| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Internal Medicine, Moses H. Cone Hospital, Greensboro, NC.
2 Department of Radiology, Wake Forest University, Winston-Salem, NC.
3 Department of Radiology, Duke University Medical Center, Box 3808, Durham, NC
27710.
Received July 19, 2007;
accepted after revision November 12, 2007.
Address correspondence to J. M. Provenzale.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Eight patients developed NSF during 2002–2006, of whom seven received the MR contrast agent gadodiamide (Omniscan), with doses of 0.10–0.31 mmol/kg. Data for the following were available for only 2005 and 2006: numbers of infected and uninfected renal failure patients who received MR contrast material and number of contrast-enhanced MR scans in all patients. We extrapolated data to 2002–2006 to approximate rates of NSF in infected and uninfected renal failure patients using Fisher's exact test for association between variables and calculated odds ratios with 95% CIs.
RESULTS. Five of seven NSF patients receiving MR contrast material had infections at the time of contrast administration. Three hundred thirty-four patients with renal failure received MR contrast material in 2005 and 2006 (29 infected). The rate of NSF was 6.7% in infected renal failure patients and 0.26% in uninfected patients. Data extrapolated to 2002–2006 yielded estimates of 75 infected and 750 uninfected renal failure patients. The association of NSF with infection was highly significant (p < 0.001) with an odds ratio of 25 and CI of 3.9–264.4.
CONCLUSION. The association of NSF with infection was highly significant, supporting the hypothesis that infection at the time of MR contrast administration in renal failure patients predisposes to NSF.
Keywords: connective tissue disorders • contrast media • dialysis • infections • MRI • nephrogenic systemic fibrosis • renal failure
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
NSF was initially described as "nephrogenic fibrosing dermopathy" due to the predominant skin findings but is now known to involve organs other than solely the skin, such as the liver, lungs, muscles, and heart [3]. Within the past few years, a strong association between impaired renal function and exposure to gadolinium-based MRI contrast media in NSF patients has been established [4, 5]. However, this association does not explain why only a few of the large number of patients with both risk factors develop NSF. This fact has led investigators to consider other, additional, risk factors that might place certain patients with both impaired renal function and exposure to gadolinium-containing contrast material at an even higher risk for developing NSF. For instance, some investigators have postulated that metabolic acidosis at the time of exposure to gadolinium-containing contrast agents may contribute to the development of NSF [4]. However, this association was not confirmed in a recent case series [5].
Investigators have suggested that major tissue injury may predispose patients with advanced renal dysfunction who receive MR contrast media to the development of NSF [2, 6]. In support of this hypothesis, other authors have shown increased C-reactive protein levels and erythrocyte sedimentation rates (ESRs) in patients with NSF [3]. Recently, proinflammatory conditions have been advanced as playing a role in the development of NSF [7]. One study, which defined a proinflammatory state as one in which an intense healing response is initiated with activation of major inflammatory pathways, identified three major types of proinflammatory states possibly predisposing to NSF: major surgery, major infection, and arterial or venous thrombosis [7]. These studies, in combination, suggest that pathologic tissue changes in various organs, referred to cumulatively as "inflammatory burden," may predispose patients with chronic renal failure to NSF after exposure to gadolinium-containing contrast agents. With this in mind, we evaluated patients with chronic kidney disease who developed NSF at our institution with respect to major surgery, major infection, and arterial or venous thrombosis at the time of administration of gadolinium-based contrast agents.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Seven patients underwent MR contrast material administration during a period ranging from 2 days to 5 months (median, 3 weeks) before the onset of NSF symptoms (Table 1). We subsequently queried the dermatologists and pathologists at our hospital to determine whether they were aware of other NSF patients who might not have been recorded on our list and found no additional patients. This study was compliant with HIPAA and was approved by our institutional review board, which provided a waiver of informed consent.
|
Electronic medical records were reviewed to determine patient age, sex, and presence of renal disease; time interval between contrast material administration and NSF symptom onset; type and dose of contrast agent; and indication of major surgery, major infection, or vascular thrombosis at the time of contrast material administration. Laboratory values examined included, but were not limited to, hemoglobin, serum bicarbonate, albumin, C-reactive protein level, and ESR. We were thus able to determine whether the eight NSF patients had renal failure and whether they underwent contrast-enhanced MRI. We were further able to determine whether they had a documented infection at the time of administration of MR contrast material.
