Switch to a sirolimus-based immunosuppression in long-term renal transplant recipients: reduced rate of (pre-)malignancies and nonmelanoma skin cancer in a prospective, randomized, assessor-blinded, controlled clinical trial.

Renal transplant recipients (RTR) have a 50-200-fold higher risk for nonmelanoma-skin cancer (NMSC) causing high rates of morbidity and sometimes mortality. Cohort-studies gave evidence that a sirolimus-based immunosuppression may inhibit skin tumor growth. This single-center, prospective, assessor-blinded, randomized trial investigated if switching to sirolimus treatment inhibits the progression of premalignancies and moreover how many new NMSC occur compared to continuation of the original immunosuppressive therapy. Forty-four RTR (mean age 59.9 years, mean duration of immunosuppression 229.5 months) with skin lesions were randomized to sirolimus or continuation of their original immunosuppression. Blinded dermatological assessment at month 6 and 12 by the same dermatologist evaluated the clinical change compared to baseline. Biopsy was performed in suspected malignancy. Already the 6-month-assessment showed significant superiority of sirolimus-therapy: a stop of progression, even regression of preexisting premalignancies (p < 0.0005). This effect was increased at month 12 (p < 0.0001). Nine patients developed histologically confirmed NMSC: one in the sirolimus group, eight in the control group, p = 0.0176. Sirolimus-based immunosuppression in RTR, even when established many years after transplantation, can delay the development of premalignancies, induce regression of preexisting lesions and decelerate the incidence of new NMSC.

Postoperative voiding dysfunction in older male renal transplant recipients.

OBJECTIVE: To evaluate the incidence of voiding dysfunction in older male renal transplant recipients. PATIENTS AND METHODS: Data for 103 patients aged 60 years or older (mean age, 65.7 years; group 1) who underwent transplantation at our center between January 1999 and August 2007 were compared with data for a group of 139 younger patients (mean age, 50.1 years; group 2) treated within the same time frame. RESULTS: Postoperatively, 28 group 1 recipients (27%) and 26 group 2 recipients (19%) experienced voiding dysfunction after removal of the transurethral catheter (P = .12). The most common cause was bladder outlet obstruction due to benign prostatic hyperplasia in 26 patients in group 1 (25%) and 17 patients in group 2 (12%) (P = .009). Bladder neck contracture, urethral stricture, and detrusor underactivity were diagnosed in the other patients. Transurethral resection of the prostate gland was performed in 21 group 1 patients (20%) and 14 group 2 patients (10%) (P = .02) at a mean of 31.1 and 29.5 days, respectively (P = .23) after transplantation. Surgical procedures were performed without complication, and symptoms did not recur postoperatively. CONCLUSIONS: Our data reveal a high incidence of voiding dysfunction in older male renal transplant recipients. High residual urine and urinary retention after renal transplantation may induce recurrent urinary tract infections, cause relevant complications, and seriously affect graft function. Recognizing the substantial effects of postoperative voiding dysfunction will enable optimum management of older kidney transplant recipients.

[Quality of kidney procurement in Germany. Ten years experience and 486 renal allografts in a single centre]. / Qualität der Leichennierenentnahme in Deutschland. 10-Jahres-Erfahrungen mit 486 transplantierten Leichennieren an einem Zentrum.

BACKGROUND: Organ damage during organ procurement is believed to be an increasing problem among transplant centres. However, only very few published data are available. The purpose of our study was to examine the quality of kidney procurement in Germany. METHODS: We retrospectively analyzed all allograft renal transplants performed at our centre from 1996 to 2005. All kidneys were retrieved in Germany and allocated by Eurotransplant. RESULTS: From a total of 486 cadaveric kidneys, 103 (21.2%) were not correctly retrieved. Nevertheless, none of the organs had to be rejected. In 18 (3.7%), a technically insufficient organ retrieval was associated with a considerable extension of the surgical procedure or complications. CONCLUSIONS: Technically insufficient kidney procurement rarely results in clinical consequences. However, surgeons performing organ retrieval should be better trained. Whether adequate technical proficiency is achieved with ten supervised cases, as requested by the German Medical Association, remains to be determined. In our opinion, a further interdisciplinary course that trains surgeons in more refined techniques of organ procurement is desirable.

