- Email: [email protected]
Hypomagnesemia and renal magnesium wasting in renal transplant recipients receiving cyclosporine
Hypomagnesemia and Renal Magnesium Wasting iri Renal Transplant Recipients Receiving Cyclosporine
CYRIL H. BARTON, M.D. NOSRATOLA D. VAZIRI, M.D. DONALD C. MARTIN, M.D. SUNNY CHOI, M.D. SHAHRIAR ALIKHANI Irvine, California
From the Divisions of Nephrology and Urology, University of California, Irvine, Irvine, California. Requests for reprints should be addressed to Dr. Cyril H. Barton, Division of Nephrology, University of California, Irvine Medical Center, 101 The City Drive, Orange, California 92668. Manuscript submitted July 7, 1986, and accepted May 15, 1987.
Following the adoption and use of cyclosporine as the drug of choice in the management of renat allograft recipients, several,cases of symp tomatic hypomagnesemia were noted. These observations prompted the current prospective study of serum concentration and urinary excretion of magnesium in 27 renal transplant recipients treated with cyclosporine and prednisone. Relevant laboratory measurements were obtained shortly before and regularly after transplantation. The results were compared with those obtained in a group of 17 allograft recipients treated with azathioprine and prednisone. The cyclosporine-treated patients showed a significant reduction In the serum magnesium concentratlon and an inappropriately increased urinary excretion and fractional excretion of magnesium, suggesting renal magnesium wasting. The observed hypomagnesemia required magnesium supplemen; tation in nearly all cyclosporine-treated patients. In contrast, azathioprine-treated patients showed normal serum magnesium concentrations and required no magnesium supplementation. In conclusion, administration of cyclosporine in renal allograft recipients appears to be commonly associated with renal magnesium wasting and hypomag nesemia. Therefore, Wis recommended that serum levels of magnesium be monitored regularly in renal allograft recipients receiving cyclosporine and that magnesium supplementation be employed as needed to avoid magnesium depletion. Currently, cyclosporine is widely used in the management of patients who have undergone organ transplantation. Although use of cyclosporine affords unquestionable benefits such as improved allograft survival and reduced steroid requirement, it is associated with various toxic effects. Nephrotoxicity with functional renal insufficiency remains the major adverse effect of this drug. Other side effects and complications include infection, lymphoma, hypertension, hyperkalemia, seizures, gingival hyperplasia, hirsutism, and hepatotoxicity [l-8]. We have observed symptomatic hypomagnesemia in several renal allograft recipients treated with cyclosporine but in none of those treated with azathioprine. We, therefore, undertook the current study to examine the possible association between cyclosporine and magnesium depletion in renal allograft recipients treated with cyclosporine and prednisone. The results were compared with those obtained in allograft recipients given azathioprine and prednisone. PATIENTS AND METHODS Cyclosporlne-Treated Group. Between July of 1984 and December of 1985, 42 renal transplant procedures were performed at our institution in which cyclosporine and prednisone were used for immunosuppression.
October
1987
The American
Journal
of Medicine
Volume
83
693
HYPOMAGNESEMIA
AND
RENAL
MAGNESIUM
WASTING
WITH
CYCLOSPORINE-BARTON
We were unable to include 15 of these patients in our study because of early allograft failure (three patients), death (three patients), and loss of follow-up (nine patients). Of the 27 allograft recipients who were studied, 25 were cadaveric, whereas two were living related donors. Sixteen were men and 11 were women, with an age range of 23 to 63 years and a mean age of 44 years. The cause of end-stage renal disease was chronic glomerulopathy (eight patients), diabetic nephropathy (seven patients), nephrosclerosis (five patients), polycystic kidney disease (two patients), chronic interstitial nephropathy (two patients), and renal failure of unknown etiology (three patients). Prior to renal transplantation, 25 patients had been treated with long-term hemodialysis, whereas two patients were managed on continuous ambulatory peritoneal dialysis. lmmunosuppression was accomplished using cyclosporine and prednisone. A dose of 5 mg/kg per day of cyclosporine was administered intravenously for the first three days of engraftment followed by orally administered cyclosporine (10 mg/kg per day). The oral dose was rapidly tapered so that by the sixth week most patients were receiving approximately 5 mg/kg per day with individual adjustments being made to maintain trough serum levels between 60 and 200 rig/ml. Patients were also treated with intravenously administered methyl prednisolone, 1,000 mg per day in four divided doses, for three days followed by a rapidly tapering schedule of orally administered prednisone so that by two to three months after engraftment most patients were receiving 15 to 20 mg per day. Other medications included the routine use of oral antacids (aluminum hydroxide or aluminum phosphate gel) or ranitidine during periods of high-dose steroid administration. Antihypertensive medications including cionidine, hydralazine, and/or propranoioi were used when indicated to control blood pressure. Eleven patients (41 percent) also received furosemide as an adjunct in the management of hypertension. Since furosemide is known to cause renal magnesium-wasting, data from these patients were compared with those obtained in the remaining 16 patients who did not receive furosemide. Arathioprine-Treated Group. Seventeen patients (11 cadaveric and six living related donors) who underwent transplantation at our institution and were treated with azathioprine and prednisone were included for comparison. Eleven were men and six were women, with an age range of 25 to 51 years and a mean age of 36 years. The cause of endstage renal disease was chronic glomeruiopathy (six patients), diabetic nephropathy (three patients), nephroscierosis (one patient), chronic interstitial nephritis (two patients), Aiport’s disease (one patient), lupus erythematosis (one patient), and renal failure of unknown etiology (three patients). Prior to transplantation, 16 patients had been treated with hemodlaiysis, and one patient was managed with continuous ambulatory peritoneal dialysis. Eight patients (47 percent) were receiving furosemide in addition to other antihypertensive medications including clonidine, hydraiazine, and/or propranoloi. None of the patients was receiving any magnesiu.in-containing preparations. Although all patients were studied prospectively, only three could be entered into the study at the time of engraftment. We were not able to study a larger number of azathioprine-treated patients in a parallel fashion due to the fact
694
October
1967
The American
Journal
of Medicine
Volume
ET AL
that the vast majority of patients who received transplants at our institution after July of 1964 were given cyciosporine to achieve immunosuppression. The additional I4 azathioprine-treated patients were selected from our clinic popuiai tion and represent a group of consecutive recipients witii good aiiograft function who underwent transplantation prior to the introduction of cyciosporine at our center. The mean duration of engraftment of this group at the time of entering the study was 23 weeks, while the mean serum creatinine level was 2.1 mg/dl. Control Group. The control group consisted of 30 normal volunteers, 16 women and 12 men, with a mean age of 39 years. None of the participants in the control group was taking any drugs or receiving magnesium supplementation. Mid-morning serum and urine samples were obtained for measurement of magnesium and creatinine concentration. Methods. In the cyclosporine group, baseline laboratory measurements were obtained shortly prior to engraftment and were repeated at weekly intervals for eight weeks and at monthly intervals for four months after transplantation. In the azathioprine group, laboratory measurements were also obtained at weekly intervals for eight weeks and at monthly intervals for four months; however, in only three of the 17 patients could baseline measurements be obtained. Blood specimens were drawn in the morning, after approximately 10 to 12 hours of fasting. Serum concentrations of creatinine, sodium, chloride, potassium, bicarbonate, caiciurn, inorganic phosphorus, and albumin were determined using standard methods, while serum magnesium levels were measured using a DuPont automatic clinical analyzer. Plasma trough levels of cyciosporine were measured using radioimmunoassay approximately 12 hours after administration of the last dose, with blood specimens being kept at 37’C prior to plasma separation. Both timed and spot urine collections were obtained for measurement of creatinine and magnesium concentrations. The urine magnesium concentration was measured using atomic absorption. Twentyfour-hour urinary magnesium excretion was calculated from the urine concentration of magnesium and 24-hour urine volume. Fractional excretion of magnesium (Feyg) was calculated using the following formula: FeMg = UMg/SMg
f UCr/SCr
X 100
where UMg and SMg represent urine,and serum magnesium concentrations, respectively, while UCr and SCr represent urine and serum creatinine concentrations, respectively. Since approximately 20 percent of serum magnesium is protein bound and not ultrafilterable, using the serum magnesium level in the foregoing formula slightly underestimates true fractional magnesium excretion. This relatively small systematic error, however, is common to both groups and all collection periods and as such should not affect the significance of the observed changes. ‘” Data Expression and Analysis. The data are presented as mean f SD. The Student t test and regression analysis were used where appropriate. Changeover time in the serum and urine measurements was evaluated by means of a repeated measures, multiple analysis of variance in order to determine the significance of any trend. p Values less than 0.05 were considered to be significant.