We then set out to compare the rate of development of NSF in two groups of renal failure patients—that is, those with and those without a documented infection at the time of administration of contrast material. However, although we had data on the total number of NSF patients over the period of 2002–2006, the electronic medical records for a history of infection were available solely for the calendar years 2005 and 2006. Data before those years were kept using paper records and were not readily available to us.
Therefore, we first obtained the numbers of infected and uninfected renal failure patients for the years 2005 and 2006 and extra polated them to the 5-year period 2002–2006. We first obtained a list of MR examinations that were ordered by physicians in the nephrology group practice at our institution during 2005 and 2006. This nephrology group treats all patients with renal failure at our institution, and we are confident that no renal failure patients who underwent MRI were referred by other sources. We reviewed the medical records of the patients on this list to determine which examinations were performed on patients with advanced renal failure, which was defined as stage IV chronic kidney disease, acute renal failure, or hemodialysis dependency. This review identified 327 patients with advanced renal insufficiency who receiv ed a gadolinium-based MR contrast agent during 2005 and 2006, other than the seven NSF patients mentioned earlier. Next, we reviewed the medical records of these 327 patients for evidence of an active infection at the time of MR contrast agent exposure and found that 29 had a documented infection and 298 did not. Extrapolating these data to the full 2002–2006 period, we estimated that 750 renal failure patients without documented infection and 75 renal failure patients with a documented infection underwent contrast-enhanced MRI during the period 2002–2006.
As a secondary analysis, we set out to compare the rates of development of NSF in infected renal failure patients and infected patients without renal failure. However, we were again limited to actual totals for solely 2005 and 2006. For both years combined, the total was 22,948 contrast-enhanced MR examinations. The number of studies performed in 2005 and 2006 was similar for all sites except one, which was an outpatient imaging center acquired by our radiology group during 2005. Then, we used the average of the number of studies performed in 2005 and 2006 for each of the years 2002, 2003, and 2004 except for that site, which was assigned zero studies for those years. On this basis, we estimated that 46,633 contrast-enhanced MR examinations were performed from 2002–2006, the time period during which our NSF cases were identified.
It was not feasible to review the medical records of the estimated 22,948 patients with out renal failure who received MR contrast material, which made it necessary to estimate the infection rate in this population. As an approximation, we applied the infection rate in renal failure patients during 2005 and 2006 (29/327 or 9%) to the estimated general population of patients undergoing contrast-enhanced MRI during 2002–2006. Although we recognize that this assumption would tend to overestimate the infection rate in non–renal failure patients, it seems reasonable for our purposes. Using this information, we constructed a table to calculate odds ratios for developing NSF (Table 2).
|
The Fisher's exact test was used as a test of association between variables. The estimated event probability, together with 95% CIs, for developing NSF was determined for the following four groups: patients with renal failure but no infection, patients with renal failure and a documented infection, patients with an infection but not renal failure, and patients with neither renal failure nor an infection. CIs for the first two groups are Wilson CIs [8] and those for the latter two groups were derived from exact binomial CIs. CIs for probabilities were based on Wilson's test as recommended by Agresti and Coull [9]. However, when the estimated probability was zero, the exact Clopper-Pearson interval was used. CIs for the odds ratio were based on the exact enumeration of possible outcome tables [10]. We used version 9.1 of the SAS software system (SAS Institute) for all calculations.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
Hypoalbuminemia (albumin < 3.9 g/dL) was present in all patients, and anemia (hemoglobin < 12.0 g/dL) was present in seven patients. No other abnormal trends in laboratory data were noted; none of our patients had protein metabolic acidosis at the time of contrast-enhanced MRI. ESRs and C-reactive levels were elevated in the one patient for whom these data were available. Medications varied among the patients, and no single medication was common to all patients. The most common medications were erythropoietin and warfarin sodium (Coumadin, Bristol-Myers Squibb), both of which were used by four of the eight patients.