Evaluation of type 1 diabetics for simultaneous pancreas-kidney transplantation with regard to cardiovascular risk.

The main cause of death for diabetic patients and patients on dialysis is coronary artery disease (CAD). The most common cause of graft loss following simultaneous pancreas and kidney transplantation (SPK) is death with a functioning graft due to CAD. Therefore, careful pretransplantation evaluation of CAD is mandatory. In our series, every patient undergoes a noninvasive cardiac function test like dobutamine stress echocardiography (DSE) or myocardial thallium scintigraphy using adenosine to induce medical stress. Thirty patients were evaluated for SPK: 15 patients with myocardial scintigraphy and 8 with DSE. Seven investigations showed pathological findings and we performed coronary angiograms, none of which showed coronary artery stenosis. Seven primary coronary angiograms were performed: four due to a history of CAD and three as a primary diagnostic. Following SPK one patient died at 21 days after transplantation due to myocardial infarction. He had a history of CAD with angioplasty and stent implantation. Noninvasive cardiac function tests like DSE or myocardial scintigraphy are reliable methods to evaluate CAD in patients with diabetic nephropathy awaiting SPK. In case of a suspicious finding or a history of CAD, a coronary angiogram should be performed to assess the need for revascularization. Following this algorithm we may further reduce the mortality of SPK.

Residual kidney function after donor nephrectomy. Assessment by 99mTc-MAG3-clearance.

AIM: We evaluated the long-term residual renal function after donor nephrectomy using 99mTc-mercaptoacetyltriglycin (MAG3)-clearance. DONORS, METHODS: Altogether 49 kidney donors were examined using 99mTc-MAG3-clearance after nephrectomy for donation to a relative (m:f = 11:38; age 55+/-27 years). The donors were examined 16+/-8 years postoperatively (1.5-26 years). 42 donors (86%) showed normal creatinine values, whereas the other seven (14%) exhibited slightly elevated levels. 20 donors were examined pre- and postoperatively and compared intraindividually. The kidney function was compared to the age adapted normal values of healthy persons with two kidneys (67-133% of age related mean). RESULTS: After nephrectomy all donors showed a normal perfusion, good secretion, merely physiological intrarenal transit and a normal elimination from the kidneys. The 99mTc-MAG3-clearance was 69+/-15% of the normal mean value of healthy carriers of two kidneys regardless of the gender. 20 donors with a preoperative examination showed a significantly reduced total renal function from 84+/-15% of the mean normal value preoperatively to 60+/-15% postoperatively (p <0.0005). 15 donors of this group exhibited a significant functional increase of the residual kidney from 40% initially to 60% after nephrectomy (p = 0.003). No correlation was found between the initial-99mTc-MAG3-clearance measured prior to nephrectomy and the clearance levels after nephrectomy. Also, no correlation between the preoperative 99mTc-MAG3-clearance and the postoperative serum creatinine values could be observed. Altogether, 22% of the donors (11/49) developed arterial hypertension 10+/-8 years after donation (1-23 years). This corresponds to the normal age prevalence of hypertension in the carriers of two kidneys. Three donors suffered from arterial hypertension prior to the operation. CONCLUSION: Kidney donors with normal or slightly elevated creatinine values postoperatively show a 99mTc-MAG3-clearance value of 69% of the mean value of healthy carriers of two kidneys. This may serve as a reference value for healthy carriers of one kidney. In our study we demonstrated a good compensation of the contralateral kidney via renal scintigraphy by means of 99mTc-MAG3-clearance.

Multiple adenomas and hepatocellular carcinoma in a renal transplant patient with glycogen storage disease type 1a (von Gierke disease).

