Correspondence *University of Vermont/FAHC, Department of Medicine, Division of Nephrology, 208 South Park Drive, Suite 2, Colchester, VT 05446, USA

Email marios.prikis@vtmednet.org

The case

A 73-year-old man with a history of chronic low back pain was admitted to hospital with worsening back pain, as well as nausea and vomiting. His past medical history included atrial fibrillation, hypertension, hyperlipidemia, and coronary artery disease. Results of a physical examination, as documented by caregivers at the hospital, were consistent with volume depletion. The patient's initial blood pressure measurements were in the range 90–100/50–60 mmHg. His serum creatinine level had increased to 106.1 mol/l (1.2 mg/dl), from a baseline level of 79.6 mol/l (0.9 mg/dl) 6 months earlier. At baseline, his estimated glomerular filtration rate (GFR) had been 88 ml/min/1.73 m². Following volume repletion with 1 l of intravenous 0.9% sodium chloride solution the patient's blood pressure improved, and CT of the abdomen and pelvis was performed with oral and intravenous contrast (iopamidol 76%). A left psoas major muscle abscess was discovered and the patient was transferred to another hospital for further evaluation and treatment. The time from his initial admission to transfer to this hospital was approximately 24 h.

On admission to the second hospital the patient was judged to be euvolemic and his vital signs were as follows: blood pressure 123/72 mmHg, pulse rate 90 beats/min, respiratory rate 18 breaths/min, and temperature 37.6 °C. His serum creatinine level was 92.8 mol/l (1.05 mg/dl) at this time. CT-guided drainage of the abscess was performed immediately, and purulent fluid was sent to the laboratory for culture. An additional liter of intravenous 0.9% sodium chloride solution was administered to the patient, and intravenous ceftriaxone 2,000 mg/day and vancomycin 1,500 mg/day were started empirically pending culture results. The following day, the blood cultures and the aspirate were found to be positive for Staphylococcus aureus with susceptibility to vancomycin, nafcillin, cephalosporins, and fluoroquinolones. The urine culture was negative. After consultation with the infectious diseases team, ceftriaxone and vancomycin were replaced with intravenous nafcillin 2,000 mg every 4 h and intravenous levofloxacin 750 mg/day. On this day (hospital day two) the patient's serum creatinine level was 95.5 mol/l (1.08 mg/dl). The next day, his serum creatinine level was not checked.

By hospital day four, the patient's serum creatinine level had risen to 107.0 mol/l (1.21 mg/dl), and it increased to 197.1 mol/l (2.23 mg/dl) and 233.4 mol/l (2.64 mg/dl) on hospital days five and six, respectively. On hospital day six the nephrology consultation team was called in to evaluate the patient. Medications at the time of this consultation included nafcillin, levofloxacin, gabapentin, heparin, metoprolol, pantoprazole, and zinc sulphate. The patient had previously been taking paracetamol (acetaminophen), amlodipine, benazepril, fenofibrate, oxycodone, and warfarin, but these medications had all been discontinued on admission to the second hospital. The patient denied use of over-the-counter medications including NSAIDs. There was no documented systolic blood pressure below 100 mmHg before the nephrology team consultation.

Physical examination on hospital day six revealed a man who seemed acutely ill. His vital signs were as follows: blood pressure 100/70 mmHg, pulse rate 83 beats/min, respiratory rate 16 breaths/min, and temperature 37.8 °C. Examination of the patient's skin revealed a diffuse erythematous maculopapular rash with multiple excoriations. The rash was most prominent on the trunk but also involved the face and the extremities. Cardiovascular, respiratory, neurological, and abdominal examinations were all normal. There was no edema, and the patient's urine output was 50–60 ml/h.

Blood tests revealed a white blood cell count of 7.4 10⁹/l with 3% eosinophils, hemoglobin 98 g/l (9.8 g/dl), platelets 381 10⁹/l, sodium 138 mmol/l, potassium 4.2 mmol/l, chloride 106 mmol/l, total carbon dioxide concentration 19 mmol/l, blood urea nitrogen 10 mmol/l (28 mg/dl), and serum creatinine 233.4 mol/l (2.64 mg/dl). A renal ultrasound showed that the kidneys were normal in size, and there was no evidence of obstruction. Urine dipstick analysis was positive for protein, blood, and leukocyte esterase, and urine specific gravity was 1.010.

