Abstract

The exclusion of bone metastases is important in the initial staging of non-small cell lung cancer, though there is debate about whether bone scans should be performed routinely or restricted to patients who present with clinical or laboratory indicators suggesting skeletal metastases. In a prospective study of 110 consecutive patients referred for initial staging of non-small cell lung cancer, we assessed the sensitivity of a group of clinical indicators (chest pain, skeletal pain, bone tenderness on physical examination, serum alkaline phosphatase, and serum calcium) for the presence of skeletal metastases as determined by bone scanning. The final staging result was validated with follow up data over at least three years. At the initial staging 37 of 110 bone scans (34%) showed areas of increased uptake, of which only nine were confirmed to be metastases (by tomography, computed tomography, or biopsy). Half the patients (55) had at least one clinical indicator suggesting skeletal metastases, including all patients with proved skeletal metastases. Thus the sensitivity of these clinical indicators was 100% and the specificity 54%. Within one year three of 27 patients with non-confirmed positive bone scans had skeletal metastases, one of which was in the area that had shown increased uptake initially. All these patients had clinical indicators for skeletal metastases and all had inoperable advanced tumours. Four of 69 patients with an initially negative bone scan developed skeletal metastases within one year. It is concluded that in non-small cell lung cancer bone scanning can be restricted to patients with clinical indicators for skeletal metastases. This approach reduces the number of bone scans and consecutive investigations without loss of sensitivity in the detection of skeletal metastases.