NephrologieCHU (A.F., M.E.C.S., R.O., H.M., P.M., G.C.), Amiens, France; and LEGENDO University Hospital (R.B.) Leuven, Belgium
To the editor:
We have read with interest the rapid communication of John et al. (1) showing that the new Scandibodies laboratory immunoradiometric assay (IRMA) for measurement of intact PTH detects full-length human PTH but not amino-terminally truncated fragments, in contrast to most commercially available IRMAs for so-called intact PTH, which measure also 784 PTH as shown by Lepage et al. (2).
In contrast to the Nichols Institute Allegro IRMA for intact PTH (N-IRMA), our PTH IRMA (B-IRMA) (Ref. 3 ; later commercialized by Biosource for the last 10 yr) had the particularity to have a much lower normal range of 340 pg/mL vs. 1065 pg/mL for N-IRMA. The reason for this discrepancy was not clear until 1992 when we compared the recovery of 784 PTH fragments at a wide concentration range in our assay and in that of Nichols Institute. The recovery of fragment (784) was 100% with N-IRMA and less than 1% in B-IRMA. Therefore, the nondetection of 784 was a satisfactory explanation for the lower normal range.
The clinical relevance of such difference is quite obvious for the indirect evaluation of bone remodeling in hemodialysis patients. In 1991, Cohen Solal et al. (4) published in this journal the (184) PTH levels with the B-IRMA of 23 hemodialysis patients for whom a bone biopsy had been performed. This allowed us to classify six of them in the adynamic bone disease (ABD) group [low bone formation rate (BFR) with low osteoid thickness and low osteoblastic and osteoclastic surfaces], eight in the mild lesion group (with normal BFR and mild increase in osteoblastic and osteoclastic surface), and nine in the osteitis fibrosa group (with high BFR and increased osteoblastic and osteoclastic surfaces). All the nine patients with osteitis fibrosa had PTH levels greater than or equal to 69 pg/mL (i.e. 1.7 the upper limit of normal). On the contrary, all the six patients with ABD had PTH levels within the normal range. Four of eight patients with mild lesion had PTH levels within the normal range, three being above and one below. Considering that an intact PTH range associated with normal BFR could define its optimal level in hemodialysis patients, we proposed that their Biosource B-IRMA PTH should be between 40 and 60 pg/mL (i.e. between 1 and 1.5 the upper limit of normal to prevent adynamic bone disease). This proposed optimal range is quite lower than that proposed with N-IRMA not only when expressed in absolute concentrations but also when expressed in folds of the upper limit of normal (ULN) since Quarles et al. (5) in 1992 and Wang et al. (6) in 1995 proposed in hemodialysis patients, 1.52.5 and 24 times the ULN, respectively, the upper threshold being a little higher in CAPD patients according to Wang et al. (Ref. 6 ; 5 times the ULN).
Four hypotheses may be discussed to explain the discrepancy between our proposition for optimal PTH range and those of our American colleagues: 1) the role of a possible 784 PTH fragment accumulation in renal failure; 2) the role of the antagonistic effect of 784 PTH fragments on the bone remodeling effect of endogenous 184 PTH; 3) the role of a difference in aluminum overload leading to a higher bone resistance to endogenous PTH remodeling effect; and 4) the role of calcitriol treatment.
1. Accumulation of 784 in renal failure is a possibility to consider since Brossard et al. (7) showed with high-performance liquid-chromatography determination that the proportion of non-184 PTH fragments in the intact PTH detected by N-IRMA was greater in hemodialyzed patients than in healthy controls whether the patients are in stable condition (50 vs. 21%), have PTH stimulation by hypocalcemia (36 vs. 13%), or PTH suppression by hypercalcemia (65 vs. 25%). However, in the article by John et al. (1) the proportion of non-184 PTH recognized by N-IRMA is not much greater in dialysis patients than in healthy controls. Indeed, we calculated it by using the ratio of the ULN range to be 52% in controls (6531/65) and 59% in the eight dialysis patients (688277/688) in basal condition and 50% after stimulation by citrate-induced hypocalcemia. Only when PTH secretion was suppressed by calcium infusion was the proportion much higher (78%), this latter phenomenon being easily explained by the well known increased catabolism of intact 184 PTH to C-terminal fragments within the parathyroid glands (8). Furthermore, Slatopolsky et al. (9) recently showed that the proportion of non-184 PTH fragment recognized with N-IRMA was not lower in transplanted patients with better renal function than in dialysis patients since the contrary is rather observed: 44% vs. 34%. Thus, the 1.5- to 5-fold higher PTH levels found necessary to maintain normal BFR in the dialysis populations of Quarles et al. (5) and Wang et al. (6) cannot be totally explained by a disproportionate retention or secretion of non-184 fragments, since the proportion of these latter has been documented only marginally higher in stable dialysis patients than in nonuremic patients in the study of John et al. (Ref. 1 ; 59% vs. 52%).
2. The hypothesis of Slatopolsky et al. (9) that the 784 PTH fragments could, furthermore, have an antagonistic action vs. endogenous 184 PTH on bone remodeling in uremic patients is, therefore, quite attractive. This hypothesis is based on the observation that human 784 PTH (hPTH 784) did not increase cAMP production in osteoblast-like cells and antagonized in nonuremic rats the hypercalcemic and phosphaturic effects of hPTH 184. However, to support the conclusion that this antagonistic effect proven in nonuremic rats is the explanation for the development of ABD in dialysis patients, one would like to document a higher proportion of non-184 PTH measured by N-IRMA in patients with ABD than in those with osteitis fibrosa. Since according to John et al. (1) this proportion is rather lower (52% and 55%) in the two patients with ABD than in the six patients with osteitis fibrosa (61% and 59%), we do not think that this explanation is quite satisfactory.
3. Therefore, we think that the main reason for the discrepancy between our proposition regarding the optimal range for intact PTH is related to a greater bone resistance to the PTH remodeling effect in the American dialysis patients than in the Amiens dialysis patients, in relation to their higher exposition to aluminum. Indeed, none of the Amiens patients had traces of aluminum staining on their bone biopsy because they had never been exposed to aluminum either by the dialysate or by the phosphate-binders, these latter having been definitively excluded since 1980. On the contrary, even though Quarles et al. (5) excluded patients with stainable aluminum, aluminum phosphate-binders were still prescribed in their center. As regards the patients of Wang et al. (6), only those with more than 25% of their osteoid/calcified bone interfaces positive for aluminum were eliminated, so that most of them had positive aluminum staining from 124% of their bone interfaces because of previous aluminum-phosphate-binder ingestion. Since it is well documented that aluminum has a direct inhibitory effect on osteoblasts, even when no stainable aluminum is present (10), it is very likely that the difference in aluminum overload is the main explanation for the differences in optimal range of PTH between our center and the American ones.
4. The role of treatment by calcitriol has to be eliminated since calcitriol treatment even without aluminum phosphate binder can further down-regulate PTH-PTHrP receptor (11) and decrease high bone remodeling without a significant decrease of plasma intact PTH (12, 13). This factor can, however, be eliminated as an explanation for our discrepancy with the proposal of Wang et al. (6) since none of our patients and of those of Wang et al. (6) were receiving calcitriol.
The fact that in our unique population of patients without exposition to aluminum, or calcitriol, we documented a subpopulation with truly normal secretion of PTH and low BFR, formally establishes for the first time that the uremic bone of these patients is not only resistant to the hypercalcemic effect of exogenous PTH [shown by Massry et al. (14) in 1974] but also to the remodeling effect of endogenous PTH.
References