Examination of the urine sediment (12 ml of urine spun at 400 g for 5 min) showed numerous white blood cells and white blood cell casts (Figure 1A). A diagnosis of drug-associated acute interstitial nephritis (AIN) was strongly considered, prompting a careful review of the patient's medications. The agents most likely to have caused AIN in this patient were nafcillin, levofloxacin, and pantoprazole (see below), and it was, therefore, recommended that these medications be discontinued. On the basis of input from all consultants involved in the case, however, the primary team managing the patient decided to discontinue only the nafcillin. All other medications—including levofloxacin—were continued, and intravenous vancomycin was restarted. The patient's serum creatinine level peaked at 247.5 mol/l (2.80 mg/dl) the following day, and improved rapidly over the next few days (Figure 2A).

Figure 1 Photomicrographs of urine sediment showing white blood cell casts

(A) White blood cell casts from the main case. (B) White blood cell casts from case #2.

Full figure and legend (27K)Figures & Tables indexDownload Power Point slide (137K)

Figure 2 Plots of daily serum creatinine values

(A) Daily serum creatinine values from the main case showing a decrease in serum creatinine level after discontinuation of nafcillin. The dotted line represents the baseline creatinine level of 79.6 mol/l (0.9 mg/dl). (B) Daily serum creatinine values from case #4 showing that both initial exposure and re-exposure to nafcillin resulted in increases in serum creatinine level. The dotted line represents the patient's baseline creatinine level. ^aTo convert creatinine from mg/dl to mol/l, multiply by 88.4.

Full figure and legend (24K)Figures & Tables indexDownload Power Point slide (134K)

Discussion of diagnosis

AIN is a well-known complication of therapy with a number of medications, including proton pump inhibitors,¹ NSAIDs, and antibiotics such as vancomycin,^{2, 3, 4} cephalosporins, sulphonamides,⁵ and several penicillin derivatives.^{6, 7, 8, 9, 10} Meticillin (formerly methicillin) is the penicillin derivative most commonly reported as associated with nephritis, with up to one-third of exposed patients developing the complication.¹⁰ By contrast, nafcillin, a semisynthetic penicillin extensively used as the drug of choice for treatment of a number of meticillin-sensitive S. aureus (MSSA) infections such as endocarditis, catheter-related infections, and complicated soft tissue infections,^{11, 12} was reported to be an unlikely cause of AIN in 1980.¹³ A prospective analysis of 210 consecutive patients given nafcillin at the University of California, San Francisco, indicated the complete absence of clinically significant nephrotoxicity.¹³ Furthermore, few suspected cases of nafcillin-associated interstitial nephritis have been reported over the past 25 years.^{6, 9} The recent clinical experience at the institution where the present case was treated, however, suggests that this 'rare' complication of nafcillin might be underappreciated.

Over the course of 1 year, a total of four patients with MSSA infections presented to the same institution and developed acute kidney injury (AKI) most likely secondary to nafcillin-associated AIN (see Table 1). Discontinuation of nafcillin (while other medications were continued) resulted in improvement in kidney function in all cases. In one case (case #4 in Table 1), re-exposure to nafcillin led to a relapse of renal inflammation (Figure 2B). The cases reported here provide compelling data to support the diagnosis of nafcillin-associated AIN as the primary cause of each patient's AKI and they highlight the important diagnostic clues that can be used to establish this diagnosis.

Table 1 Clinical summary of four cases of nafcillin-associated acute interstitial nephritis

Full tableFigures & Tables indexDownload Power Point slide (176K)

Sterile pyuria with white blood cell casts and kidney dysfunction were universal findings in all cases. With these abnormalities, the diagnosis of drug-induced AIN seemed probable. While it is well known that multiple classes of drugs can precipitate AIN,^{1, 2, 3, 4, 5, 6, 7, 8, 9, 10} careful review of the medication administration records demonstrated a clear relationship between nafcillin initiation and the development of AKI in each case. This temporal association was not as apparent for other medications known to cause AIN, such as vancomycin, ceftriaxone and pantoprazole. In each case, kidney function improved with cessation of nafcillin (Figure 2). A strong preference of the infectious disease consultant for use of nafcillin to treat the S. aureus infection in case #4 led to this antibiotic being restarted in this patient. Re-exposure to nafcillin promptly led to a relapse of kidney inflammation (Figure 2B). Additionally, both hematuria and proteinuria were seen in all four patients. Previous case reports of nafcillin-associated AIN have also described hematuria and proteinuria,^{6, 9} and these findings are also common in those receiving meticillin.⁸ Finally, rash (which occurred in the main case) and peripheral eosinophilia (which occurred in case #3) have both been reported with nafcillin^{6, 9} and meticillin.⁸ While a kidney biopsy remains the gold standard for a pathological diagnosis of AIN, the diagnosis can also be based on strong clinical evidence. In each case presented here, pyuria with frequent white blood cell casts was the dominant urinary finding as the GFR declined. No red blood cell casts or 'muddy brown' casts were present in any of the cases. These observations, combined with the timing of nafcillin exposure, made the clinical diagnosis of nafcillin-induced AIN the most compelling. Kidney function promptly improved in each patient with cessation of nafcillin exposure, obviating the need for renal biopsy. Gallium scans were not pursued in the diagnostic work-ups as their utility in the diagnosis of drug-associated AIN has not been clearly established.¹⁴

