Abstract
Nephrogenic systemic fibrosis is a debilitating disorder seen in chronic kidney disease patients and is characterized by stiffening of the joints and thickening of the skin. Treatment options are limited, but some patients have had an improvement of their clinical symptoms after renal transplantation and the use of immunosuppression. Although there is a variable response to renal transplantation, it is currently unknown what factors promote a favorable outcome. Our objective was to evaluate if the response to renal transplantation was superior to other treatment modalities and to determine which characteristics allowed for a positive response to occur. We retrieved the data from the literature of 298 reported patients, compared the response to renal transplantation and to other treatments, and analyzed their characteristics. We found that more patients had a higher response to renal transplantation, as determined by softening of the skin and improved joint mobility, and among those that did respond, they had a shorter dialysis vintage. We suggest that if renal transplantation is to be considered as a treatment modality, it should be initiated at the earliest possible in the course of the disease to achieve maximum clinical benefit.
1. Introduction
Nephrogenic systemic fibrosis (NSF) is primarily seen in chronic kidney disease (CKD) patients and is characterized by large areas of hardened skin with slightly raised plaques, papules, and hyperpigmentation. Biopsies of the affected areas show increased numbers of fibroblasts, alteration of the normal pattern of collagen bundles, and often increased dermal deposits of mucin [1]. The skin may have a “cobblestone” [2], “woody” [3, 4], or peau d’ orange appearance [5]. These lesions may be pruritic and accompanied by sharp pain or burning sensations [3, 6]. Movement of the joints may be so severely limited by the fibrosis that the flexibility is lost.
The first cases were noted between 1997 and 2000 in hemodialysis (HD) patients or patients with a failed renal allograft who developed severe skin indurations that were initially thought to be scleromyxedema [7, 8]. Since Grobner reported a correlation between the use of gadolinium (Gd) in end-stage renal disease (ESRD) patients and NSF [9], additional reports supporting this correlation have been published [10, 11]. It is now recognized that low-stability gadolinium-based contrast agents, in the presence of permissive factors, most likely trigger the disease. The prevalence of NSF after exposure to gadodiamide (Omniscan) has been reported to be between 3% and 7% in patients with reduced renal function in several different studies [12–14]. In a study evaluating ESRD patients over an 18-month period [15], it was found that the incidence of NSF was 4.3 cases per 1000 patient years and each radiologic study using gadolinium presented a 2.4% risk for developing the disease. Other studies have shown that the exposure to gadolinium-based contrast agents was associated with an increased risk of developing cutaneous changes of NSF (odds ratio, 14.7; 95% confidence interval, 1.9–117.0) compared with patients who were not exposed to gadolinium [16]. Patients suffering from NSF were found to have higher mortality rate compared to matched controls with no cutaneous lesions. It was found to be as high as 48% over a 24-month followup [16]. As of May 2011, over 335 cases of NSF have been reported to the International NSF Registry at Yale University [1].
There is no proven treatment available to patients suffering from NSF. Although various treatments such as high-dose intravenous immunoglobulin [17], plasmapheresis [18], extracorporeal photopheresis (ECP) [19, 20], imatinib [21], and sodium thiosulfate (STS) [22] have been reported to provide benefit, it has only been shown in a limited number of cases, and therefore no generalizations can be made. However, the progressive course of NSF may be altered by improving kidney function as the remission of NSF has been described in patients who recovered from acute kidney injury [4, 23, 24] or after receiving a kidney transplant [25]. The response to transplantation has been variable ranging from a complete resolution of symptoms [9, 26] to minimal or no improvement reported in some patients [27, 28]. Although there are more documented cases of clinical improvement of NSF after renal transplantation, it is still unknown what factors confer a favorable response to transplantation versus those that do not. The database of 298 patients that we have formulated from the existing literature compares the response to renal transplantation and to other treatments, with an analysis of their characteristics. We attempted to determine if there were any patient factors that could predict a better response to transplantation and if this treatment modality was superior to others with respect to better clinical outcomes.
