Between October 1994 and December 2004, a total 60 patients were referred for palliative radiation therapy for bone metastasis from HCC. Fifty-three patients completed planned radiation therapy for 84 lesions. The remaining seven patients stopped treatment because of either their own refusal or deterioration of performance status regardless of treatment. The median age of all patients was 53 years-old (range 33 to 70 years). With respect to performance status, 13 patients were scored as ECOG 1, 28 patients were scored as ECOG 2, and the remaining 12 patients were scored as ECOG 3. Thirty-seven patients had a single metastatic site, while 16 patients had multiple lesions. Twenty-three patients presented with synchronous bony metastasis with a diagnosis of primary HCC. The median interval between the day of the initial HCC diagnosis and that of the metachronous bone metastasis was 8.5 months, ranging from 2 months to 70 months (Table 1).

Of the 84 lesions treated, 48 lesions were examined with CT/MRI. Among them, 41 lesions were also examined with a bone scan. Thirty-one lesions were examined with a bone scan only. The remaining 5 lesions, despite of a lack of evaluation with either a CT/MRI or bone scan, revealed abnormal findings on a plain x-ray as symptomatic palpable masses.

Radiation therapy was delivered with a linear accelerator, utilizing 6 or 15 MV photon beams or electrons, according to the depth of the lesion. The treatment volume covered the gross tumor volume (GTV), which was identified with imaging data, adding a margin of 1.5 to 2 cm in all directions. If vertebral bodies were targeted for treatment, the treatment volume included the vertebral bodies, plus one additional segment in both the cephalad and caudal directions. The median total dose was 30 Gy, ranging from 7 Gy to 40 Gy. The daily fraction size was 2.5 or 3 Gy, with the exception of a single fraction treatment of 7 Gy.

We focused on the comparison of the size of the radiation field as determined with a bone scan only to the field size determined by both a bone scan and CT or MRI. We evaluated the need of adjustment of the radiation field according to the imaging modality.

After completion of radiation therapy, all patients received follow-up by either a hospital visit or phone call. The median period of follow-up was 6 months. Patient data was analyzed using the Statistical Package for the Social Science (SPSS, Chicago, IL USA) software package. The Kaplan-Meier method was also used. The end-point was the day of the patient death, both from the day of the initial diagnosis of primary disease and from the day of the initial diagnosis of a bone metastasis.

Total 84 metastatic sites received radiation treatment. The C-T-L spines were the most common site (40 lesions), followed by the pelvic bone, including the hip joints (19 lesions), ribs (9 lesions), upper and lower extremities (7 lesions), shoulder joint (4 lesions), skull (3 lesions), and sternum (2 lesions).

Among the 48 lesions evaluated by either CT or MRI, 41 lesions showed bone destruction combined with soft tissue formation (Fig. l). There were additional 5 lesions with soft tissue formation, which were not examined with CT or MRI, but presented with clinically palpable soft tissue at the site of bone metastasis as observed on plain x-rays.

Metastatic lesions at the vertebrae were the most frequent ones with both bone destruction and soft tissue formation (27/29). Three soft tissue lesions were expressed as a cold spot (or as a photon defect) rather than with hot uptake as seen on the bone scan. However, there was no way to detect existing soft tissue formation around metastatic bone in lesions evaluated with a bone scan only (Table 2).

Forty-one lesions were examined with both a bone scan and by CT/MRI. We compared the radiation field size based on a bone scan image to the field size based on a CT/MRI image. Thirty-one lesions required an adjustment of the radiation field size when the radiation field size was determined with CT/MRI findings rather than with bone scan findings (Fig. 2). These lesions showed characteristic destruction of the bony framework and infiltration into the surrounding muscle or fat tissue extensively. The extent of infiltration into the surrounding muscle or fat tissue determined the size of modification of the radiation field. The amount of radiation field adjustment was an increment of an average 2.5 cm in length (range, 1~7 cm).

At the time of analysis, we identified 50 patient deaths. Three patients were lost to follow-up. The median survival period of all patients from the day of the initial HCC diagnosis was 8 months (ranging from 2 to 71 months). The survival rate was 53.9%, 38.6% and 30.1% at 1, 2 and 3 years, respectively. From the time of the initial diagnosis of bone metastasis, the overall survival rate was 8.7% at 1 year, with a median survival period of 5 months (ranging from 1 to 38 months). Only one patient survived for more than 2 years. This male patient was diagnosed with a bone metastasis 6 years after the initial diagnosis of primary HCC. The median survival period of patients diagnosed with primary HCC and bone metastasis synchronously and metachronously were nearly the same-5 months and 4.5 months, respectively (ranging from 2 to 38 months, and 1 to 21 months, respectively). We evaluated the influence of prognostic factors by use of the log-rank test. By univariate analysis, the presence of metastasis to other solid organs was the only significant prognostic factor. The 1-year survival rate of those patients that had another solid organ metastasis at the time of the diagnosis of a bone metastasis was 0%, with a median survival period of 3 months, but the 1-year survival rate of patients that had only a bone metastasis was 14.5% with a median survival period of 8 months (p=0.03). However, age, sex, performance status, primary disease status and multiplicity of bony metastasis did not have an impact on patient prognosis.