Bone Histologic Response To Deferoxamine In Aluminum .

4m ago
2 Views
0 Downloads
579.62 KB
7 Pages
Last View : 3m ago
Last Download : n/a
Upload by : Ciara Libby
Transcription

View metadata, citation and similar papers at core.ac.ukbrought to you byCOREprovided by Elsevier - Publisher ConnectorKidney International, Vol. 31 (1987), pp. 1344—1350CLINICAL INVESTIGATIONBone histologic response to deferoxamine in aluminum—relatedbone diseaseDENNIS L. ANDRESS, HENRY G. NEBEKER, SUSAN M. OTT, DAVID B. ENDRES,ALLEN C. ALFREY, EDUARDO A. SLATOPOLSKY, JACK W. COBURN,and DONALD J. SHERRARDDepartments of Medicine and Nephrology at the Seattle Veterans Administration Medical Center and the University of Washington, Seattle,Washington; University of Southern Cal(fornia Medical Center, and Veterans Administration Medical Center, Wadsworth Division, LosAngeles, California; University of Colorado, Denver, Colorado; and Washington University,St. Louis, Missouri, USABone histologic response to deferoxamine in aluminum—related bonedisease. We have examined the changes in bone histology in 28 uremicnum accumulation [13, 14], long—term studies in a large population have not been reported. In the present study, we reportpatients after long—term treatment with the aluminum chelator, the histologic response of bone to long—term aluminum cheladeferoxamine. Marked declines in stainable bone—surface aluminumwere associated with increases in bone formation rate and osteoblasticosteoid following deferoxamine. The increased bone formation resultedfrom increases in bone apposition and length of double—tetracyclinetion with deferoxamine in uremic patients with increasedsurface—aluminum staining of bone.labels, the latter being highly correlated with the increase inMethodsosteoblastic osteoid (r 0.85). While bone surface aluminum washighly correlated with bone formation rate (r .69, p .001), bonePatientsaluminum content did not correlate with bone formation (r 0.13) andwas often elevated after treatment despite an improvement in boneTwenty—eight patients (23 males and 5 females) underwentiliac crest bone—biopsies (after double tetracycline labeling)parathyroid glands. We conclude that aluminum chelation therapy with both before and after intravenous therapy with deferoxamine.deferoxamine is effective in ameliorating the bone histology of patientsTwenty—one patients were symptomatic with bone pain and/orwith chronic renal failure and bone aluminum accumulation, and that nontraumatic fractures and seven were asymptomatic for bonethe change in stainable bone—surface aluminum is a more sensitive disease at the start of the study. Patients were included forindicator than the change in bone aluminum content in assessingadequacy of chelation therapy. Patients who need deferoxamine treat- study if aluminum staining occupied at least 30% or more of thement but have undergone a prior parathyroidectomy will probably mineralized bone surface. Seven patients had undergone a priorrequire a more intensive treatment schedule than those who have intactparathyroidectomy, and one patient had insulin—dependentparathyroid glands.diabetes mellitus.Maintenance deferoxamine was administered as an intravehistology. Patients who had undergone prior parathyroidectomy wereless likely to have improved bone histology than those with intactnous infusion during the last two hours of dialysis, except in onepatient with chronic renal failure not on dialysis who receivedweekly deferoxamine subcutaneously by infusion pump. Theindividual weekly dosage of deferoxamine ranged from 2 to 6grams and the duration of treatment ranged from 6 to 18 monthsSD, 11.72.7 months). In the dialysis patients,(meantients with chronic renal failure not yet receiving dialysis [5, 6], deferoxamine was given either as a single dose or as two dosesin long—term dialysis patients [71 and in diabetic dialysis pa- divided by at least a three—day interval while dialyzing on atients [8] not exposed to water contaminated with aluminum, thrice weekly schedule. Eleven patients dialyzed in the Seattlepresumably from intestinal absorption of aluminum from area where the aluminum content of the water was consistentlybelow 15 jig/liter during deferoxamine treatment; all but one ofaluminum—containing phosphate binders.Successful removal of aluminum during dialysis has been the remainder dialyzed with water treated by deionization andrecently demonstrated using the aluminum chelating agent, reverse osmosis either at home or in regional dialysis centers.Bone aluminum accumulation in chronic renal failure wasfirst