Treatment failure

Abstract

A number of drugs for the treatment of osteoporosis have been developed in the last four decades. Calcitonin was the first available agent specifically developed for osteoporosis treatment and widely employed. Other options like hormone replacement therapy (both estrogens and androgens) had been used before but sporadically and in isolated trials. The effects on bone mineral density, as the first radiological and isotopic methods became available, were used for monitoring drug response. However, as research in the field progressed and new drugs were incorporated into the clinical armamentarium, the evaluation of the efficacy of these treatments became increasingly complex. A hallmark in the evaluation of the efficacy of the treatments for osteoporosis was the design of the trials with alendronate, and Gideon Rodan was the key figure that settled up what still today are the standard criteria for measuring the effect of drug interventions in osteoporosis. The design of the FIT trial established the reduction of fracture incidence as the main outcome for the first time. This criterion required a large sample size and a long period of follow-up to achieve enough statistical power for detecting differences between arms. Moreover, like in other areas of medicine, FIT focused on a final endpoint representing the pathological event in clinics, fracture; additionally, intermediate endpoints, basically bone mineral density (BMD) variation and bone turnover markers (BTM), were used as surrogates of the efficacy. This scheme has been persisting, and therefore, efficacy estimates have relied in these three pillars. Regulatory agencies and evidence evaluation have been working on these grounds. Pivotal trials have been carried out in a highly controlled setting, with excellent follow-up and adherence, in populations with selected clinical profile where strict inclusion and exclusion criteria were applied. However, for the practicing clinician using these drugs, two problems immediately came out. First, patients attending clinics in real-life environment often do not match with the ideal patients included in clinical trials, raising potential issues of applicability (external validity) of the results to the “common patient.” It has been estimated that up to 80 % of patients attending an osteoporosis clinics would have been excluded from participation in the pivotal trials. Second, adherence to osteoporosis medications is poor, and this is associated with a reduction in effectiveness. Therefore, a gap between evidence gathered under ideal conditions and daily practice was an element of uncertainty for the clinician. Furthermore, even under ideal controlled conditions, some patients in the pivotal trials still suffered fractures in the treatment arm. The consequence for the clinician is the lack of firm basis for making decisions in the individual patient.