63
HYPOMAGNESEMIA
TABLE I
Laboratory
Data in Serum Cyclosporine-Treated
Serum Magnesium* (mg/W
Patients Baseline Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 6 Month 3 Month 4 Month 5 Month 6
2.6 2.1 1.6 1.6 1.6 1.7 1.7 1.7 1.6 1.6 1.6 1.5 1.7
f f f f f f f f f f f f f
AND RENAL
Serum Creatinine’ (WdU
0.4 0.4 0.4 0.2 0.3 0.3 0.4 0.3 0.3 0.3 0.1 0.2 0.3
10.7 f 4.0 f 3.1 f 2.4 f 2.3 f 2.3 f 2.4 f 2.1 f 2.1 f 2.0 f 1.6f0.7 2.2 f 2.0 f
NA = not available. * p
Urinary Magnesium (mg124 hours)
3.0 2.5 1.5 1.0 1.4 1.1 1.4 1.0 0.6 0.7
NA 115f75 92f 66 f 114f41 147 f 101 f 146 f 101 f 95 f 103f 120 f 125 f
0.7 0.7
were evaluated trends.
by a multiple
The mean serum and urine values collected over month study period in the 27 cyclosporine-treated are
shown
magnesium to engraftment, ease. nesium
However, levels
Table I. The
in
level
of 2.6 f 0.4 is characteristic
by Week dropped to
elevated
the
baseline
mg/dl, measured of end-stage
2 of engraftment, 1.8 f 0.4 mg/dl
analysis
prior dis-
of
with
in
baseline
values.
serum magnesium supplementation (750 to 1,500 mg further significant concentration majority of levels fell magnesium whom pletion increase
the
To
prevent
below 1.6 was initiated per day) reduction
occurred patients.
as required. in mean beyond the However,
to values between supplementation
exhibited (muscle doses
further mg/dl, using
oral magnesium magnesium oxide Consequently, serum magnesium third serum
0.9 and in five
signs and symptoms weakness and tremors) of magnesium oxide.
reductions
no
despite two of
of magnesium that responded The percentage
deto of
cyclosporine-treated patients requiring supplemental magnesium oxide at given intervals during the study is shown in Table I. As can be seen, by the end of five to six months,
virtually
all patients
m?gnesium supplementation. The urinary magnesium tion of magnesium in the
were
requiring
excretion and cyclosporine-treated
and fr@ional
receiving
renal magnesium-wasting. correlation between
There the
serum
pendency function
was magne-
October
NA 379 f 294 f 453 f 367 f 299 f 165 f 196f70 202 f 266 f 365 f 169 f 251 f
repeated-measure
ET AL
of and
fluctuations throughout
in the
with
significant concentration
of calcium, inorganic chloride, bicarbonate,
97 161 146 56 170
the
p Values
less than 0.01
0.56, p the de-
hypomagnesemia on renal magnesium-wasting.
renal
correlation between the and serum concentrations
phosphorus, or albumin.
serum study,
NA 11 11 26 46 60 62 56 65 67 64 95 96
concentrations (r = further demonstrates
the observed is consistent no
Percent Receiving Magnesium Oxide
254 330 264 260 196 105
design.
serum creatinine This association
There was magnesium
se-
potassium, sodium, Although considerable
cyclosporine somewhat
levels higher
were noted levels present
during the early weeks of engraftment no doubt reflect the higher dosages of cyclosporine administered at that time. Our data, however, show no significant correlation bethe
serum
levels
of cyclosporine
renal magnesium-wasting A comparison of the who did not did receive there were subgroups fractional cate not
receive furosemide this drug is shown no significant in terms of serum
excretion only that
TABLE II
with in
the
Table
differences magnesium
of magnesium. a factor (or factors)
the degree
of
patients
11 patients who II. Surprisingly, between the concentration
These other
two or
findings indithan furose-
hypomagnesemia were not substan-
Comparison of Serum and Urinary Magnesium Values In CyclosporlneTreated Patients Receiving and Not Receiving Furosemide No Furosemide
Serum magnesium @W-W Fractional excretion of magnesium (percent) Serum creatinine (mg/di)
1987
and
and/or hypomagnesemia. 16 cyclosporine-treated
mide was responsible for the observed but also that the effects of furosemide
excrepatients at
the end of Week 1 were 115 f 75 mg per 24 hours and 15 f 8.0 percent of filtered magnesium load, respectively. These values remained inappropriately high throughout the study period despite persistent hypomagnesemia, which suggests also a significant
sium and
tween
week in the magnesium
1.1 mg/dl patients,
CYCLOSPORINE-BARTON
Serum Cyc!osporine Wml)
6.0 6.7 7.7 8.7 6.7 6.5 5.1 4.0 4.6 7.0 9.9 a.7
of variance,
rum
serum magand by the end
NA f f f f f f f f f f f f
15.0 12.6 13.0 14.2 15.2 13.5 14.2 9.7 10.1 15.6 14.5 13.3
52 32 39 14 a 11 31 42
sixpa-
of Week 3, frank hypomagnesemia was present (levels 1.6 f 0.2 mg/dl). These changes in serum magqesium levels were highly significant (p
WITH
Fractidnal Excretion (percent)
19 34
serum just renal
WASTING
Patlents
RESULTS
tients
MAGNESIUM
The American
Journal
Furosemide
1.6 f
0.4
1.7f0.3
16.1 f
6.0
14.0 f
2.7 f
1.6
2.4fl.O
of Medicine
Volume
p Values NS
6.9
NS NS
83
696
HYPOMAGNESEMIA
TABLE III
AND
RENAL
Comparison
MAGNESIUM
of Relevant
sM9 (m9W
Groups Cyclosporine-treated Azathioprine-treated p Values
1.7 f 0.3 2.3 f 0.4
WASTING
WITH
Laboratory
FeMg (percent) 13.5 f 7.0 5.0 f 2.8
scr (mgldl) 2.6 f 1.5 2.1 f 1.0 NS
CYCLOSPORINE-BARTON
Values
ET AL
In Cyclosporlne-Treated
(me$ter)
(rnzkr
)
4.5 f 0.6 4.5 f 0.7 NS
142 f 3.5 142 f 5.2 NS
(me:iter
696
October
1967
The American
Journal
of Medicine
Volume
)
107 f 8.0 104 f 7.9 NS
sMg = serum magnesium; FeMg = fractional excretion of magnesium; sCr = serum chloride; HC03= bicarbonate; Ca = calcium; Pi = phosphorus. * Cumulative mean values collected over a six-month study period are shown.