2. Patients and Methods
PubMed, MEDLINE, Google, Google scholar, and Embase databases were searched for articles including early release publications, limited to a date range of January 1, 2000 through May 1, 2011. Keywords and medical subject heading (MeSH) were used in the search. MeSH term used was “nephrogenic fibrosing dermopathy” and search keywords included “nephrogenic systemic fibrosis,” “nephrogenic fibrosing dermopathy,” “case report,” “dialysis-associated dermatopathy”, “renal transplant” and “survival.” Abstracts of retrieved citations were reviewed and prioritized by relevance. Full articles were obtained and reference lists were reviewed for additional articles when appropriate. Data from these studies on benefits of different NSF treatment modalities on both patient survival and disease response along with patients’ characteristics were extracted.
Inclusion criteria incorporated articles describing a newly or previously diagnosed case of NSF, delineation of the management, NSF response to treatment, time of followup, and patient outcomes. The exclusion criteria included review articles, articles with insufficient data, and papers limited to biopsy studies only due to the lack of clinical information. The manuscripts reviewed were not just limited to English; even non-English language papers were translated and analyzed. Both the dialysis-dependent patients as well as patients who were not on maintenance dialysis were included. To reduce the chance of duplication, data regarding patient characteristics were compared to the articles wherein they were originally reported. Treatment modalities of NSF patients were divided into four groups, namely, those who either continued on maintenance dialysis or received no treatment at all (T1), those who received treatment other than renal transplantation (T2), patients who received only renal transplantation as a treatment modality (T3), and those patients who received alternative treatment initially which was followed later by renal transplantation (T4). The unanalyzed and unreported data from the original publications were considered as missing data.
Time to treat was defined as the duration between the beginning of NSF symptoms and any treatment provided to each group, with zero weeks being the time to treat in patients who received maintenance dialysis only, and time to dialysis for those who were initiated on dialysis either as maintenance dialysis or as a treatment for acute kidney injury (AKI) after NSF was diagnosed and did not receive any other NSF directed therapy (in group T1). Improvement was defined as a complete resolution of symptoms, softening of the skin, improved joint mobility, and improved functional status, as outlined in respective publications. Decrease of edema alone was not considered as a response to treatment. Skin biopsy after treatment was neither performed nor reported in the original case reports. Detailed biopsy results and immunohistochemical staining were reported in only a few articles. Most of the cases that were labeled as NSF were initially biopsy proven NSF.
Statistical analysis was conducted using NCSS statistical program. All categorical variables were analyzed for statistical significance using chi-square test. Noncategorical variables were analyzed using analysis of variance. All noncategorical variables were presented as mean ± standard error (SE) of mean. To evaluate the outcome of various treatments on NSF, data were analyzed using Cox-regression analysis with response to treatment considered as an event, no response during followup as censored, and follow up as reported in the literature using the Efron method to analyze time ties. Hazard ratio (HR) was calculated to evaluate the response rate and two-sided log rank probability test ( value) to evaluate for statistical significance.
3. Results
Of the articles retrieved, 157 were case reports, and of these, 95 had sufficient information with respect to the inclusion criteria [2, 3, 5, 6, 9, 10, 14–16, 18, 19, 21–23, 25, 26, 28–106]. After application of the inclusion and exclusion criteria, 298 patients were selected (Figure 1).
Patients’ age range was from 8 to 87 years with a mean of years (mean ± SE), and 46.89% () were male. The response to treatment was monitored for an average of weeks (1–336 weeks), and patient survival was followed up to an average of weeks (1–336 weeks). The time until any treatment was initiated following the diagnosis of NSF was weeks (0–364 weeks). Patients on maintenance dialysis were 255 (86.4% on HD) for a duration of months (0–385 months). After extensive review of patients’ medication and imaging studies, authors of the original case reports reported 153 patients who had Gd exposure, while 15 patients were never exposed to Gd.
