Radiation therapy in the management of symptomatic bone metastases: the effect of total dose and histology on pain relief and response duration

Purpose: In order to better define variables and factors that may influence the pain response to radiation, and to look for a radiation regimen that can assure the highest percentage and the longest duration of pain relief, we performed a prospective, although not randomized, study on patients with bone metastases from various primary sites. Methods and Materials: From December 1988 to March 1994, 205 patients with a total of 255 solitary or multiple bone metastases from several primary tumors were treated in our radiotherapy center with palliative intent. Irradiation fields were treated with three main fractionation schedules: (1) Conventional fractionation: 40–46 Gy/20–23 fractions in 5–5.5 weeks; (2) Short course: 30–36 Gy/10–12 fractions in 2–2.3 weeks; (3) Fast course: 8–28 Gy/1–4 consecutive fractions. Pain intensity was self-assessed by patients using a visual analogic scale graduated from 0 (no pain) to 10 (the strongest pain one can experience). Analgesic requirement was assessed by using a five-point scale, scoring both analgesic strength and frequency (0 = no drug or occasional nonopioids; 1 = Nonopioids once daily; 2 = Nonopioids more than once daily; 3 = Mild opioids (oral codeine, pentazocine, etc.), once daily; 4 = Mild opioids more than once daily; 5 = Strong opioids (morphine, meperidine, etc.). Complete pain relief meant the achievement of a score ≤ 2 in the pain scale or 0 in the analgesic requirement scale. Partial pain relief indicated a score of 3 to 4 or of 1 to 2 on the former and latter scale, respectively. Results: Total pain relief (complete + partial) was observed in 195 (76%) sites, in 158 of which (62%) a complete response was obtained. Metastases from NSC lung tumors appeared to be the least responsive among all primary tumors, with 46% complete pain relief in comparison to 65% and 83% complete relief in breast (p = 0.04) and in prostate metastases (p = 0.002), respectively. A significant difference in pain relief was detected among the several ranges of total dose delivered to the painful metastases, with 81%, 65%, and 46% complete relief rates in the 40–46 Gy, 30–36 Gy (p = 0.03), and 8–28 Gy (p = 0.0001) dose ranges respectively. A straight correlation between total dose and complete pain relief was confirmed by the curve calculated by the logistic model which shows that doses of 30 Gy or more are necessary to achieve complete pain relief in 70% or more of bone metastases. This correlation holds also for the duration of pa