tially additive. It should be pointed out, however, that approximately half of the patients included in the furosemide-treated subgroup were only intermittently receiving this agent, which may explain at least in part why more severe hypomagnesemia was not observed. In contrast to the findings in the cyclosporine-treated patients, serum magnesium levels in the 17 azathioprinetreated patients were at the upper limits of normal (normal range for serum magnesium concentration reported by our laboratory is 1.7 to 2.4 mg/dl) at 2.3 f 0.4 mg/dl during the six-month observation period. Furthermore, none of the azathioprine-treated patients, including the three who entered the study at the time of engraftment, exhibited serum magnesium levels below 1.7 mg/dl, and none required or received magnesium supplementation. Table III compares the cumulative serum and urine measurements in the azathioprine and cyclosporine groups. It can be seen that serum magnesium levels were significantly higher while the fractional excretion of magnesium was significantly lower in the azathioprine-treated group as compared with the cyclosporine-treated group. Overall, these findings show no evidence for renal magnesium wasting or hypomagnesemia in the patients treated with azathioprine. The serum magnesium concentration in our 30 normal control subjects was 2.12 f 0.17 mg/dl. This value is significantly higher than the mean serum magnesium concentration in the cyclosporine group for the six-month study period (p
versus Arathloprlne-Treated
creatinine;
(n!$&
Groups* Albumin (Wdl)
(miydl)
24 f 3.9 28 f 5.7
9.9 f 1.4 9.7 f 0.8 NS
K+ = potassium;
3.6 f 1.8 3.2 f 1.0 NS Na+
3.6 f 0.4 4.0 f 0.4
= sodium;
Cl-
=
tion of magnesium appears to be consistent with the mild to moderate functional renal insufficiency present in this group. COMMENTS
‘Hypomagnesemia developed in all 27, transplant redpients treated with cyclosporine atid prednisone. This condition was usually apparent by the second to third week after renal transplantation and was dependent upon allograft function as evidenced by a significant correlation between serum magnesium and cr&atinine concentrations. In most patients, a substantial drop in serum magnesium occurred following several days of good allograft function. In contrast, serum magriesitim levels remained normal or slightly elevated in those with poor or delayed allograft function. Since oral magnesium supplementation was routinely instituted when serum mtignesium levels fell below 1.6 mg/dl, severe hypomagnesemia was usually not encountered. However, moderate to severe hype magnesemia (serum magnesium 0.9 to 1.1 mg/dl)’ did develop in five patients, two of whom exhibited overt signs and symptoms of magnesium depletion, despite the just mentioned precautions. It is very likely that more cases of severe magnesium depletion would have occurred had not magnesium supplementation been routinely instituted. Supplemental magnesium in the form of oral magnesium oxide (750 mg daily) was generally adequate to raise serum magnesium levels to the low normal range in the majority of patients. Three patients, however, re’ quired much larger doses (1,500 mg of magnesium oxide per 24 hours). Based on our results, renal magnesium wasting was at least partially responsible for the observed hypomagnesemia as evidenced by the presence of inappropriately high urinary magnesium excretion and fractional excretion of magnesium in the face .of hypomagnesemia noted throughout the six-month study period. These findings would rule against inadequate dietaiy intake, malabsorp: tion, and/or gastrointestinal losses as the major factors contributing to hypomagnesemia since these conditidns are accompanied by renal magnesium conservation
83
HYPOMAGNESEMIA
AND
[9, lo]. It should be noted that inadequate dietary magnesium intake may have occurred during the initial week of engraftment when most of the patients were unable to eat a well-balanced diet. However, reduced magnesium intake could not be implicated after this period when the patients received balanced diets as well as magnesium supplementation. Since we did not measure fecal magnesium excretion, we cannot exclude the possibility that such losses may have also contributed to the observed hypomagnesemia. This is unlikely, however, as none of our patients had discernible evidence of gastrointestinal disease on pre-transplant evaluation nor did they exhibit anorexia, nausea, vomiting, diarrhea, or weight loss during the study period. A number of factors may be operative in the genesis of increased urinary magnesium excretion following renal transplantation. First, hypermagnesemia is known to suppress tubular reabsorption of magnesium [9]. Although the slight elevation in serum magnesium observed prior to and shortly after transplantation could have initially contributed to the increased renal magnesium excretion, it could not explain the persistent renal magnesium-wasting in the face of hypomagnesemia. In addition, osmotic diuresis associated with urinary excretion of urea and other solutes is known to suppress tubular reabsorption of magnesium and could have transiently contributed to renal magnesium wasting during the initial post-transplant period [l I]. However, persistent renal magnesium wasting and hypomagnesemia long after cessation of osmotic diuresis argue against this possibility. Other factors that could have caused renal magnesium wasting in this setting include diuretic administration, hypercalcemia, hypophosphatemia, renal tubular acidosis, and other tubular disorders often present after renal transplantation [9,12-151. Loop diuretics may have contributed to the observed renal magnesium wasting and hypomagnesemia since 11 of our cyclosporine-treated patients (45 percent) received furosemide at some point during the study period. Interestingly, however, no significant differences were found in serum magnesium, fractional excretion of magnesium, and urinary magnesium excretion rates between the patients treated with furosemide and those who did not receive the drug. Certain metabolic, electrolyte, and acid-base disturbances, including hypercalcemia, hypophosphatemia, and acidosis, have also been reported to inhibit tubular reabsorption of magnesium [9,12,14,15]. As a group, however, our patients were normocalcemic and normophosphatemic throughout the duration of the study. Furthermore, serum bicarbonate levels were generally within the normal range, thereby excluding significant metabolic acidosis. As a final consideration, the observed persistent renal magnesium wasting could be explained on the basis of some type of renal tubular injury per se. Clearly considerable potential for tubular injury exists in the engrafted
October
RENAL
MAGNESIUM
WASTING
WITH
CYCLOSPORINE-BARTON
ET AL
kidney due to warm and cold ischemia, immunologic rejection, obstruction, infection, and/or nephrotoxic drugs. Not surprisingly, a variety of tubular abnormalities have been described involving renal allografts including the Fanconi syndrome, proximal and distal renal tubular acidosis, urinary phosphate wasting, impaired ammonium production as well as abnormalities in urine concentration and dilution [ 16-201. The most commonly reported electrolyte disturbances have been hypokalemia, hypophosphatemia, and hypercalcemia [ 16,20,21]. In contrast, hypomagnesemia appears to have been quite uncommon in the pre-cyclosporine era, with only a few isolated cases reported in renal allograft recipients [22,23]. Although a mild defect in the tubular reabsorption of magnesium has been recently described in renal transplant patients treated with azathioprine and prednisone, this was not associated with hypomagnesemia [24]. In fact, serum magnesium levels were actually higher in the transplant recipients as compared with the control group, with no significant difference in urinary magnesium excretion noted between the two groups. Although the precise cause(s) of the reported reduction in the tubular reabsorption of magnesium was not known, the substantially reduced glomerular filtration rate (34.0 f 24.1 ml/minute/ 1.73 m*) and elevated serum magnesium concentrations present in the patients could have been responsible for this phenomenon. Based on the available information, serum magnesium levels are generally normal in renal allograft recipients treated with azathioprine and prednisone. Accordingly, in the study previously cited, the 36 transplant recipients had serum magnesium levels in the high normal range [24]. In addition, Ulmann et al [21] evaluated 44 patients with excellent renal allograft function (serum creatinine 1.2 mg/dl) and found serum magnesium concentrations and renal magnesium excretion rates to be normal. Moreover, in our own series of 17 renal allograft recipients treated with azathioprine and prednisone, serum magnesium and fractional excretion of magnesium measured from one week to 48 months after engraftment were 2.3 f 0.4 mg/dl and 5.0 f 2.8 percent, respectively. These values were significantly different than the corresponding measurements found in patients treated with cyclosporine (p
1987
The American
Journal
of Medicine
Volume
83
697
HYPOMAGNESEMIA
AND
RENAL
MAGNESIUM
WASTING
WITH
CYCLOSPORINE-BARTON
renal allograft recipients as well as in patients with essentially normal native kidney function. Moreover, the findings suggest renal magnesium wasting to be operative in the genesis of the observed hypomagnesemia. Although the mechanism(s) responsible for the observed renal magnesium wasting remain unknown, there is considerable evidence implicating cyclosporine as a proximal tubular toxin [28-301. Specifically, dose-related increases in urinary excretion of several proximal tubular enzymes, including N-acetyl-fi-glucosaminidase and gamma-glutamyl transferase, have been reported [30]. In addition, a number of morphologic changes have been described involving the proximal convoluted tubules, including brush border abnormalities, cytoplasmic vacuolization, dystrophic microcalcifications, and cell necrosis [28,29]. It is, therefore, tempting to speculate that cyclosporine could cause renal magnesium wasting through proximal tubular injury and impairment of magnesium reabsorption. The available data, however, do not allow for such a conclusion and, to the contrary, cyclosporine may actually enhance proximal tubular reabsorption of filtrate as evidenced by the reductions in the fractional excretion of phosphate, lithium, sodium, and potassium shown in experimental animals treated with this agent [31,32]. A more important consideration is that the thick ascending limb of Henle’s loop is the major site of magnesium reabsorption and as such may be the principal locus of action of cyclosporine on renal magnesium transport. However, the available data do not allow for localization
ET AL
within the nephron of this apparent cyclosporine-induced magnesium transport defect. In summary, renal magnesium wasting and hypomagnesemia occurred in all 27 renal allograft recipients treated with cyclosporine and prednisone included in this study. In the vast majority, hypomagnesemia was apparent within two weeks of adequate allograft function and was for the most part correctable with oral magnesium supplementation. Renal magnesium wasting persisted throughout the six-month study period. In contrast, hypomagnesemia and marked renal magnesium wasting appear to be uncommon in renal allograft recipients treated with azathioprine and prednisone. In the cyclosporinetreated patients, renal magnesium wasting occurred independently of osmotic diuresis, diuretic administration, hypercalcemia, hypophosphatemia, or acidosis. The consistent association between cyclosporine administration, renal magnesium wasting, and hypomagnesemia, in the absence of other identifiable reasons, suggests a cause and effect relationship. Further studies are required, however, to determine the mechanism and site of the apparent defect within the tubules. Likewise, the relationship between the dosage of cyclosporine and the severity of the renal magnesium transport defect needs to be investigated. Since severe symptomatic hypomagnesemia can occur in association with cyclosporine administration, we recommend routine monitoring of serum magnesium levels and early institution of supplemental magnesium therapy when indicated in patients receiving this agent.