No patients had vascular thrombosis. One patient (patient 1) had major surgery (cholecystectomy 7 days after the second contrast-enhanced MR examination). The types of in fections in the five NSF patients with documented infection are as follows: Patient 1 was hospitalized with a catheter infection causing bacteremia and underwent contrast-enhanced MRI twice within 1 week during the period of bacteremia. She also underwent open cholecystectomy during that hospitalization, 15 days after the first contrast-enhanced MR examination, 11 days after bacteremia was documented, and 7 days after the second contrast-enhanced MR examination. The cholecystectomy operative report noted a leathery texture to the peritoneal tissue, which was considered consistent with NSF. Patient 2 was administered MR contrast material in the setting of a urinary tract infection, for which urine culture was positive for Enterococcus organisms, and was treated with antibiotics. Ten days later, he developed fever, rash, endocarditis, and bacteremia; blood cultures grew coagulase-negative Staphylococcus organisms. Transesophageal echocardiography 1 month later showed aortic valve vegetation. Patient 3 twice underwent contrast-enhanced MRI in the setting of infection within 3 weeks. First, he underwent contrast-enhanced MRI the same day that pneumonia was diagnosed on a chest radiograph. Two weeks later, he underwent contrast-enhanced MRI 4 days after the onset of fever and cellulitis. Patient 4 underwent contrast-enhanced MRI in the setting of chronic cellulitis treated with vancomycin at the time of imaging. Patient 5 underwent contrast-enhanced MRI of the foot that showed osteomyelitis. One day later, he again underwent contrast-enhanced MRI for evaluation of the brain, which showed normal findings.
Two individuals, patients 7 and 8, had renal failure and received MR contrast material. However, neither patient had a proinflammatory condition. The patient who did not receive MR contrast medium (patient 6) was a 64-year-old woman with breast cancer who suffered from calciphylaxis. This condition, which typically progresses from reticulated erythema to the development of bullae, necrosis, and ulcers, has previously been reported in a patient who developed NSF [11]. On the basis of the nature of this process, calciphylaxis could potentially have been considered a proinflammatory condition for the purposes of our study. However, because it was not mentioned in the prior study describing proinflammatory conditions in NSF [7], we did not include it in our list of proinflammatory states.
Statistical Analysis
Using the known number of infected renal failure patients who developed NSF
(i.e., five) and the projected number of renal failure patients with an
infection undergoing contrast-enhanced MRI during 2002–2006 (i.e., 75),
a 6.7% rate of NSF development among infected renal failure patients was
determined.
Given the known number of renal failure patients who developed NSF but did not have a documented infection (i.e., two) and the projected number of renal failure patients without infection undergoing contrast-enhanced MRI during 2002–2006 (i.e., 750), the estimated rate of NSF development was 0.26%. This 33-fold difference between the rates of NSF development in the two populations was highly significant, p < 0.001 by Fisher's exact test. The estimated odds ratio was 25 and the 95% CI was 25.7 (3.9–264.4).
We next estimated the rates of NSF development in infected patients with end-stage renal failure compared with infected patients without renal failure. For this calculation, we used the known number of patients with renal failure and a documented infection who developed NSF (i.e., five patients) and the estimated number of all renal failure patients with an infection during the 5-year period (75), which yielded a rate of 6.7%. We next estimated the number of patients without renal failure but with an infection who underwent contrast-enhanced MRI during the period 2002–2006 by applying the 9% rate of infection to the estimated 46,633 patients who underwent contrast-enhanced MRI during that period. This calculation yielded 4,197 patients, of whom none developed NSF, for a rate of 0%.
Estimated rates for developing NSF and 95% CIs for the four groups are as follows: patients with renal failure but no infection, 0.27% (0.07–0.96%); patients with renal failure and a documented infection, 6.3% (2.7–13.8%); patients with infection but not renal failure, 0% (0–0.07%); and patients with neither renal failure nor infection, 0% (0–0.006%).
For patients with an infection, the association of NSF with renal failure was highly significant, p < 0.0001 by Fisher's exact test. Furthermore, for patients without infection, the association of NSF with renal failure was also highly significant, p < 0.0001 by Fisher's exact test.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The major finding in our study is that the rates of development of NSF in renal failure patients undergoing contrast-enhanced MRI appear to be substantially higher than in renal failure patients without an infection. The differences between the two groups were statistically significant in our admittedly small sample. Our findings, although suggestive, require confirmation by a larger study before they are used as guidelines for decision making about the administration of MR contrast material to renal failure patients. Such confirmation could likely be obtained by a retrospective analysis of previous studies. Nonetheless, our findings, if confirmed, could have important ramifications for triaging renal failure patients for contrast administration for MRI.
Presently, the primary emphasis in determination of whether a patient with renal failure can be considered for MR contrast administration is on the degree of renal insufficiency. If the presence of infection indeed proves to be a risk factor for the development of NSF, then some renal failure patients presently judged to be acceptable risks for MR contrast administration on the basis of the degree of renal failure might be reconsidered as high-risk patients. Such factors have recently been addressed in a review article to which the reader is referred [12].