We report on a 42-year-old female patient with glycogen storage disease type 1a (von Gierke disease, GSD 1a) who developed hepatic adenomas and finally a hepatocellular carcinoma 10 years after renal transplantation. The tumor was resected; however, the patient died 6 months later as a result of fulminant carcinoma recurrence. In patients who have GSD 1a with terminal renal failure, combined liver and kidney transplantation may be considered at an early stage of the disease.

Pathophysiology of cyclosporine-induced nephrotoxicity in humans: a role for nitric oxide?

BACKGROUND: The causes for the nephrotoxicity of cyclosporine A (CsA) have not been fully elucidated. Intrarenal vasoconstriction induced by several different mediators, both in humans and experimental animals, have been proposed. METHODS: We studied prostaglandin metabolites, endothelin and nitric oxide in kidney transplant patients receiving their first CsA dose. Prostaglandin metabolites in the urine and endothelin and nitric oxide (NO2/NO3 in urine and plasma were measured in 14 patients before and 3 and 6 h after oral ingestion of CsA (10 mg/kg b.w.). Clearances for inulin and p-aminohippuric acid (PAH) were measured before and in two separate 3-hour periods after CsA. Blood pressure, heart rate, and CsA blood levels were also determined. RESULTS: Clearances of inulin and PAH decreased progressively after CsA dosage while renal vascular resistance increased. Nitric oxide plasma levels decreased in nearly all patients from 21.0 +/- 2.8 to 19.1 +/- 2.6 (p = 0.003) and then rose slightly to 19.5 +/- 2.5 micromol/l (p = 0.1) 3 and 6 h after CsA ingestion, respectively. Urinary excretion of NO2/NO3 decreased nonsignificantly from 269 +/- 38.8 to 259 +/- 27.7 and 254 +/- 41.6 micromol/min (p = 0.5 and 0.5). At the same time, urinary prostaglandin E2 and 6-keto-prostaglandin F(1 alpha) excretion rate declined significantly [from 1,187 +/- 254 to 1,186 +/- 351 and 730 +/- 148 pg/min (p = 0.27 and 0.02) and from 697 +/- 115 to 645 +/- 134 and 508 +/- 58.2 pg/min (p = 0.34 and 0.05)]. Urinary thromboxane B2 and plasma and urinary endothelin first increased and then decreased nonsignificantly. Mean arterial pressure rose from 107 +/- 2.5 to 110 +/- 2.6 and 114 +/- 3.4 mm Hg (p = 0.1 and 0.05). CONCLUSION: The pathophysiology of CsA-induced acute renal vasoconstriction involves several different mechanisms including a decrease of the vasodilating prostaglandins E2 and 6-keto-prostaglandin F(1 alpha) and possibly nitric oxide.

Angiotensin II infusion increases plasma erythropoietin levels via an angiotensin II type 1 receptor-dependent pathway.

BACKGROUND: Angiotensin-converting enzyme inhibitors (ACEIs) have been shown to lower hematocrit and erythropoietin (EPO), but a direct link between angiotensin II (Ang II) and EPO in humans has not been shown. METHODS: Placebo or Ang II was infused for six hours in nine healthy male volunteers with and without blockade of the Ang II subtype 1 receptor (AT1R). EPO concentrations were measured 3, 6, 12, and 24 hours after the start of the infusion. RESULTS: Ang II raised the mean arterial pressure by about 20 mm Hg. Consistent with the known diurnal variation, EPO levels rose significantly (P < or = 0.02) during the day in all groups. During Ang II infusion, EPO levels rose to significantly higher levels after 6 and 12 hours compared with placebo [9.9 +/- 3.5 vs. 7.2 +/- 3.1 mU/mL (3 h, P = NS); 16.9 +/- 4.5 vs. 8.8 +/- 3.7 mU/mL (6 h, P = 0.01); 17.0 +/- 8.6 vs. 11.1 +/- 4.7 mU/mL (12 h, P = 0.01)] and returned to baseline after 24 hours (7.9 +/- 3.8 vs. 10.6 +/- 8.6 mU/mL, P = NS). With AT1R blockade, blood pressure remained normal during Ang II infusion, and EPO levels were never significantly different from placebo [6.8 +/- 4.8, 10.5 +/- 5.6, 13.1 +/- 9.0, and 12.4 +/- 10.1 mU/mL at 3, 6, 12, and 24 h after infusion, respectively, P = NS]. CONCLUSIONS: Ang II increases EPO levels in humans. This increase requires the participation of AT1R.