identified in groups of uremic patients who underwentdialysis with aluminum contaminated water [1—4]. More recently, aluminum—related bone disease has developed in pa-deferoxamine [9—12]. While reports have suggested that The one patient without water treatment consistently haddeferoxamine may be effective in the treatment of bone alumi- dialysate aluminum levels that ranged from 50 to 70 jig/literthroughout the treatment period. Each patient continued toReceived for publication August 21, 1986and in revised form January 12, 1987 1987 by the International Society of Nephrologyreceive aluminum-containing phosphate binders throughout thestudy period, in doses adjusted to maintain serum phosphoruslevels of 4.5 to 6.0 mg/l00 ml. Baseline serum biochemistriesare given in Table 1.1344

1345Treatment of aluminum bone diseaseTable 1. Baseline serum biochemistriesPatientsaCalcium, mg/dlPhosphorus, mg/dlAlkaline phosphatase, U/literAluminum jig/literSerum parathyroid hormone (PTH)Normal 5206130 10Values are mean SD.a N 27 except for phosphorus and aluminum where N 23Bone biopsyIliac crest bone—specimens were taken before and aftertreatment with deferoxamine, fixed in iced neutral formalin andprocessed as previously described [15]. Histomorphometricanalysis of undecalcified, Goldner—stained sections was doneusing a computerized digitizer. Preliminary data in 10 of thepatients have been reported previously [14].Static bone histologic—measurements included osteoid surface (as percent of total surface), osteoblastic osteoid (cuboidalor "plump" osteoblasts as percent of total surface), total andmineralized bone area and endosteal fibrosis (as percent oftissue area), osteoid area (as percent of total bone area), andosteoid width (jim). Acid phosphatase staining was utilized forthe identification of osteoclasts [16], which were expressed asthe number of osteoclasts per millimeter of total bone surface.Sections were also stained with aurin—tricarboxylic acid for theA radioimmunoassay with an antiserum (CH 9) specific to themid-region and carboxy-terminal portions of the PTH molecule[20] was employed for PTH determinations before and afterdeferoxamine treatment in 13 patients. The normal limit for thisassay is 10 jilEq/ml. To assess acute PTH secretion in sevenother patients, antiserum specific for the amino—terminal portion of PTH was used to measure the PTH response during anacute hypocalcemic challenge [21] before and after therapy withdeferoxamine. This antiserum does not cross—react with biolog-ically inactive mid-region/carboxy-terminal P1'H fragments[22]. The normal range is 11 to 24 pg/mI.StatisticsAll results are expressed as the mean SEM. The Wilcoxonsign—rank test was used for the comparisons of all paired datawith P 0.05 as the minimum level of significance. Linearregression analysis was used for the comparisons of bonehistology.ResultsBone biopsiesSelected parameters of bone histology are displayed in Table2 for each type of renal bone disease. Before treatment withdeferoxamine, 11 patients had osteomalacia, 12 had aplasticdisease, four had mild disease, and one patient had mixed bonedisease. Stainable bone—surface aluminum was significantlygreater in the patients with osteomalacia than in the aplastic anddetection of aluminum on the mineralized bone surface and mild groups (P 0.01). Osteoblastic osteoid was less in bothwithin cement lines [17]. The Prussian blue stain was applied for the osteomalacic and aplastic groups when compared to thethe detection of iron. The amount of aluminum present was group with mild disease (P 0.01). Osteoclast number was notexpressed as: stainable bone—surface aluminum (as percent of statistically different among the groups. Tetracycline labelstotal bone surface), stainable aluminum on the cement—line (asfailed to separate in eight patients with osteomalacia and in fivepercent of total stainable aluminum), and total stainable patients with aplastic disease. The mean BAR was lower in theosteomalacic group than in the mild group (P 0.01) but notsignificantly lower than in the aplastic group (P 0.057). Therewas no difference in bone apposition rate between the mild andaplastic groups.As shown in Table 3, stainable bone—surface aluminum wasscribed [18]. Normal bone aluminum content is 2.4 1.2 mg/kg61 3% before and 21 4% after deferoxamine treatment (P dry weight.Dynamic bone parameters were quantitated on unstained 0.001). Total stainable bone—aluminum also decreased signifisections. The bone apposition rate (BAR) was determined by cantly following deferoxamine therapy (2.57 0.19 vs. 1.32dividing