We recognize that our findings are based on a number of assumptions, such as extrapolation of some patient numbers over the 5-year period for which we have documentation of NSF cases. Furthermore, our study, like almost all reports of NSF patients, is limited by a small number of patients. Despite these limitations, our findings raise the distinct possibility that infection in patients with renal failure is a risk factor for NSF and it is appropriate for confirmation to be sought from other sources.
The mechanism by which gadolinium-containing MR contrast administration leads to NSF in some patients with renal failure is a matter of active debate. Some investigators have proposed that NSF may be triggered by a proinflammatory condition or event including vascular thromboses, surgery, or infection [7]. These authors have postulated that such proinflammatory conditions may create an inflammatory burden in patients who are predisposed to develop NSF on the basis of renal dysfunction. Specifically, the inflammatory or infective state may lead to the proliferation of cytokines, which recruit circulating fibrocytes subsequently leading to the clinical symptoms of NSF. One proposed mechanism is formation of gadolinium deposits within areas of cutaneous calcium phosphate deposition, a common finding in cutaneous biopsies of patients undergoing hemodialysis [13]. Such gadolinium deposits are thought to serve as a nidus for the development of NSF by acting as a target for circulating fibrocytes, which normally function in fibrotic processes such as wound repair and in response to infection [14]. According to one hypothesis, intracellular gadolinium deposited within various tissues serves as a surrogate target for circulating fibrocytes (Figs. 1A and 1B), which leads to dermal or systemic fibrosis [15]. Circulating fibrocytes have been described in skin biopsies of NSF patients [15] and have been implicated in the pathogenesis of NSF [16].
On the basis of the proposed role of a proinflammatory state in the development of NSF outlined, we set out to examine the frequency of these events in our clinical population. We found that infection was common in our patients at the time of administration of MR contrast agents, being seen in five of seven NSF patients to whom an MR contrast agent was administered. Major surgery was much less common, being seen in only one patient (who had coexistent infection). On the other hand, we did not find evidence of the other two proinflammatory events thought to be associated with NSF—that is, major surgery and vascular thrombosis. In our study population, both focal infections, such as osteomyelitis, and systemic infections, such as bacteremia, were seen. In most cases, the interval between the onset of infection and subsequent administration of MR contrast medium was a few days, but in one case of chronic cellulitis, the proinflammatory state began many months before, and continued during, the time of infusion of MR contrast material. Thus, no one typical duration or type of infection was seen in all patients.
In a previous study, investigators reported proinflammatory events at the time of development of NSF in 13 patients [7]. That study population is similar to ours in terms of the age of patients and prevalence of renal failure and contrast administration. Six patients in that study had an infection, and most of the remaining patients had vascular thrombosis. Some overlap in the infections seen in that study population and our patients exists; three patients in that study had osteomyelitis and one patient had pneumonia and bacteremia (compared with one patient with osteomyelitis and two patients with bacteremia in our study). Unlike our study, in which two patients had cellulitis at the time of contrast administration, the previously published study did not report any patients with cellulitis. Nonetheless, the potential importance of cellulitis in the development of NSF is worth mentioning. Because the skin is the major site of deposition of gadolinium particles in NSF, an inflammatory condition involving the skin is potentially an important condition predisposing to NSF. The initial stages of NSF are frequently mistaken for cellulitis. Similarly, calciphylaxis, which was seen in one of our patients, is a process arising within skin, one of the sites of fibrotic change in patients with NSF. In a previous report, the coexistence of calciphylaxis and NSF was postulated to be due to increased levels of transforming growth factor-β in patients with renal failure [11]. Transforming growth factor-β regulates signaling proteins that at once initiate fibrosis and, under appropriate conditions, calcification, which could explain the coexistence of calciphylaxis and NSF. The fact that cutaneous calcium phosphate deposition has been implicated in the development of NSF (as outlined earlier) also suggests that calciphylaxis, in which subcutaneous vascular calcification occurs, may be a risk factor for the development of NSF [13].