Effect of angiotensin II infusion with and without angiotensin II type 1 receptor blockade on nitric oxide metabolism and endothelin in human beings: a placebo-controlled study in healthy volunteers.

BACKGROUND: Angiotensin II has been shown to induce the synthesis of endothelium-derived relaxing factor nitric oxide (NO) and endothelin in vitro. In human beings, to our knowledge, no data on NO release in response to angiotensin II and on the influence of angiotensin II type 1 receptor blockade have been published. METHODS: In a placebo-controlled study in nine healthy volunteers, angiotensin II was administered intravenously for 6 hours with and without pretreatment with valsartan, a specific angiotensin II type 1 receptor antagonist. NO (NO2 + NO3) and endothelin plasma concentrations, clearance values for inulin and paraaminohippuric acid and NO (NO2 + NO3) excretion in urine were determined. RESULTS: During angiotensin II infusion NO plasma concentrations remained unaltered compared with placebo after 3 hours: 6.66 +/- 5.49 versus 5.56 +/- 3.09 micromol/L (P = ns) but increased after 6 hours: 18.36 +/- 20.02 versus 7.13 +/- 3.87 micromol/L (P < .04). The same was noted after pretreatment with valsartan: 7.61 +/- 5.69 versus 5.56 +/- 3.09 micromol/L (P= ns) after 3 hours, and 21.70 +/- 11.51 versus 7.13 +/- 3.87 micromol/L (P = .02) after 6 hours. In urine fractional NO excretion decreased after angiotensin II infusion: 0.87 +/- 0.72 versus 0.95 +/- 0.71 (P = .5) during the first 3 hours, and 0.44 +/- 0.39 versus 0.78 +/- 0.43 (P = .01) during the following 3 hours. After valsartan pretreatment the decrease in fractional urinary NO excretion began earlier: 0.40 +/- 0.15 versus 0.95 +/- 0.71 (P = .04) during the first 3 hours, and 0.17 +/- 0.11 versus 0.78 +/- 0.43 (P = .01) during the following 3 hours. Endothelin plasma concentrations showed no difference after angiotensin II infusion with or without valsartan. CONCLUSIONS: Our observations demonstrate for the first time that angiotensin II increases NO plasma concentrations in human beings and that this response is not mediated by angiotensin II type 1 receptor. In spite of increased NO plasma levels, urinary NO excretion decreased. Endothelin plasma levels remained unchanged during angiotensin II infusion.

Exhaustive exercise-induced tissue hypoxia does not change endothelin and big endothelin plasma levels in normal volunteers.

