Background: Fewer than 20% of patients with bone cancer who are treated with surgery alone are cured. Even with the best current treatment, surgery combined with chemotherapy, only 60%-80% of patients with nonmetastatic bone cancer and 10% of patients with metastatic bone cancer are cured. Thus far, the reason for treatment failure in the nonresponding subset has not been identified. It has been hypothesized that P-glycoprotein, which confers multidrug resistance, might be the cause. We sought to determine whether the expression of P-glycoprotein is associated with poor treatment outcome in osteosarcoma.
Methods: In a retrospective study, we correlated P-glycoprotein expression with the outcome of conventional chemotherapy in 62 consecutive, clinically staged patients diagnosed as having osteosarcoma between 1980 and 1989.
Results: P-glycoprotein was overexpressed in 27 patients but not in another 34 patients, and expression was ambiguous in the sample from one patient. At a median follow-up of 8.9 years, the 34 patients whose tumors did not express P-glycoprotein had significantly better relapse-free rates than the 27 subjects whose tumors expressed the protein (87% versus 0%; P<.00001) and had improved survival rates (94% versus 35%; P<.00001). Among the 46 patients who received chemotherapy before surgery, the 23 whose tumors were negative for P-glycoprotein showed significantly better long-term outcomes (P<.00002), although differences in tumor necrosis in response to therapy were only of borderline significance (P = .057).
Conclusions: P-glycoprotein expression does correlate with treatment failure in patients with osteosarcoma. This correlation raises the possibility that inhibiting the action of P-glycoprotein as part of therapy for this disease would improve outcome.