no effect on lymphocytes (P=0.1) and mild erythrocytosis (P=0.06) that failed to reach statistical significance.

In a recent analysis of patients with hepatocellular carcinoma treated with sunitinib, a greater decrease in monocyte count after 14 days of treatment significantly predicted progression-free and overall survival times (16). In the current study, patients with a greater than average decline (≥200/µl) in absolute monocyte count following sunitinib and radiation therapy had significantly improved rates of 2-year progression-free (54% vs. 0%, P=0.03) and overall survival (60% vs. 17%, P=0.05). Progression-free and overall survival curves are shown in Fig. 2. Due to limited sample size and the hypothesis-generating nature of the analysis, multivariable analysis was not attempted.