Chronic hypoxia has been shown to increase plasma endothelin levels. The current study was undertaken to examine the effect of exercise-induced tissue hypoxia on plasma levels of endothelin-1 (ET-1) and its precursor big endothelin-1 (Big-ET-1). After approval by the local ethical committee an incremental dynamic exercise test was performed in 12 physically trained volunteers (aged 20 to 40 years), using an electrically braked bicycle ergometer. The protocol included a step-wise increase of the workload until a heart rate of 130/min was reached, followed by a maintenance period of 25 min, and a further step-wise increase until exhaustion. Blood was drawn before, at several time points during, and 5 min after termination of the study for determination of ET-1, Big-ET-1, plasma renin activity (PRA), angiotensin converting enzyme (ACE), norepinephrine, epinephrine, and lactate. Lactate levels at baseline were 14.5 +/- 1.6 mg/dL (mean +/- SEM), which increased to 76.5 +/- 4.8 mg/dL at the time of exhaustion (P < .01). Baseline values for ET-1 and Big-ET-1 were 0.264 +/- 0.061 and 0.637 +/- 0.130 fmol/ml, respectively, which remained essentially unaltered throughout the exercise test. PRA was 1.46 +/- 0.45 ng/mL/h before exercise and increased to 3.55 +/- 0.96 ng/mL/h at exhaustion (P < .001). Norepinephrine and epinephrine were also increased at exhaustion. The study demonstrates that exhaustive physical exercise with acute development of pronounced tissue hypoxia--in contrast to chronic hypoxia--does not influence the release of ET-1 or Big-ET-1 or the conversion of the precursor to the active compound. Unlike endothelin, circulating renin and the catecholamines were markedly stimulated by this maneuver.

Mechanism of angiotensin converting enzyme inhibitor-related anemia in renal transplant recipients.

To delineate the pathogenesis of the reduction in hemoglobin occurring in renal transplant patients treated with angiotensin converting enzyme inhibitors (ACEI) and azathioprine (AZA) a controlled, prospective trial of ACEI withdrawal was conducted. The ACEI was replaced by nifedipine or clonidine in 15 kidney transplant patients immunosuppressed with AZA and prednisone (enalapril in 14 and captopril in 1). Before and during 10 to 12 weeks after withdrawal of the ACEI, AZA metabolites, renal function parameters and hematological parameters including erythropoietin and reticulocytes were evaluated. Enalaprilat levels were measured and compared with 15 similar patients matched for transplant function and enalapril dosage immunosuppressed with cyclosporine and prednisone. AZA metabolites did not differ significantly in the presence or absence of the ACEI. Enalaprilat levels also showed no significant difference between the two patient groups treated with AZA or cyclosporine. Hematocrit and hemoglobin increased significantly from 37.5 +/- 6.4 to 39.7 +/- 3.6% (mean +/- SD, P = 0.02) and 12.8 +/- 2.2 to 13.5 +/- 1.2 g/dl, P = 0.04, respectively, 10 to 12 weeks after ACEI treatment had been discontinued. Simultaneously numbers of reticulocytes and erythropoietin concentrations rose significantly after 2, 4 and 10 weeks, with a peak at two weeks (from 14.1 +/- 3.8 to 20.6 +/- 8.0/1000, P < 0.05 and from 14.3 +/- 12.4 to 29.3 +/- 54.5 mU/ml, P < 0.05, respectively). In conclusion, ACEI-related anemia in renal transplant recipients seems to be due to the erythropoietin-lowering effect of this group of drugs. A pharmacokinetic interaction between AZA and enalapril is not likely since plasma enalaprilat levels were independent of the immunosuppressive regimen and AZA metabolite levels were unchanged in the presence and absence of the ACEI. Several mechanisms by which angiotensin converting enzyme blockade may cause a decrease in circulating erythropoietin are discussed.

HPLC analysis of azathioprine metabolites in red blood cells, plasma and urine in renal transplant recipients.