One patient in our series had both an infection and an open cholecystectomy, which was performed 7 days after contrast-enhanced MRI. However, this patient was found to have fibrotic changes consistent with NSF within abdominal tissues during the surgical procedure. Therefore, undiagnosed NSF must have existed before the surgery and, in this patient, surgery cannot be considered a risk factor for the development of NSF. Previous authors have commented on the possible role of major tissue injury, such as major surgery, as a risk factor for patients with renal failure and receiving MR contrast material in developing NSF [7, 17]. In one series of six patients who developed NSF after the administration of MR contrast material, all patients had undergone some form of surgery within the time period of 11 weeks before NSF was diagnosed [17]. In another series, it appears that five patients underwent major surgery at the time of MR contrast administration before the development of NSF [7]. In another series of patients, six of 12 patients had vascular surgery within 17 days after administration of MR contrast material [18].
Based on a review of the medical literature and our patients, the time interval between MR contrast administration and the development of NSF symptoms appears to vary widely among patients. In a previous report, all five NSF patients were noted to develop NSF symptoms within 2–4 weeks after MR contrast administration [4]. In another study, the mean time interval from MR contrast material exposure to the development of NSF symptoms was 25 days [5]. In yet another study, investigators reported that six of 15 affected patients developed NSF symptoms within 30 days of MR contrast administration [7]. Furthermore, symptoms can apparently follow very quickly after MR contrast administration. Two of our patients developed symptoms consistent with NSF within 1 week of MR contrast administration. Four of 13 patients in another study developed NSF symptoms within 7 days of administration of MR contrast material [5].
One specific MR contrast agent, gadodiamide, has been implicated in most cases of NSF. Gadodiamide was the only gadolinium-containing contrast agent available in the two European medical centers at the time of the earliest reports of MR contrast agent–related NSF in Europe [4, 5]. Gadodiamide is almost exclusively renally excreted and has a half-life that is increased from about 2 hours in patients with normal renal function to 30–120 hours in patients with chronic renal failure [19]. Gadodiamide is distinct from the four other U.S. Food and Drug Administration–approved MR contrast agents by having excess chelate (12 mg/mL) and being less stable [19]. Free gadolinium ions result from a process of transmetallation, which is facilitated by the excess chelate and relative instability of gadodiamide. This diminished stability, under conditions of prolonged circulation time due to renal impairment, may lead to deposition of free gadolinium within tissues and cause NSF, as has been suggested by some investigators [5]. Indeed, two recent articles have documented gadolinium deposition in the tissues of NSF patients [13, 20].
All six patients who were hemodialysis-dependent at the time of contrast-enhanced MRI were dialyzed 1 day after gadodiamide administration, suggesting that prompt hemodialysis may not be protective against the development of NSF. Other authors have described cases in which NSF occurred despite prompt dialysis [18, 21]. Furthermore, other investigators have also pointed out that hemodialysis after the administration of gadolinium agents may not have a protective effect [12].
As with any study, our study has some limitations. First, the number of patients in our study is small and our results need to be confirmed by evaluation of a larger number of patients. Second, our list of patients was compiled over years by a number of medical personnel, who added names of affected individual patients to the list. Some patients with NSF may not have been listed. In addition, some patients with NSF at our institution may not have been seen by the nephrology group that served as the patient referral base for our study population. Thus, for various reasons, the number of patients with NSF we report may underestimate the number of all NSF patients at our hospital. In that case, the rate of infection in the larger population of NSF patients may actually be lower than what we report. Third, it is possible that NSF patients with infection were more likely to be brought to medical attention, which would again overestimate the rate of infection.
In summary, we found a high rate of infection at the time of MR contrast administration in a small group of patients who later developed NSF. Among patients with renal failure, the association of NSF with infection status was highly significant. Our findings lend support to the hypothesis that infection is a predisposing factor for the development of NSF after exposure to gadolinium-containing contrast agents. Our results, if validated, would have important ramifications for NSF risk assessment in patients with impaired renal function who are being considered for administration of MR contrast material. As such, our findings deserve further study in a larger group of patients and in animal models of NSF.
Acknowledgments
We gratefully acknowledge the contributions of the medical staff at
Carolina Kidney Associates, Greensboro, NC; David C. Clark, Jr., Greensboro
Radiology, Greensboro, NC; and David DeLong, Duke University Medical Center,
Durham, NC, for statistical support.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  P. H. Kuo NSF-Active and NSF-Inert Species of Gadolinium: Mechanistic and Clinical Implications Am. J. Roentgenol., December 1, 2008; 191(6): 1861 - 1863. [Full Text] [PDF]
|
|
|
|
|
|
J. M. Provenzale Nephrogenic Systemic Fibrosis: Some Considerations on the Debate Regarding Its Cause Am. J. Roentgenol., December 1, 2008; 191(6): 1867 - 1869. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