the distance between the two tetracycline labels by the 0.25 mm/mm2 tissue area; P 0.001). Concomitantly, stainablenumber of days between tetracycline administration. The bone aluminum on the cement—line increased from 11 2% to 39formation rate (BFR) was calculated by multiplying the BAR by 5% of the total stainable aluminum (P 0.001). No iron wasthe length of bone surface occupied by double tetracycline detected on the mineralized bone surface. Among the staticbone—aluminum (surface and cement lines stained, as mm/mm2tissue area). In 13 patients, two separate bone samples wereavailable for measurement of total aluminum content usingflameless atomic absorption spectroscopy as previously de-labels. Data from bone in biopsies of 19 normal males (age 21 to69) from the Seattle area were used for comparisons. The BARin the normals ranged from 0.51 to 0.78 jim/day and the BFRranged from 106 to 602 jim2/mm2/day.Biopsies were classified histologically, as previously described [19], as mild (osteold area 15%, fibrosis 0.5% andBFR 106 jim2/mm2/day), osteitis fibrosa (osteoid area 15%,fibrosis 0.5% and BFR 106 jim2/mm2/day), osteomalacia(osteoid area 15%, fibrosis 0.5% and BFR 106bone histologic—parameters, there was a significant increase inosteoblastic osteoid and osteoclast number and among thedynamic bone parameters there were increases in BFR andBAR following deferoxamine treatment. The increase indouble—tetracycline label length from 0.11 0.04 to 0.40 0.09mm/mm2 (P 0.001) represents a greater than 250% change.There was a negative correlation of stainable bone—surfacealuminum with BFR (r —0.47, P 0.001), BAR (r —0.61,P 0.001), and double—tetracycline labeled surfaces (r —0.54,P 0.001). BFR correlated with osteoblastic osteoid (r 0.85,jim2/mm2/day), aplastic (osteoid area 15%, fibrosis 0.5% P 0.001) primarily because of the high correlation betweenand BFR 106 jim2/mm2/day), or mixed (osteoid area 15%, double—tetracycline label length and osteoblastic osteoid (r fihrncic 0 5tr,).0.85. P 0.001). Osteoblastic osteoid also correlated with BAR

Andress et a!1346Table 2. Surface bone aluminum, osteoid width, osteoblastic osteoid, osteoclast number, and bone apposition rate in28 dialysis patients before treatment with deferoxamineType ofBone histologyOsteomalacia (N Aplastic (N 12)Mild (N 4)Mixed (N .71a P 0.01 vs. aplastic and mildbP 0.01 vs. mildP 0.01 vs. osteomalacia25Table 3. Bone histologic changes in 28 uremic patients before andafter treatment with le bone—surfacealuminum, %Total stainable bonealuminum, mm/mm2Stainable cement—linealuminum, % of totalOsteoblastic osteoid %of total surfaceOsteoclasts, no. permm bone surfaceBone apposition ratewn/dayDouble tetracycline labellength, mm/mm2Bone formation rate,wn2/mm2/day612.5730.19211.32400.2505000SSS0. 1tE8027311.0.60.239Values are mean SE.P 0.005 vs. Before valuesb 19 normal patients normal subjects, aged 21 to 59 years, except forosteoblast number where N 71530456075Stainable bone—surface aluminum% total surfaceFig. 1. Relationship between log osteoblastic osteoid and stainablebone—surface aluminum. Closed circles indicate values before(r 0.41, P 0.01). As shown in Figure 1, there was a negativecorrelation between the log osteoblastic osteoid and stainablebone—surface aluminum (r —0.64, P 0.001).deferoxamine treatment and open circles represent post-treatmentvalues. (r —0.64; P 0.001; N 56)Table 4 shows the changes in bone aluminum content,bone formation rate (r 0.13), but there was a significantand serum PTH in 13 dialysis patients before and after negative correlation between stainable bone—surface aluminumstainable bone—surface aluminum, BFR, osteoblastic osteoiddeferoxamine treatment. Bone aluminum content increased in and bone formation rate (r —0.69, P 0.01) in these patients.two patients, showed no change in one patient, and decreased One patient (number 12 in Table 4) with consistently high levelsby a mean of 23 4% in the remaining 10 patients. Only one of aluminum in the dialysate (50 to 70 pg/liter), failed to show anpatient had more than a 50% decrease in bone aluminum increase in BFR and osteoblastic osteoid despite significantcontent following deferoxamine. There was a significant de- reductions in bone aluminum content and stainable bone—crease in mean bone aluminum—content from 115 14 mg/kg to surface aluminum. There were no correlations of the duration of14 mg/kg (P 0.02). In the same patients, the stainable deferoxamine treatment with the change in BFR (r —0.05) or94bone—surface aluminum decreased in all but one patient, with with the change in