Differential diagnosis

The four patients reported here all experienced GFR decreases, and urinary findings that were most consistent with a diagnosis of AIN (hematuria, proteinuria, and pyuria with frequent white blood cell casts), signs that reversed when nafcillin was stopped, and, in case #4, recurred with re-exposure to nafcillin. Each patient in the series received other medications that also have the potential to cause AIN, so those drugs must also be considered in the differential diagnosis. For example, the main case was receiving a proton pump inhibitor (pantoprazole), a class of drugs recognized as a common cause of drug-associated AIN.¹ The timing of kidney injury in each case, however, fitted best with nafcillin exposure, and improvement in kidney function was observed with prompt cessation of nafcillin alone. Kidney function improved in the face of continued exposure to the other medications that should be considered in the differential diagnosis for drug-associated AIN.

Patients who receive nafcillin are often at high risk for other forms of AKI, and an element of prerenal azotemia as a contributor to the decreased GFR cannot be excluded; however, clinical volume depletion was not evident at the time of kidney injury in any of the cases. Ischemic and/or nephrotoxic acute tubular necrosis must also be considered. The 'muddy brown' casts classically seen in patients with acute tubular necrosis, however, were not observed on urine microscopy in any case. Patients with infective endocarditis, catheter-related infections, or soft tissue infections are also likely to receive additional nephrotoxic agents such as radiocontrast agents. While the main patient described here did receive a nonionic, low osmolar (iopamidol 76%) radiocontrast agent, his GFR did not decline until 4 days after its administration, making contrast-induced nephropathy unlikely. Both MSSA and meticillin-resistant S. aureus (MRSA) infections have been associated with glomerulonephritis.^{15, 16} While hematuria and proteinuria occur in glomerulonephritis, these were not the dominant urinary findings in the four cases presented here. Furthermore, one would not expect glomerulonephritis to resolve with the rapidity observed in these cases. Kidney injury secondary to atheroembolic disease can also present with hematuria, proteinuria, pyuria, and peripheral eosinophilia, but the clinical course observed in all four cases was not consistent with this chronic process. Urinary tract obstruction was also considered in the differential diagnosis. In each case, this diagnosis was excluded by a renal ultrasound. Finally, given the acute nature of the presentations, tubulointerstitial diseases such as sarcoidosis were thought to be unlikely.

Discussion of treatment

Prompt removal of the suspected offending medication is the primary therapy during any adverse reaction or injury that might be associated with a drug. Intense tubulointerstitial inflammation is the histopathologic lesion in AIN and can be easily appreciated by examination of the urine sediment (Figure 1). Corticosteroids have been suggested as part of the treatment strategy for this inflammation because of its potentially destructive nature. No prospective, randomized, controlled trials have examined the efficacy of this therapy, although a large retrospective study published in 2004 found no benefit of corticosteroid treatment in terms of outcome (determined by serum creatinine level) at 1, 6 and 12 months.¹⁷ The potential for incomplete recovery from medication-related AIN does exist and patients should be followed up by a nephrologist if this occurs.