REFERENCES 1.
2.
3.
4.
5.
6.
7.
6.
9.
698
Cohen DJ, Loertscher R, Rubin MF, et al: Cyclosporine: a new immunosuppressive agent for organ transplantation. Ann Intern Med 1984; 101: 667-682. The Canadian Multicentre Transplant Study Group: A randomized clinical trial of cyclosporine in cadaveric renal transplantation. N Engl J Med 1983; 309: 809-815. European Multicentre Trial Group: Cyclosporine in cadaveric renal transplantation: one-year follow-up of a multicentre trial. Lancet 1983; II: 986-989. Oyer PE, Stinson EB, Jamieson SW, et al: Cyclosporine in cardiac transplantation: a 2-‘/z year follow-up. Transplant Proc 1983; 15: 2546-2552. Lauparic A, for the Canadian Transplant Study Group: Complications of cyclosporine therapy-a comparison to azathioprine. Transplant Proc 1983; 15: 2748-2753. Joss DV, Barrett AJ, Kendra JR, Lucas CF, Desai S: Hypertension and convulsions in children receiving cyclosporin A (letter). Lancet 1982; I: 906. Starzl TE, Klintmalm GB, Weil R Ill, et al: Cyclosporine A and steroid therapy in sixty-six cadaver kidney transplants. Surg Gynecol Obstet 198 1; 153: 486-494. Klintmalm GB, lwatsuki S, Starzl TE: Cyclosporine A hepatotoxicity in 66 renal allograft recipients. Transplantation 1981; 32: 488-489. Dirks JH: The kidney in magnesium regulation. Kidney Int 1983; 23: 771-777.
October
1987
The
American
Journal
of Medicine
Volume
10.
11.
12.
13.
14.
15.
16. 17. 18.
19.
83
Carney SL, Wong NLM, Quamme GA, Dirks JH: Effects of magnesium deficiency on renal magnesium and calcium transport in the rat. J Clin Invest 1980; 65: 180-188. Wong NLM, Quamme GA, Dirks JH: Effect of urea on electrolyte transport in the dog kidney. J Lab Clin Med 1981; 98: 741-749. Shareghi GT, Agus ZS: Magnesium transport in the cortical thick ascending limb of Henle’s loop of the rabbit. J Clin Invest 1982; 69: 759-769. Duarte CG: Effects of ethacrynic acid and furosemide on urinary calcium, phosphorous and magnesium. Metabolism 1968; 17: 867-876. Massry SG, Coburn JW: The hormonal and non-hormonal control of renal excretion of calcium and magnesium. Nephron 1973; 10: 66-112. Wong NLM, Quamme GA, Dirks JH: The effect of acidosis and alkalosis on renal handling of magnesium. J Clin Invest Med 1982; 5: 443. Better OS: Tubular dysfunction following kidney transplantation. Nephron 1980; 25: 209-213. Wilson DR, Siddigui AA: Renal tubular acidosis after kidney transplantation. Ann Intern Med 1975; 79: 352-361. Vertuno LL, Preuss HG, Argy WP, Schrier GE: Fanconi syndrome following homotransplantation. Arch Intern Med 1974; 133: 302-305. DeFronzo RA, Goldberg M, Cook CR, et al: Investigation into
HYPOMAGNESEMIA
20.
21.
22.
23.
24.
25.
28.
AND
RENAL
the mechanism of hyperkalemia following renal transplantation. Kidney Int 1977; 11: 357-384. Vaziri ND, Nellans RE, Brueggemann RM, Barton CH, Martin DC: Renal tubular dysfunction in transplanted kidneys. South Med J 1979; 72: 530-537. Ulmann A, Chekoff N, Lacour B: Disorders of calcium and phosphorous metabolism after successful kidney transplantation. Adv Nephrol 1983; 12: 331-340. Davis BB, Preuss HG, Murdaugh V: Hypomagnesemia following the diuresis of post-renal obstruction and renal transplant. Nephron 1975; 14: 275280. Revusova V, Zvara V, Gratzlova J, et al: Magnesium deficit after renal transplantation with secondary post-transplantation hypocalcemia and hypophosphatemia. Urol Int 1975; 30: 313-320. Revusova V, Gratzlova J, Zvara V, et al: Impaired tubular reabsorption of magnesium after renal transplantation. Int Urol Nephrol 1984; 18: 77-82. Thompson CB, June CH, Sullivan KM, Thomas ED: Association between cyclosporine neurotoxicity and hypomagnesemia. Lancet 1984; II: 1118-1120. June CH, Thompson CB, Kennedy MS, et al: Correlation of hypomagnesemia with the onset of cyclosporine-associ-
October
27.
28.
29.
30.
31.
32.
1987
MAGNESIUM
WASTING
WITH
CYCLOSPORINE-BARTON
ET AL
ated hypertension in marrow transplant patients. Transplantation 1988; 41: 47-51. June CH, Thompson CB, Kennedy MS, et al: Profound hypomagnesemia and renal magnesium wasting associated with the use of cyclosporine for marrow transplantation. Transplantation 1985; 39: 820-824. Mihatsch M, Oliveri W, Marbet UA, et al: Giant mitochondria in renal tubular cells and cyclosporine A. Lancet 1981; 1: 1182-1183. Simonton SC, Rynasiewicz J, Sibley RK: Light microscopic and electron microscopic features of experimental cyclosporine A nephrotoxicity. Lab Invest 1983; 43: 78A-79A. Devineni R, McKenzie N, Duplan J, et al: Renal effects of cyclosporine: clinical and experimental observations. Transplant Proc 1983; 15: 28952898. Tonnesen AS, Hamner RW, Weinmann EJ: Cyclosporine and sodium and potassium excretion in the rat. Transplant Proc 1983; 15: 2730-2735. Dieperink H, Starklint H, Leyssac PP: Nephrotoxicity of cyclosporine an animal model: study of the nephrotoxic effect of cyclosporine on overall renal and tubular function in conscious rats. Transplant Proc 1983; 15: 27382741.