stainable bone—surface aluminum (r an average decline of 70 6%, and 10 showing more than a 50% —0.06).As shown in Figure 2, 16 of the patients (59%) exhibited adecrease. The mean stainable bone—surface aluminum decreased significantly (57 4 vs. 21 5%; P 0.01). Concom- change in the type of renal osteodystrophy after the treatmentitantly, BFR and osteoblastic osteoid increased following of deferoxamine. Of the 11 patients with osteomalacia, twodeferoxamine (119 54 vs. 530 196 m2/mm2/day; P 0.01 changed to aplastic disease, three changed to mild disease, two1.5 vs. 6.9and 3.02.0%; P 0.01, respectively). The changed to osteitis fibrosa, and one changed to a transitionalpretreatment bone aluminum-content did not correlate with stage of mild histology in which osteoid area was 31% (baseline,

1347Treatment of aluminum bone diseaseTable 4. Bone aluminum content, stainable bone—surface aluminum, bone formation rate, osteoblastic osteoid,and serum PTH in 13 patients before and after DFO 48589131237466mean SEM115103149414b57 25.72.97.93544064726810.32.411210300119 554PTHB262438044631424921AB51378350bl. 55720994419897 30395764116361Abbreviations are: BAC, bone aluminum content (mg/kg); SBA, stainable bone—surface aluminum (% total surface); BFR, bone formation rate(sm2/mm2/day); ObI. Ost,, osteoblastic osteoid (% total surface); PTH, mid-region assay (normal, 10 dEq/ml); B, before; A, after. a Patientswho had a prior parathyroidectomy. b P 0,02 vs. before and C P 0.01 vs. before.Before DFOAfter y none of the patients in the asymptomatic groupdeveloped bone symptoms or fractures during treatment withdeferoxamine. Among the symptomatic patients (N 21), 17showed moderate to marked improvement in bone pain and fourdemonstrated no change in their bone symptoms. None hadworsening of musculoskeletal symptoms or developed fracturesduring therapy.Among the seven patients with a prior parathyroidectomy,stainable bone—surface aluminum decreased after deferoxamine7 vs. 46treatment (649%; P 0.05), but no significantchanges were observed in osteoblastic osteoid (0.6or BFR0.2 vs.1.2vs. 6920 m2/mm2/day). Thiscontrasts to the changes seen in the group with intact parathyroid glands as shown in Figure 3. There were no differences in0.4%)(2514the total dose of deferoxamine administered or duration ofFig. 2. Changes in bone histologic classification following deferoxamine therapy. (*) indicates mild bone histology except that osteoidtreatment between the two groups.Serum PTHAs shown in Table 4, the mean serum PTH did not changesignificantly following deferoxamine treatment (before, 97 3061 pg/mI, P 0.48) in the 13 patients inpg/mi; after, 163which these measurements were made. Six of these patientshad osteomalacia (patients 1, 3, 6, 7, 12, and 13), three hadaplastic disease (patients 2, 8, and 10), three had mild bone41%) after eight months of treatment. Among the 12 patients disease (patients 4, 5, and 11), and one had mixed histologywith aplastic bone disease, eight changed to mild disease and (patient 9). The three patients with prior parathyroidectomythree remained aplastic. One patient with aplastic disease (patients 1 to 3) had lower PTH levels than the other patients.initially could not be categorized on the final biopsy because Five patients showed elevations of serum PTH aftertetracycline was not given. However, since there was only a deferoxamine, which ranged from 37 to 512 pg/mI above basemodest increase in osteoblastic osteoid (from 0.9% to 1.3%), line (mean change, 203 94 pg/mI). Two patients had declinesany increase in BFR from baseline (BFR 0) was probably not in serum PTH which were 31 and 86 pg/mI below baseline.sufficient to change the disorder from the aplastic category. All There was a positive correlation between serum PTH, both preof the patients with mild or mixed bone histology remained with and post-treatment values, and osteoblastic osteoid (r 0.78, Pthat diagnosis on the follow—up biopsy. Among the patients 0.001; N 26).with osteomalacia and aplastic disease who did not change theirAmong the seven patients who had amino—terminal PTHhistologic classification (6 patients), five had undergone a pre- levels determined during a hypocalcemic challenge, before andvious parathyroidectomy.after deferoxamine treatment, six had aplastic bone disease andarea was 31% compared to 41% before deferoxamine treatment (that is,transitional stage). Closed circles indicate patients who had undergoneprior parathyroidectomy. Open box indicates one patient who did nothave tetracycline labeling on final biopsy to determine bone formationrate.

