NONINVASIVE PREDICTION OF ALUMINUM BONE-DISEASE IN HEMOTONEAL AND PERITONEAL-DIALYSIS PATIENTS

Between October 1987 and October of 1989, we conducted a prospective study to evaluate non-invasive test strategies for predicting aluminum bone disease (ABD) in a group of largely unselected dialysis patients based on their deferoxamine (DFO) test alone, or the combined results of their DFO test and intact 1-84 parathyroid hormone (PTH) levels. These test parameters were evaluated against the pathological diagnosis of ABD based on bone biopsy ("gold standard"). A total of 445 patients in three dialysis centers in Toronto were serially followed for their clinical, laboratory and risk parameters for renal osteodystrophy during the study, and 259 (142 PD and 117 HD) patients underwent a series of investigations which included the DFO test, measurement of intact 1-84 PTH levels, and an iliac crest bone biopsy. Serum aluminum ([Al]) level greater-than-or-equal-to 3700 nM (or 100-mu-g/liter) had a positive predictive value (PPV) of 75% for ABD in our PD and 88% in our HD patients, but its sensitivity was low (10 and 37%). Delta [Al] (that is, incremental rise of serum [Al] from baseline post-DFO) was useful in predicting ABD in our PD but not HD patients. Test combination based on delta [Al] greater-than-or-equal-to 5550 nM (or 150-mu-g/liter) and PTH level < 20 pM (or 200 pg/ml) yielded the best PPV greater-than-or-equal-to 95% for ABD in both PD and HD patients. This test cut-off would remain highly predictive of ABD even if the prevalence of ABD decreases to as low as 5% for the PD patients and 10% for the HD patients. However, for patients who have discontinued their aluminum gels for greater-than-or-equal-to 6 months these test strategies would not be useful due to a very high frequency of false-negative cases (or low sensitivity). Thus, the main utility of these non-invasive strategies will be in those patients who continue to require treatment of aluminum gels for their phosphate control.