The American
Journal
of Medicine
Volume
83
699
CYRIL H. BARTON, M.D. NOSRATOLA D. VAZIRI, M.D. DONALD C. MARTIN, M.D. SUNNY CHOI, M.D. SHAHRIAR ALIKHANI Irvine, California
From the Divisions of Nephrology and Urology, University of California, Irvine, Irvine, California. Requests for reprints should be addressed to Dr. Cyril H. Barton, Division of Nephrology, University of California, Irvine Medical Center, 101 The City Drive, Orange, California 92668. Manuscript submitted July 7, 1986, and accepted May 15, 1987.
Following the adoption and use of cyclosporine as the drug of choice in the management of renat allograft recipients, several,cases of symp tomatic hypomagnesemia were noted. These observations prompted the current prospective study of serum concentration and urinary excretion of magnesium in 27 renal transplant recipients treated with cyclosporine and prednisone. Relevant laboratory measurements were obtained shortly before and regularly after transplantation. The results were compared with those obtained in a group of 17 allograft recipients treated with azathioprine and prednisone. The cyclosporine-treated patients showed a significant reduction In the serum magnesium concentratlon and an inappropriately increased urinary excretion and fractional excretion of magnesium, suggesting renal magnesium wasting. The observed hypomagnesemia required magnesium supplemen; tation in nearly all cyclosporine-treated patients. In contrast, azathioprine-treated patients showed normal serum magnesium concentrations and required no magnesium supplementation. In conclusion, administration of cyclosporine in renal allograft recipients appears to be commonly associated with renal magnesium wasting and hypomag nesemia. Therefore, Wis recommended that serum levels of magnesium be monitored regularly in renal allograft recipients receiving cyclosporine and that magnesium supplementation be employed as needed to avoid magnesium depletion. Currently, cyclosporine is widely used in the management of patients who have undergone organ transplantation. Although use of cyclosporine affords unquestionable benefits such as improved allograft survival and reduced steroid requirement, it is associated with various toxic effects. Nephrotoxicity with functional renal insufficiency remains the major adverse effect of this drug. Other side effects and complications include infection, lymphoma, hypertension, hyperkalemia, seizures, gingival hyperplasia, hirsutism, and hepatotoxicity [l-8]. We have observed symptomatic hypomagnesemia in several renal allograft recipients treated with cyclosporine but in none of those treated with azathioprine. We, therefore, undertook the current study to examine the possible association between cyclosporine and magnesium depletion in renal allograft recipients treated with cyclosporine and prednisone. The results were compared with those obtained in allograft recipients given azathioprine and prednisone. PATIENTS AND METHODS Cyclosporlne-Treated Group. Between July of 1984 and December of 1985, 42 renal transplant procedures were performed at our institution in which cyclosporine and prednisone were used for immunosuppression.
October
1987
The American
Journal
of Medicine
Volume
83
693
HYPOMAGNESEMIA
AND
RENAL
MAGNESIUM
WASTING
WITH
CYCLOSPORINE-BARTON
We were unable to include 15 of these patients in our study because of early allograft failure (three patients), death (three patients), and loss of follow-up (nine patients). Of the 27 allograft recipients who were studied, 25 were cadaveric, whereas two were living related donors. Sixteen were men and 11 were women, with an age range of 23 to 63 years and a mean age of 44 years. The cause of end-stage renal disease was chronic glomerulopathy (eight patients), diabetic nephropathy (seven patients), nephrosclerosis (five patients), polycystic kidney disease (two patients), chronic interstitial nephropathy (two patients), and renal failure of unknown etiology (three patients). Prior to renal transplantation, 25 patients had been treated with long-term hemodialysis, whereas two patients were managed on continuous ambulatory peritoneal dialysis. lmmunosuppression was accomplished using cyclosporine and prednisone. A dose of 5 mg/kg per day of cyclosporine was administered intravenously for the first three days of engraftment followed by orally administered cyclosporine (10 mg/kg per day). The oral dose was rapidly tapered so that by the sixth week most patients were receiving approximately 5 mg/kg per day with individual adjustments being made to maintain trough serum levels between 60 and 200 rig/ml. Patients were also treated with intravenously administered methyl prednisolone, 1,000 mg per day in four divided doses, for three days followed by a rapidly tapering schedule of orally administered prednisone so that by two to three months after engraftment most patients were receiving 15 to 20 mg per day. Other medications included the routine use of oral antacids (aluminum hydroxide or aluminum phosphate gel) or ranitidine during periods of high-dose steroid administration. Antihypertensive medications including cionidine, hydralazine, and/or propranoioi were used when indicated to control blood pressure. Eleven patients (41 percent) also received furosemide as an adjunct in the management of hypertension. Since furosemide is known to cause renal magnesium-wasting, data from these patients were compared with those obtained in the remaining 16 patients who did not receive furosemide. Arathioprine-Treated Group. Seventeen patients (11 cadaveric and six living related donors) who underwent transplantation at our institution and were treated with azathioprine and prednisone were included for comparison. Eleven were men and six were women, with an age range of 25 to 51 years and a mean age of 36 years. The cause of endstage renal disease was chronic glomeruiopathy (six patients), diabetic nephropathy (three patients), nephroscierosis (one patient), chronic interstitial nephritis (two patients), Aiport’s disease (one patient), lupus erythematosis (one patient), and renal failure of unknown etiology (three patients). Prior to transplantation, 16 patients had been treated with hemodlaiysis, and one patient was managed with continuous ambulatory peritoneal dialysis. Eight patients (47 percent) were receiving furosemide in addition to other antihypertensive medications including clonidine, hydraiazine, and/or propranoloi. None of the patients was receiving any magnesiu.in-containing preparations. Although all patients were studied prospectively, only three could be entered into the study at the time of engraftment. We were not able to study a larger number of azathioprine-treated patients in a parallel fashion due to the fact
694
October
1967
The American
Journal
of Medicine
Volume
ET AL
that the vast majority of patients who received transplants at our institution after July of 1964 were given cyciosporine to achieve immunosuppression. The additional I4 azathioprine-treated patients were selected from our clinic popuiai tion and represent a group of consecutive recipients witii good aiiograft function who underwent transplantation prior to the introduction of cyciosporine at our center. The mean duration of engraftment of this group at the time of entering the study was 23 weeks, while the mean serum creatinine level was 2.1 mg/dl. Control Group. The control group consisted of 30 normal volunteers, 16 women and 12 men, with a mean age of 39 years. None of the participants in the control group was taking any drugs or receiving magnesium supplementation. Mid-morning serum and urine samples were obtained for measurement of magnesium and creatinine concentration. Methods. In the cyclosporine group, baseline laboratory measurements were obtained shortly prior to engraftment and were repeated at weekly intervals for eight weeks and at monthly intervals for four months after transplantation. In the azathioprine group, laboratory measurements were also obtained at weekly intervals for eight weeks and at monthly intervals for four months; however, in only three of the 17 patients could baseline measurements be obtained. Blood specimens were drawn in the morning, after approximately 10 to 12 hours of fasting. Serum concentrations of creatinine, sodium, chloride, potassium, bicarbonate, caiciurn, inorganic phosphorus, and albumin were determined using standard methods, while serum magnesium levels were measured using a DuPont automatic clinical analyzer. Plasma trough levels of cyciosporine were measured using radioimmunoassay approximately 12 hours after administration of the last dose, with blood specimens being kept at 37’C prior to plasma separation. Both timed and spot urine collections were obtained for measurement of creatinine and magnesium concentrations. The urine magnesium concentration was measured using atomic absorption. Twentyfour-hour urinary magnesium excretion was calculated from the urine concentration of magnesium and 24-hour urine volume. Fractional excretion of magnesium (Feyg) was calculated using the following formula: FeMg = UMg/SMg
f UCr/SCr
X 100
where UMg and SMg represent urine,and serum magnesium concentrations, respectively, while UCr and SCr represent urine and serum creatinine concentrations, respectively. Since approximately 20 percent of serum magnesium is protein bound and not ultrafilterable, using the serum magnesium level in the foregoing formula slightly underestimates true fractional magnesium excretion. This relatively small systematic error, however, is common to both groups and all collection periods and as such should not affect the significance of the observed changes. ‘” Data Expression and Analysis. The data are presented as mean f SD. The Student t test and regression analysis were used where appropriate. Changeover time in the serum and urine measurements was evaluated by means of a repeated measures, multiple analysis of variance in order to determine the significance of any trend. p Values less than 0.05 were considered to be significant.