1348Andress et a!was associated with improvement in bone mineralization as60—C0.— 'E'shown by increases in bone apposition and bone formation rate,which paralleled the increase in osteoblast number. As a result,there was a favorable change in the histologic classification ofbone disease in the majority of patients.40.0 m—CQ0(oc.-20—BeforeAfterThe most common type of bone histology observed afterdeferoxamine treatment in the patients who had a change inbone histology was the mild lesion. The patients with eithermild (4 patients) or mixed bone disease (1 patient) beforetreatment remained with that diagnosis on the followup bonebiopsy. Thus, only the patients with low turnover osteodystrophy (osteomalacia and aplastic disease) had a change in bonehistologic classification after deferoxamine.Improvements in bone formation and bone histology afterdeferoxamine treatment occurred despite the absence of anincrease in PTH (Table 4). However, there was a highly6significant positive correlation between the serum PTH andQ.0osteoblastic osteoid (r 0.78, P,.4,.n.0o 0.001, N 26) among the 13patients who had PTH determinations, suggesting that smallincreases in serum PTH may have been partially responsible forsome of the increases in osteoblast number. Interestingly, in theseven patients in which amino—terminal PTH levels were measured after an acute hypocalcemic challenge, no increase in the2BeforeAfterPTH response was seen after treatment. However, stainablebone—surface aluminum decreased and bone formation increased significantly in this group, suggesting that site—specificaluminum removal caused an enhancement of bone formationindependent of any change in PTH in these patients.The favorable changes we observed in bone apposition,stainable bone—surface aluminum and osteoblastic osteoid afterlong—term deferoxamine treatment are in agreement with thew.bone histologic response of the three patients reported byMalluche et al [13]. While the mean bone apposition ratecoSonincreased by more than 80% after treatment, there was a largerpercent increase in the mean double—tetracycline label length(250%) which we attribute to the marked increase in the numberof active osteoblasts. In contrast to the findings of Malluche etRfforeAfterFig. 3. Stainable bone—surface aluminu,n, osteoblastic osteoid andbone formation rate before and after treatment with deferoxamine inpatients with a prior parathyroidectomy ( —I; N 7) and patientswith intact parathyroid glands (O--O; N 20). ( ) indicates P 0.05vs. before; (*) indicates P 0.01 vs. the group with parathyroidectomy.al [13], we found that the bone aluminum content decreasedonly minimally in the 13 patients in which it was measured,compared to the decline in stainable bone—surface aluminum.Bone aluminum content decreased by only 18% while bonesurface aluminum decreased approximately 60% and boneformation increased by more than 300% (Table 4). Moreover, inthis group of patients, the pretreatment bone aluminum—contentdid not correlate with bone formation rate (r 0.13), whileone had osteomalacia; one patient also had a prior parathy-there was a significant negative correlation between stainablebone—surface aluminum and bone formation rate (r —0.69, P 0.01). Thus, the small amounts of aluminum removed fromroidectomy. Before deferoxamine, the maximum PTH increasebone in our patients appeared to have little if any role in thefrom baseline was 269 pg/mi and after treatment the improved bone formation. Rather, the translocation of alumi-10 pg/mI. In this group theremaximum PTH increase was 28was a significant decrease in stainable bone—surface aluminum(56 5 vs. 10 3%, P 0.02) and an increase in BFR (63 