Anemia has been frequently reported in renal transplant recipients receiving azathioprine for immunosuppression and enalapril for treatment of hypertension. During the course of a prospective trial in such patients we determined azathioprine metabolites in erythrocytes, plasma, and urine as well as erythropoietin and hemoglobin levels in order to evaluate a potential interaction between these 2 drugs, possibly leading to anemia. Two specific high performance liquid chromatography (HPLC) methods for determination of azathioprine metabolites, both employing a mercurial cellulose resin for extraction, are presented. One method using a strong anion exchange column allows detection of 6-thioguanosine di- and triphosphate (thioguanine nucleotides) in red blood cells (RBC) with a sensitivity of 30 pmol/100 microliters RBC. 6-mercaptopurine (MP) and 6-thiouric acid (TUA) in plasma and urine were analyzed simultaneously by reversed-phase HPLC with a sensitivity of 5 ng/ml. The average (median values are given) steady state concentrations of thioguanine nucleotides in erythrocytes came to 267 pmol/100 microliters RBC (range 53-613) with and to 246 pmol/100 microliters RBC (range 39-629) without concomitant enalapril medication. Mean plasma concentrations of MP and TUA 3 hours after drug intake came to 14.8 +/- 9.9 ng/ml and 398 +/- 262 ng/ml, respectively, during enalapril comedication. Withdrawal of enalapril did not influence these metabolite levels coming to 15.3 +/- 9.1 and 451 +/- 253 after stopping enalapril treatment. Thioguanine nucleotides in RBCs were neither related to the dose of azathioprine given (r = -0.113, p > 0.05) nor to hemoglobin levels (r = 0.278, p > 0.05). However, azathioprine dose/kg body weight seemed to be related to hemoglobin concentration, with and without enalapril comedication. We conclude that enalapril therapy does not influence the measured azathioprine metabolites, the reported cases of anemia may rather be due to a pharmacodynamic interaction as shown by the significant increase in erythropoietin after withdrawal of enalapril. The assays described here are suitable to study the metabolism of azathioprine in patients with various diseases.

Murine hepatitis caused by lymphocytic choriomeningitis virus. II. Cells involved in pathogenesis.

BACKGROUND: Human virus hepatitides are often assumed to result from pathogenic immune responses rather than from direct viral cytopathic effects, but the details are largely unknown. Hepatitis of the mouse undergoing infection with lymphocytic choriomeningitis (LCM) virus is an immunopathologic phenomenon, and its analysis may help us to understand some of the events leading to the human illnesses. EXPERIMENTAL DESIGN: Mice were infected with LCM virus and were depleted of T cells or their subsets by inoculation of monoclonal antibodies; other infected mice lacked all T lymphocytes or the CD8+ subset because of genetic defects. Also, mice were infected and transfused with unsorted or CD(4+)-enriched LCM-immune spleen cells. Subsequently, the infectious titers were determined, the cytolytic activities of mononuclear cells isolated from the livers were measured, and the disease process was studied. RESULTS: In LCM virus-infected mice devoid of all T lymphocytes, pathologic alterations remained undetectable. In contrast, immunocompetent animals responded with a severe hepatitis, at the height of which the liver contained large numbers of cytolytic mononuclear leukocytes. Experiments with mice depleted of subset T lymphocytes revealed a predominantly CD8+ T lymphocyte-mediated phase, which was characterized by panlobular inflammation, whereas later there was a periportal inflammatory reaction, in which mainly CD4+ T lymphocytes were involved. Infusion of syngeneic immune spleen cells from immunocompetent donor mice into infected thymus-less mice resulted in virus elimination and damage to liver cells. With a similar protocol and the use of congenic mice, CD8+ T lymphocytes were observed to rapidly enter the recipients' livers, where they were present at the time liver cell injury was apparent. In mice genetically deficient in CD8+ T lymphocytes due to disruption of the gene for beta 2-microglobulin, a somewhat different type of LCM hepatitis developed that was largely dependent on CD4+ T lymphocytes. Liver cells were also damaged in infected nude mice that had been infused with positively selected CD4+ spleen cells from infected +/+ mice. CONCLUSIONS: Our findings published previously (Löhler J, Gossmann J, Kratzberg T, Lehmann-Grube F. Lab Invest 1994; 70:263-78) and related here suggest that the hepatitis in mice undergoing infection with LCM virus consists of three consecutive phases, which are mediated predominantly by NK cells, CD8+ T lymphocytes, and CD4+ T lymphocytes, respectively. Presumably, other elements of the immune system, such as mononuclear phagocytes and B lymphocytes, contribute to the pathogenesis.

Murine hepatitis caused by lymphocytic choriomeningitis virus. I. The hepatic lesions.