63
HYPOMAGNESEMIA
TABLE I
Laboratory
Data in Serum Cyclosporine-Treated
Serum Magnesium* (mg/W
Patients Baseline Week 1 Week 2 Week 3 Week 4 Week 5 Week 6 Week 7 Week 6 Month 3 Month 4 Month 5 Month 6
2.6 2.1 1.6 1.6 1.6 1.7 1.7 1.7 1.6 1.6 1.6 1.5 1.7
f f f f f f f f f f f f f
AND RENAL
Serum Creatinine’ (WdU
0.4 0.4 0.4 0.2 0.3 0.3 0.4 0.3 0.3 0.3 0.1 0.2 0.3
10.7 f 4.0 f 3.1 f 2.4 f 2.3 f 2.3 f 2.4 f 2.1 f 2.1 f 2.0 f 1.6f0.7 2.2 f 2.0 f
NA = not available. * p
Urinary Magnesium (mg124 hours)
3.0 2.5 1.5 1.0 1.4 1.1 1.4 1.0 0.6 0.7
NA 115f75 92f 66 f 114f41 147 f 101 f 146 f 101 f 95 f 103f 120 f 125 f
0.7 0.7
were evaluated trends.
by a multiple
The mean serum and urine values collected over month study period in the 27 cyclosporine-treated are
shown
magnesium to engraftment, ease. nesium
However, levels
Table I. The
in
level
of 2.6 f 0.4 is characteristic
by Week dropped to
elevated
the
baseline
mg/dl, measured of end-stage
2 of engraftment, 1.8 f 0.4 mg/dl
analysis
prior dis-
of
with
in
baseline
values.
serum magnesium supplementation (750 to 1,500 mg further significant concentration majority of levels fell magnesium whom pletion increase
the
To
prevent
below 1.6 was initiated per day) reduction
occurred patients.
as required. in mean beyond the However,
to values between supplementation
exhibited (muscle doses
further mg/dl, using
oral magnesium magnesium oxide Consequently, serum magnesium third serum
0.9 and in five
signs and symptoms weakness and tremors) of magnesium oxide.
reductions
no
despite two of
of magnesium that responded The percentage
deto of
cyclosporine-treated patients requiring supplemental magnesium oxide at given intervals during the study is shown in Table I. As can be seen, by the end of five to six months,
virtually
all patients
m?gnesium supplementation. The urinary magnesium tion of magnesium in the
were
requiring
excretion and cyclosporine-treated
and fr@ional
receiving
renal magnesium-wasting. correlation between
There the
serum
pendency function
was magne-
October
NA 379 f 294 f 453 f 367 f 299 f 165 f 196f70 202 f 266 f 365 f 169 f 251 f
repeated-measure
ET AL
of and
fluctuations throughout
in the
with
significant concentration
of calcium, inorganic chloride, bicarbonate,
97 161 146 56 170
the
p Values
less than 0.01
0.56, p the de-
hypomagnesemia on renal magnesium-wasting.
renal
correlation between the and serum concentrations
phosphorus, or albumin.
serum study,
NA 11 11 26 46 60 62 56 65 67 64 95 96
concentrations (r = further demonstrates
the observed is consistent no
Percent Receiving Magnesium Oxide
254 330 264 260 196 105
design.
serum creatinine This association
There was magnesium
se-
potassium, sodium, Although considerable
cyclosporine somewhat
levels higher
were noted levels present
during the early weeks of engraftment no doubt reflect the higher dosages of cyclosporine administered at that time. Our data, however, show no significant correlation bethe
serum
levels
of cyclosporine
renal magnesium-wasting A comparison of the who did not did receive there were subgroups fractional cate not
receive furosemide this drug is shown no significant in terms of serum
excretion only that
TABLE II
with in
the
Table
differences magnesium
of magnesium. a factor (or factors)
the degree
of
patients
11 patients who II. Surprisingly, between the concentration
These other
two or
findings indithan furose-
hypomagnesemia were not substan-
Comparison of Serum and Urinary Magnesium Values In CyclosporlneTreated Patients Receiving and Not Receiving Furosemide No Furosemide
Serum magnesium @W-W Fractional excretion of magnesium (percent) Serum creatinine (mg/di)
1987
and
and/or hypomagnesemia. 16 cyclosporine-treated
mide was responsible for the observed but also that the effects of furosemide
excrepatients at
the end of Week 1 were 115 f 75 mg per 24 hours and 15 f 8.0 percent of filtered magnesium load, respectively. These values remained inappropriately high throughout the study period despite persistent hypomagnesemia, which suggests also a significant
sium and
tween
week in the magnesium
1.1 mg/dl patients,
CYCLOSPORINE-BARTON
Serum Cyc!osporine Wml)
6.0 6.7 7.7 8.7 6.7 6.5 5.1 4.0 4.6 7.0 9.9 a.7
of variance,
rum
serum magand by the end
NA f f f f f f f f f f f f
15.0 12.6 13.0 14.2 15.2 13.5 14.2 9.7 10.1 15.6 14.5 13.3
52 32 39 14 a 11 31 42
sixpa-
of Week 3, frank hypomagnesemia was present (levels 1.6 f 0.2 mg/dl). These changes in serum magqesium levels were highly significant (p
WITH
Fractidnal Excretion (percent)
19 34
serum just renal
WASTING
Patlents
RESULTS
tients
MAGNESIUM
The American
Journal
Furosemide
1.6 f
0.4
1.7f0.3
16.1 f
6.0
14.0 f
2.7 f
1.6
2.4fl.O
of Medicine
Volume
p Values NS
6.9
NS NS
83
696
HYPOMAGNESEMIA
TABLE III
AND
RENAL
Comparison
MAGNESIUM
of Relevant
sM9 (m9W
Groups Cyclosporine-treated Azathioprine-treated p Values
1.7 f 0.3 2.3 f 0.4
WASTING
WITH
Laboratory
FeMg (percent) 13.5 f 7.0 5.0 f 2.8
scr (mgldl) 2.6 f 1.5 2.1 f 1.0 NS
CYCLOSPORINE-BARTON
Values
ET AL
In Cyclosporlne-Treated
(me$ter)
(rnzkr
)
4.5 f 0.6 4.5 f 0.7 NS
142 f 3.5 142 f 5.2 NS
(me:iter
696
October
1967
The American
Journal
of Medicine
Volume
)
107 f 8.0 104 f 7.9 NS
sMg = serum magnesium; FeMg = fractional excretion of magnesium; sCr = serum chloride; HC03= bicarbonate; Ca = calcium; Pi = phosphorus. * Cumulative mean values collected over a six-month study period are shown.