17vs. 14019 m2/mm2/day; P 0.02), but no significant changein osteoblastic osteoid (0.9 0.2 vs. 2.6 0.6; P 0.063).num from the bone surface to the mineralized bone compartment, as exemplified by the increase in stainable cement—lineDiscussionaluminum chelation, and is, therefore, a better guide to chelation therapy than is the change in bone aluminum content.We have shown that long—term deferoxamine therapy iseffective in removing aluminum from mineralized bone inuremic patients with elevated bone aluminum. This responsealuminum, and the increase in osteoblast number, were themajor determinants of the increased bone—formation rate. Fromthis we conclude that the amount of aluminum on the bonesurface is a more sensitive indicator of the bone response toThe optimum duration of therapy and the amount ofdeferoxamine needed for successful treatment of aluminum—as-

1349Treatment of aluminum bone diseasesociated bone disease cannot be determined from this study. Itappears, however, that a rather prolonged treatment schedulemay be necessary in the majority of patients. It is possible thatthe concomitant administration of aluminum—containing phosphate binders in our patients may have contributed to this longtreatment period. Perhaps the use of calcium carbonate as aphosphate binder[23] would adequately control serum phosphorus levels and allow for the discontinuation of aluminum bindersduring chelation therapy. Whether the use of newer methods toenhance aluminum removal during dialysis [24] can also makechronic deferoxamine treatment more efficient remains to bedetermined.Patients who have had a parathyroidectomy prior todeferoxamine therapy may require more intensive treatmentthan patients with intact parathyroid glands. Among the sixpatients with low turnover bone disease (3 with osteomalaciaand 3 with aplastic disease) who failed to change their bonehistologic classification (Fig. 2), five had undergone a parathyroidectomy prior to receiving deferoxamine. When all sevenpatients with parathyroidectomy were analyzed together, avery modest decrease in bone surface aluminum was noted inthe group, although osteoblastic osteoid and bone formation didnot improve (Fig. 3). The reason for their poor response totreatment is unclear but may be related to a requirement for anoptimum level of circulating PTH for the maintenance of normalbone formation. Thus, a greater dosage of deferoxamine may berequired in patients with prior parathyroidectomy. This dosageadjustment should probably also include the discontinuation ofaluminum-containing phosphate binders. These data and thefinding that bone aluminum accumulation is enhanced afterparathyroidectomy [25—271 emphasize the need to developbetter treatments for secondary hyperparathyroidism.In summary, we have shown that long—term aluminum chela-ride: Relationship to impaired bone mineralization. J Clin Pathol32:832—844, 19794. Pierides AM, Edwards WG, Cullum UX, McCall iT, Ellis HA:Hemodialysis encephalopathy with osteomalacia fractures andmuscle weakness. Kidney mt 18:115—124, 19805. FELSENFELD AJ, GUTMAN RA, LLACH F, HARRELSON JMOsteomalacia in chronic renal failure: A syndrome previouslyreported only with maintenance dialysis. Am J Nephrol 2:147—154,19826. ANDREOLI SP, BEROSTEIN JM, SHERRARD Di: Aluminum intoxi-cation from aluminum—containing phosphate binders in childrenwith azotemia not undergoing dialysis. N Engl J Med 310:1079—1084, 19847. ANDRESs DL, MALONEY NA, ENDRE5 DB, SHERRARD Di:Aluminum—associated bone disease in chronic renal failure: Highprevalence in a long—term dialysis population. J Bone Miner Res1:391—398, 19868. ANDRESS DL, Ko JB, MALONEY NA, COBURN iW, SHERRARDDi: Early