BACKGROUND: The hepatitis that occurs after adult mice are infected with lymphocytic choriomeningitis (LCM) virus is immune mediated, although the details of the pathogenetic mechanisms are largely unknown. To better understand the sequence of events leading to alterations typical for hepatitides with immunopathogenesis, livers of immunocompetent mice infected with LCM virus were examined. EXPERIMENTAL DESIGN: Virus replication and histopathology in the livers and concentrations of liver enzymes in the sera of C57BL/6 mice were followed from day 3 through day 14 after intraperitoneal infection with 10(6) mouse infectious units of LCM virus. Histologic, histochemical, immunohistologic, and in situ hybridization methods were used to determine the cells involved in the inflammatory process. RESULTS: Infectious virus rose to 10(9) mouse infectious units/g of liver by day 7 and declined thereafter. Viral RNA and antigen were localized in foci of hepatocytes and in Kupffer and endothelial cells of the sinusoids. Disseminated spotty necroses, steatosis, a marked sinusoidal reaction, and lobular and (later) periportal mononuclear infiltrates were observed. In the infiltrates, T cells predominated followed by macrophages and NK cells; the number of B and plasma cells rose moderately. Among T lymphocytes the CD8+ cells increased preferentially, and the CD4/CD8 ratio changed from 1.7 to 0.3. Other features were major histocompatibility complex antigen expression on hepatocytes, enhanced immunocytochemical evidence of fibronectin and ICAM-1 in sinusoids, and deposition of immunoglobulin, complement, and fibrinoid. Changed activities of liver enzymes and bilirubin levels paralleled the pathologic alterations. CONCLUSIONS: Although CD8+ T cells seem to be central in the pathogenesis of LCM hepatitis, probably more than one immunopathologic mechanism is operative.

Anemia in renal transplant recipients caused by concomitant therapy with azathioprine and angiotensin-converting enzyme inhibitors.

Immunosuppression of recipients of renal transplants with azathioprine has been associated with two major side effects: hepatotoxicity and myelotoxicity, mainly in the form of leukopenia. Reports of isolated anemia in these patients have been rare. We now observed the development of severe anemia in 9 out of 11 renal transplant recipients whose immunosuppressive regimen was converted from cyclosporine plus prednisone to azathioprine plus prednisone. A significant (P = 0.001) drop in hematocrit (from 34 +/- 4% to 27 +/- 3%, mean +/- SD) and hemoglobin (from 11.6 +/- 1.3 g/dl to 9.5 +/- 1.0 g/dl) was found. Since a common variable of all these patients was their use of an angiotensin-converting enzyme (ACE) inhibitor as antihypertensive medication, we speculated that the combination of azathioprine and ACE blocker might be the reason for the anemia. We then compared 2 groups of 10 patients each who had been on azathioprine as their regular immunosuppressive agent and who did or did not take an ACE inhibitor. Hematocrit and hemoglobin were significantly (P = 0.01) lower in the group of patients taking ACE inhibitors (33 +/- 6% versus 41 +/- 5% and 11.5 +/- 2.0 g/dl versus 14.0 +/- 1.6 g/dl, respectively). Haptoglobin levels were also significantly (P = 0.05) lower in the ACE inhibitor group (116 +/- 65 mg/dl versus 210 +/- 114 mg/dl). Erythropoietin concentration in the serum and the reticulocyte index were slightly, but not significantly, higher in the ACE inhibitor group but the values were probably too low for their degree of anemia. Comparing hematological parameters of the patients in the ACE inhibitor group before and after beginning of the antihypertensive treatment confirmed a significant reduction of hematocrit and hemoglobin following therapy with an ACE inhibitor. Hematocrit fell from 41 +/- 7% to 36 +/- 6% and hemoglobin from 14.0 +/- 2.3 g/dl to 11.3 +/- 1.5 g/dl (P < 0.05 for both). We conclude that the combination of these two drugs should probably be avoided.