tially additive. It should be pointed out, however, that approximately half of the patients included in the furosemide-treated subgroup were only intermittently receiving this agent, which may explain at least in part why more severe hypomagnesemia was not observed. In contrast to the findings in the cyclosporine-treated patients, serum magnesium levels in the 17 azathioprinetreated patients were at the upper limits of normal (normal range for serum magnesium concentration reported by our laboratory is 1.7 to 2.4 mg/dl) at 2.3 f 0.4 mg/dl during the six-month observation period. Furthermore, none of the azathioprine-treated patients, including the three who entered the study at the time of engraftment, exhibited serum magnesium levels below 1.7 mg/dl, and none required or received magnesium supplementation. Table III compares the cumulative serum and urine measurements in the azathioprine and cyclosporine groups. It can be seen that serum magnesium levels were significantly higher while the fractional excretion of magnesium was significantly lower in the azathioprine-treated group as compared with the cyclosporine-treated group. Overall, these findings show no evidence for renal magnesium wasting or hypomagnesemia in the patients treated with azathioprine. The serum magnesium concentration in our 30 normal control subjects was 2.12 f 0.17 mg/dl. This value is significantly higher than the mean serum magnesium concentration in the cyclosporine group for the six-month study period (p
versus Arathloprlne-Treated
creatinine;
(n!$&
Groups* Albumin (Wdl)
(miydl)
24 f 3.9 28 f 5.7
9.9 f 1.4 9.7 f 0.8 NS
K+ = potassium;
3.6 f 1.8 3.2 f 1.0 NS Na+
3.6 f 0.4 4.0 f 0.4
= sodium;
Cl-
=
tion of magnesium appears to be consistent with the mild to moderate functional renal insufficiency present in this group. COMMENTS
‘Hypomagnesemia developed in all 27, transplant redpients treated with cyclosporine atid prednisone. This condition was usually apparent by the second to third week after renal transplantation and was dependent upon allograft function as evidenced by a significant correlation between serum magnesium and cr&atinine concentrations. In most patients, a substantial drop in serum magnesium occurred following several days of good allograft function. In contrast, serum magriesitim levels remained normal or slightly elevated in those with poor or delayed allograft function. Since oral magnesium supplementation was routinely instituted when serum mtignesium levels fell below 1.6 mg/dl, severe hypomagnesemia was usually not encountered. However, moderate to severe hype magnesemia (serum magnesium 0.9 to 1.1 mg/dl)’ did develop in five patients, two of whom exhibited overt signs and symptoms of magnesium depletion, despite the just mentioned precautions. It is very likely that more cases of severe magnesium depletion would have occurred had not magnesium supplementation been routinely instituted. Supplemental magnesium in the form of oral magnesium oxide (750 mg daily) was generally adequate to raise serum magnesium levels to the low normal range in the majority of patients. Three patients, however, re’ quired much larger doses (1,500 mg of magnesium oxide per 24 hours). Based on our results, renal magnesium wasting was at least partially responsible for the observed hypomagnesemia as evidenced by the presence of inappropriately high urinary magnesium excretion and fractional excretion of magnesium in the face .of hypomagnesemia noted throughout the six-month study period. These findings would rule against inadequate dietaiy intake, malabsorp: tion, and/or gastrointestinal losses as the major factors contributing to hypomagnesemia since these conditidns are accompanied by renal magnesium conservation
83
HYPOMAGNESEMIA
AND
[9, lo]. It should be noted that inadequate dietary magnesium intake may have occurred during the initial week of engraftment when most of the patients were unable to eat a well-balanced diet. However, reduced magnesium intake could not be implicated after this period when the patients received balanced diets as well as magnesium supplementation. Since we did not measure fecal magnesium excretion, we cannot exclude the possibility that such losses may have also contributed to the observed hypomagnesemia. This is unlikely, however, as none of our patients had discernible evidence of gastrointestinal disease on pre-transplant evaluation nor did they exhibit anorexia, nausea, vomiting, diarrhea, or weight loss during the study period. A number of factors may be operative in the genesis of increased urinary magnesium excretion following renal transplantation. First, hypermagnesemia is known to suppress tubular reabsorption of magnesium [9]. Although the slight elevation in serum magnesium observed prior to and shortly after transplantation could have initially contributed to the increased renal magnesium excretion, it could not explain the persistent renal magnesium-wasting in the face of hypomagnesemia. In addition, osmotic diuresis associated with urinary excretion of urea and other solutes is known to suppress tubular reabsorption of magnesium and could have transiently contributed to renal magnesium wasting during the initial post-transplant period [l I]. However, persistent renal magnesium wasting and hypomagnesemia long after cessation of osmotic diuresis argue against this possibility. Other factors that could have caused renal magnesium wasting in this setting include diuretic administration, hypercalcemia, hypophosphatemia, renal tubular acidosis, and other tubular disorders often present after renal transplantation [9,12-151. Loop diuretics may have contributed to the observed renal magnesium wasting and hypomagnesemia since 11 of our cyclosporine-treated patients (45 percent) received furosemide at some point during the study period. Interestingly, however, no significant differences were found in serum magnesium, fractional excretion of magnesium, and urinary magnesium excretion rates between the patients treated with furosemide and those who did not receive the drug. Certain metabolic, electrolyte, and acid-base disturbances, including hypercalcemia, hypophosphatemia, and acidosis, have also been reported to inhibit tubular reabsorption of magnesium [9,12,14,15]. As a group, however, our patients were normocalcemic and normophosphatemic throughout the duration of the study. Furthermore, serum bicarbonate levels were generally within the normal range, thereby excluding significant metabolic acidosis. As a final consideration, the observed persistent renal magnesium wasting could be explained on the basis of some type of renal tubular injury per se. Clearly considerable potential for tubular injury exists in the engrafted
October
RENAL
MAGNESIUM
WASTING
WITH
CYCLOSPORINE-BARTON
ET AL
kidney due to warm and cold ischemia, immunologic rejection, obstruction, infection, and/or nephrotoxic drugs. Not surprisingly, a variety of tubular abnormalities have been described involving renal allografts including the Fanconi syndrome, proximal and distal renal tubular acidosis, urinary phosphate wasting, impaired ammonium production as well as abnormalities in urine concentration and dilution [ 16-201. The most commonly reported electrolyte disturbances have been hypokalemia, hypophosphatemia, and hypercalcemia [ 16,20,21]. In contrast, hypomagnesemia appears to have been quite uncommon in the pre-cyclosporine era, with only a few isolated cases reported in renal allograft recipients [22,23]. Although a mild defect in the tubular reabsorption of magnesium has been recently described in renal transplant patients treated with azathioprine and prednisone, this was not associated with hypomagnesemia [24]. In fact, serum magnesium levels were actually higher in the transplant recipients as compared with the control group, with no significant difference in urinary magnesium excretion noted between the two groups. Although the precise cause(s) of the reported reduction in the tubular reabsorption of magnesium was not known, the substantially reduced glomerular filtration rate (34.0 f 24.1 ml/minute/ 1.73 m*) and elevated serum magnesium concentrations present in the patients could have been responsible for this phenomenon. Based on the available information, serum magnesium levels are generally normal in renal allograft recipients treated with azathioprine and prednisone. Accordingly, in the study previously cited, the 36 transplant recipients had serum magnesium levels in the high normal range [24]. In addition, Ulmann et al [21] evaluated 44 patients with excellent renal allograft function (serum creatinine 1.2 mg/dl) and found serum magnesium concentrations and renal magnesium excretion rates to be normal. Moreover, in our own series of 17 renal allograft recipients treated with azathioprine and prednisone, serum magnesium and fractional excretion of magnesium measured from one week to 48 months after engraftment were 2.3 f 0.4 mg/dl and 5.0 f 2.8 percent, respectively. These values were significantly different than the corresponding measurements found in patients treated with cyclosporine (p
1987
The American
Journal
of Medicine
Volume
83
697
HYPOMAGNESEMIA
AND
RENAL
MAGNESIUM
WASTING
WITH
CYCLOSPORINE-BARTON
renal allograft recipients as well as in patients with essentially normal native kidney function. Moreover, the findings suggest renal magnesium wasting to be operative in the genesis of the observed hypomagnesemia. Although the mechanism(s) responsible for the observed renal magnesium wasting remain unknown, there is considerable evidence implicating cyclosporine as a proximal tubular toxin [28-301. Specifically, dose-related increases in urinary excretion of several proximal tubular enzymes, including N-acetyl-fi-glucosaminidase and gamma-glutamyl transferase, have been reported [30]. In addition, a number of morphologic changes have been described involving the proximal convoluted tubules, including brush border abnormalities, cytoplasmic vacuolization, dystrophic microcalcifications, and cell necrosis [28,29]. It is, therefore, tempting to speculate that cyclosporine could cause renal magnesium wasting through proximal tubular injury and impairment of magnesium reabsorption. The available data, however, do not allow for such a conclusion and, to the contrary, cyclosporine may actually enhance proximal tubular reabsorption of filtrate as evidenced by the reductions in the fractional excretion of phosphate, lithium, sodium, and potassium shown in experimental animals treated with this agent [31,32]. A more important consideration is that the thick ascending limb of Henle’s loop is the major site of magnesium reabsorption and as such may be the principal locus of action of cyclosporine on renal magnesium transport. However, the available data do not allow for localization
ET AL
within the nephron of this apparent cyclosporine-induced magnesium transport defect. In summary, renal magnesium wasting and hypomagnesemia occurred in all 27 renal allograft recipients treated with cyclosporine and prednisone included in this study. In the vast majority, hypomagnesemia was apparent within two weeks of adequate allograft function and was for the most part correctable with oral magnesium supplementation. Renal magnesium wasting persisted throughout the six-month study period. In contrast, hypomagnesemia and marked renal magnesium wasting appear to be uncommon in renal allograft recipients treated with azathioprine and prednisone. In the cyclosporinetreated patients, renal magnesium wasting occurred independently of osmotic diuresis, diuretic administration, hypercalcemia, hypophosphatemia, or acidosis. The consistent association between cyclosporine administration, renal magnesium wasting, and hypomagnesemia, in the absence of other identifiable reasons, suggests a cause and effect relationship. Further studies are required, however, to determine the mechanism and site of the apparent defect within the tubules. Likewise, the relationship between the dosage of cyclosporine and the severity of the renal magnesium transport defect needs to be investigated. Since severe symptomatic hypomagnesemia can occur in association with cyclosporine administration, we recommend routine monitoring of serum magnesium levels and early institution of supplemental magnesium therapy when indicated in patients receiving this agent.