Conclusions

Within 1 year, a single institution saw four patients with MSSA infections who developed AKI that was most likely secondary to nafcillin-associated AIN. Records from the institution show that nafcillin had been ordered 180 times over that year. If it is assumed that each order was for a different patient, the minimum incidence of nafcillin-associated AIN (with a reduction in GFR) was at least 2.2% for that year. This experience and the review of the literature suggest that nafcillin can cause AIN, and indicate that previous reports that downplay the frequency of clinically significant adverse effects of nafcillin may be overstated. In the absence of a tissue diagnosis, physicians might be unlikely to submit reports of suspected reactions based on a clinical diagnosis alone. Kidney dysfunction in patients receiving nafcillin might be incorrectly attributed to renal hypoperfusion or acute tubular necrosis. Finally, because it is possible for the estimated GFR to remain unchanged despite the presence of mild to moderate inflammation (e.g. in pyelonephritis), the actual incidence of renal inflammation caused by nafcillin can be estimated only by systematic examination of the urine before and after starting the drug. The experience in the institution reported here indicates that AIN associated with nafcillin therapy is not rare, and that careful examination of the urine should be performed in any patient receiving nafcillin who sustains even a slight decrease in GFR. The findings also indicate that this condition can be satisfactorily diagnosed and managed without an invasive kidney biopsy being performed.

References

1. Torpey N et al. (2004) Drug-induced tubulo-interstitial nephritis secondary to proton pump inhibitors: experience from a single UK renal unit. Nephrol Dial Transplant 19: 1441–1446 | PubMed | ISI | ChemPort |
2. Eisenberg ES et al. (1981) Vancomycin and interstitial nephritis. Ann Intern Med 95: 658 | PubMed | ISI | ChemPort |
3. Hsu SI (2001) Biopsy-proved acute tubulointerstitial nephritis and toxic epidermal necrolysis associated with vancomycin. Pharmacotherapy 21: 1233–1239 | PubMed | ISI | ChemPort |
4. Ratner SJ and Roberts DK (1988) Vancomycin-induced interstitial nephritis. Am J Med 84: 561–562 | Article | PubMed | ISI | ChemPort |
5. Kodner CM and Kudrimoti A (2003) Diagnosis and management of acute interstitial nephritis. Am Fam Physician 67: 2527–2534 | PubMed | ISI |
6. Bodendorfer TW (1980) Nafcillin-induced interstitial nephritis. JAMA 244: 2609 | Article | PubMed | ISI | ChemPort |
7. Cushner HM et al. (1985) Acute interstitial nephritis associated with mezlocillin, nafcillin, and gentamicin treatment for Pseudomonas infection. Arch Intern Med 145: 1204–1207 | Article | PubMed | ISI | ChemPort |
8. Ditlove J et al. (1977) Methicillin nephritis. Medicine (Baltimore) 56: 483–491 | PubMed | ChemPort |
9. Guharoy SR et al. (1993) Suspected nafcillin-induced interstitial nephritis. Ann Pharmacother 27: 170–173 | PubMed | ISI | ChemPort |
10. Nolan CM and Abernathy RS (1977) Nephropathy associated with methicillin therapy: prevalence and determinants in patients with staphylococcal bacteremia. Arch Intern Med 137: 997–1000 | Article | PubMed | ISI | ChemPort |
11. Horstkotte D et al. (2004) Guidelines on prevention, diagnosis and treatment of infective endocarditis executive summary; the task force on infective endocarditis of the European society of cardiology. Eur Heart J 25: 267–276 | Article | PubMed | ISI |
12. Ramsdale DR and Turner-Stokes L (2004) Prophylaxis and treatment of infective endocarditis in adults: a concise guide. Clin Med 4: 545–550 | PubMed | ISI |
13. Barriere SL and Conte JE Jr (1980) Absence of nafcillin-associated nephritis: a prospective analysis of 210 patients. West J Med 133: 472–477 | PubMed | ISI | ChemPort |
14. Graham GD et al. (1983) Failure of Gallium-67 scintigraphy to identify reliably noninfectious interstitial nephritis: concise communication. J Nucl Med 24: 568–570 | PubMed | ISI | ChemPort |
15. Kobayashi M and Koyama A (1999) Methycillin-resistant Staphylococcus aureus-related glomerulonephritis. Nephrol Dial Transplant 14 (Suppl 1): S17–S18 | Article |
16. Koyama A and Hirayama K (2001) Glomerulonephritis associated with Staphylococcus aureus infection. Intern Med 40: 365–367 | PubMed | ISI | ChemPort |
17. Clarkson MR et al. (2004) Acute interstitial nephritis: clinical features and response to corticosteroid therapy. Nephrol Dial Transplant 19: 2778–2783 | Article | PubMed | ISI | ChemPort |

Contact the journal about this article

Subject areas under which this article appears: Interstitial nephritis