deposition of aluminum in bone in patients with diabeteson hemodialysis. N Engi J Med 316:292—296, 19879. ACKRILL P. RALSTON Ai, DAY iP, HODGE KC: Successful removalof aluminum from a patient with dialysis encephalopathy. Lancet2:692—693, 198010. BROWN Di, HAM KN, DAWBORN iK, XIPPEL iM: Treatment ofdialysis osteomalacia with desferrioxamine, Lancet 2:343—345, 198211. IHLE BU, BUCHANAN MRC, STEVENS B, BECKER GJ,KINCAID—SMITH P: The efficacy of various treatment modalities onaluminum associated bone disease. Proc EDTA 19:195—201, 198212. ACKRILL P, DAY JP, GARSTANG FM, HODGE KC, METCALFE Pi,BENZO Z, HILL K, RALSTON Ai, DENTON i: Treatment of fractur-ing renal osteodystrophy by desferrioxamine. Proc EDTA19:203—207, 198213. MALLUCHE HH, SMITH AJ, ABREO K, FAIJGERE MC: The use ofdeferoxamine in the management of aluminum accumulation inbone in patients with renal failure. N Engi J Med 311:140—144, 198414. OTT SM, ANDRESS DL, NEBEKER HG, MILLINER DS, MALONEYNA, COBURN JW, SHERRARD Di: Changes in bone histology aftertreatment with deferrioxamine. Kidney mt 29(Suppl 18): 108—113,198615. SHERRARD Di, BAYLINKDi, WEGEDAL JE, MALONEY NA: Quan-titative histological studies on the pathogenesis of uremic bonetion therapy with deferoxamine effectively ameliorates alumidisease. J Clin Endocrinol Metab 39:119—135, 1974num associated, low—turnover renal osteodystrophy in patients 16. EVANS RA, DUNSTAN CR, BAYLINK Di: Histochemical identificawith intact parathyroid glands. Removal of aluminum from thetion of osteoclasts in undecalcified sections of human bone. MinerElectrol Metab 2:179—185, 1979bone surface is associated with marked increases in osteoblastnumber and bone formation, resulting in improved bone histol- 17. MALONEY NA, O-rr S, ALFREY AC, Cosu JW, SHERRARD Di:Histologic quantitation of aluminum in iliac bone from patients withogy. While small increases in serum PTH following aluminumrenal failure. J Lab Gun Med 99:206—216, 1982chelation may have an anabolic effect on bone function, alumi- 18. LEGENDRE GR, ALFREY AC: Measuring picogram amounts ofnum removal alone can result in increased bone formationwithout changes in PTH.aluminum in biological tissue by flameless atomic absorption analysis of a chelate. Clin Chem 22:53—56, 197619. ANDRESS DL, ENDRES DB, MALONEY NA, KoP iB, COBLJRN JW,AcknowledgmentsThis study was supported in part by General Medical Research fundsfrom the Veterans Administration.SHERRARD Di: Comparison of parathyroid hormone assays withbone histomorphometry in renal osteodystrophy. J Clin EndocrinolMetab 63:1163—1 169, 198620. HRUSKA KA, KOPELMAN R, RUTHERFORD WE, KLAHR S,SLATOPOLSKY E: Metabolism of immunoreactive parathyroid hor-mone in the dog: The role of the kidney and the effects of chronicReprint requests to Dennis L. Andress, M.D., Dialysis Unit (11/A),VA Medical Center, /660 S. Columbian Way, Seattle, Washington98108, USA.renal disease. J Gun Invest 56:39—48, 197521. ANDRESS D, FELSENFELD Ai, VOIGTS A, LLACH F: Parathyroidhormone response to hypocalcemia in hemodialysis patients withReferencesI. WARD WK, FEEST TG, ELLIS HA, PARKINSON IS, KERR DNS:osteomalacia. Kidney mt 24:364—370, 198322. SEGRE GV: Amino—terminal radioimmunoassays for human para-thyroid hormone, in Clinical Disorders of Bone and MineralMetabolism, edited by FRAME B, POTTS iT iR. Amsterdam,Osteomalacic dialysis osteodystrophy: Evidence for a water—borneaetiological agent, probably aluminum. Lancet 1:841—845, 19782. PARK

Treatment of aluminum bone disease 1345 Table 1. Baseline serum biochemistries Patientsa Normal range Calcium, mg/dl 10.3 0.7 8.5—10.5 Phosphorus, mg/dl 5.7 1.1 2.5—4.5 Alkaline phosphatase, U/liter 193 132 30—