REFERENCES 1.
2.
3.
4.
5.
6.
7.
6.
9.
698
Cohen DJ, Loertscher R, Rubin MF, et al: Cyclosporine: a new immunosuppressive agent for organ transplantation. Ann Intern Med 1984; 101: 667-682. The Canadian Multicentre Transplant Study Group: A randomized clinical trial of cyclosporine in cadaveric renal transplantation. N Engl J Med 1983; 309: 809-815. European Multicentre Trial Group: Cyclosporine in cadaveric renal transplantation: one-year follow-up of a multicentre trial. Lancet 1983; II: 986-989. Oyer PE, Stinson EB, Jamieson SW, et al: Cyclosporine in cardiac transplantation: a 2-‘/z year follow-up. Transplant Proc 1983; 15: 2546-2552. Lauparic A, for the Canadian Transplant Study Group: Complications of cyclosporine therapy-a comparison to azathioprine. Transplant Proc 1983; 15: 2748-2753. Joss DV, Barrett AJ, Kendra JR, Lucas CF, Desai S: Hypertension and convulsions in children receiving cyclosporin A (letter). Lancet 1982; I: 906. Starzl TE, Klintmalm GB, Weil R Ill, et al: Cyclosporine A and steroid therapy in sixty-six cadaver kidney transplants. Surg Gynecol Obstet 198 1; 153: 486-494. Klintmalm GB, lwatsuki S, Starzl TE: Cyclosporine A hepatotoxicity in 66 renal allograft recipients. Transplantation 1981; 32: 488-489. Dirks JH: The kidney in magnesium regulation. Kidney Int 1983; 23: 771-777.
October
1987
The
American
Journal
of Medicine
Volume
10.
11.
12.
13.
14.
15.
16. 17. 18.
19.
83
Carney SL, Wong NLM, Quamme GA, Dirks JH: Effects of magnesium deficiency on renal magnesium and calcium transport in the rat. J Clin Invest 1980; 65: 180-188. Wong NLM, Quamme GA, Dirks JH: Effect of urea on electrolyte transport in the dog kidney. J Lab Clin Med 1981; 98: 741-749. Shareghi GT, Agus ZS: Magnesium transport in the cortical thick ascending limb of Henle’s loop of the rabbit. J Clin Invest 1982; 69: 759-769. Duarte CG: Effects of ethacrynic acid and furosemide on urinary calcium, phosphorous and magnesium. Metabolism 1968; 17: 867-876. Massry SG, Coburn JW: The hormonal and non-hormonal control of renal excretion of calcium and magnesium. Nephron 1973; 10: 66-112. Wong NLM, Quamme GA, Dirks JH: The effect of acidosis and alkalosis on renal handling of magnesium. J Clin Invest Med 1982; 5: 443. Better OS: Tubular dysfunction following kidney transplantation. Nephron 1980; 25: 209-213. Wilson DR, Siddigui AA: Renal tubular acidosis after kidney transplantation. Ann Intern Med 1975; 79: 352-361. Vertuno LL, Preuss HG, Argy WP, Schrier GE: Fanconi syndrome following homotransplantation. Arch Intern Med 1974; 133: 302-305. DeFronzo RA, Goldberg M, Cook CR, et al: Investigation into
HYPOMAGNESEMIA
20.
21.
22.
23.
24.
25.
28.
AND
RENAL
the mechanism of hyperkalemia following renal transplantation. Kidney Int 1977; 11: 357-384. Vaziri ND, Nellans RE, Brueggemann RM, Barton CH, Martin DC: Renal tubular dysfunction in transplanted kidneys. South Med J 1979; 72: 530-537. Ulmann A, Chekoff N, Lacour B: Disorders of calcium and phosphorous metabolism after successful kidney transplantation. Adv Nephrol 1983; 12: 331-340. Davis BB, Preuss HG, Murdaugh V: Hypomagnesemia following the diuresis of post-renal obstruction and renal transplant. Nephron 1975; 14: 275280. Revusova V, Zvara V, Gratzlova J, et al: Magnesium deficit after renal transplantation with secondary post-transplantation hypocalcemia and hypophosphatemia. Urol Int 1975; 30: 313-320. Revusova V, Gratzlova J, Zvara V, et al: Impaired tubular reabsorption of magnesium after renal transplantation. Int Urol Nephrol 1984; 18: 77-82. Thompson CB, June CH, Sullivan KM, Thomas ED: Association between cyclosporine neurotoxicity and hypomagnesemia. Lancet 1984; II: 1118-1120. June CH, Thompson CB, Kennedy MS, et al: Correlation of hypomagnesemia with the onset of cyclosporine-associ-
October
27.
28.
29.
30.
31.
32.
1987
MAGNESIUM
WASTING
WITH
CYCLOSPORINE-BARTON
ET AL
ated hypertension in marrow transplant patients. Transplantation 1988; 41: 47-51. June CH, Thompson CB, Kennedy MS, et al: Profound hypomagnesemia and renal magnesium wasting associated with the use of cyclosporine for marrow transplantation. Transplantation 1985; 39: 820-824. Mihatsch M, Oliveri W, Marbet UA, et al: Giant mitochondria in renal tubular cells and cyclosporine A. Lancet 1981; 1: 1182-1183. Simonton SC, Rynasiewicz J, Sibley RK: Light microscopic and electron microscopic features of experimental cyclosporine A nephrotoxicity. Lab Invest 1983; 43: 78A-79A. Devineni R, McKenzie N, Duplan J, et al: Renal effects of cyclosporine: clinical and experimental observations. Transplant Proc 1983; 15: 28952898. Tonnesen AS, Hamner RW, Weinmann EJ: Cyclosporine and sodium and potassium excretion in the rat. Transplant Proc 1983; 15: 2730-2735. Dieperink H, Starklint H, Leyssac PP: Nephrotoxicity of cyclosporine an animal model: study of the nephrotoxic effect of cyclosporine on overall renal and tubular function in conscious rats. Transplant Proc 1983; 15: 27382741.
The American
Journal
of Medicine
Volume
83
699