T3 and T4 patients were younger when compared to the other two groups (); they also had the longest duration until treatment (). No statistically significant difference was found among the groups regarding gender, race, dialysis vintage, prior renal transplantation, and followup duration. All patients who received renal transplantation regardless of previous other treatments had ESRD at the time of NSF diagnosis (Table 1). Patients who received renal transplantation showed significant improvement when compared to the other two groups. Group T2 (i.e., treatments other than transplantation) also demonstrated a higher improvement rate when compared to no treatment. The same results were retrieved after correcting for age and time to treat. Table 2 compares all the treatment modalities.
The differences in patients’ characteristics between those who responded and those who did not during the followup period were also analyzed (Table 3). Patients who responded to any treatment were younger, had lesser dialysis vintage, had a longer time to treat, and had a higher rate of receiving a renal transplant previously. This longer time to treat effect contributed to a better outcome as observed among those renal transplantation patients who had longer time to treat.
Of the 26 patients treated with renal transplantation, data regarding dialysis vintage were available on only 23. Patients who were dialyzed for less than four years had a better response rate after renal transplantation. No difference was found regarding age, gender, duration to treat, and prior renal transplantation. Only eight patients’ race was reported therefore making the racial distribution analysis imprecise (Table 4). An interesting observation regarding the use of any other treatment prior to transplantation was noticed. Eight out of nine patients treated only with renal transplantation improved compared to 11 of the 17 patients who received renal transplantation preceded by other NSF specific treatment. Those patients who only received a renal allograft were dialyzed for compared to months, were transplanted earlier ( compared to weeks) and showed a faster clinical response (20.5 compared to 78.6 weeks ( for all)).
4. Discussion
Patients with NSF generally have a poor clinical course as the fibrosis affecting the joints and skin often leads to contractures, pain, and immobility. Although various treatments have been tried on patients, most have resulted in variable clinical responses, and overall consistency is lacking. For instance, the response to ECP, which induces monocyte-derived tumor necrosis factor-alpha and suppresses collagen synthesis and enhances collagenase production, was demonstrated in a case report, wherein 2 out of 3 patients had a marked resolution of skin lesions and normalization of skin distensibility after undergoing from 8 to 16 cycles at two to four week intervals [48]. In another case report [19], a patient who was two years after kidney transplantation showed significant improvement of the sclerodermiform brownish plaques on her extremities after 4 cycles of ECP. Although ultraviolet-A phototherapy (UV-A) has the ability to inhibit procollagen synthesis in human skin [107], the response has been variable [8, 19, 47, 108, 109]. Furthermore, there has been improvement reported after three to five days of plasmapheresis treatment in three patients [53]. A decrease in the transforming growth factor-beta-1 levels following plasmapheresis presumably led to the amelioration of this clinical condition. Imatinib, a treatment modality used in patients with chronic myelogenous leukemia (CML) and noted to cause a regression of concomitant bone marrow, pulmonary, hepatic, and kidney fibrosis, has also been tried [110]. Studies have shown a strong dose-dependent in vitro and in vivo inhibition of the synthesis of major dermal extracellular matrix proteins [111]. Kay and High [21] treated two patients with stage 5 CKD and NSF with oral imatinib mesylate at a dosage of 400 mg/day and found that imatinib mesylate treatment decreased fibrosis and resulted in a relatively rapid and steady improvement of the skin changes. However, these skin changes recurred after the discontinuation of the drug. Henceforth, they recommended a further evaluation to determine the duration of treatment. Imatinib mesylate has been tried as a single agent and showed marked improvement in NSF patients [112]. There are two case reports demonstrating success with using STS as well [22, 67].
It appears that renal function may be correlated with the course of NSF since remission has occurred after the recovery of renal function following renal transplantation or from resolution of an acute kidney injury [4, 23, 24] and the gradual reversal of the disease in patients not requiring dialysis. Our review of the literature showed that there are 26 NSF patients who received a renal transplant, of which 19 patients showed improvement after transplantation, leading to a response rate of 73.08%. This represents a higher number of successfully treated NSF patients when compared to other treatments used thus far. The response rate observed in the group which received renal transplantation is likely to be as a result of restoration of renal function, although other factors may have contributed as well. The half-life of gadolinium in patients with normal kidney function is approximately 90 minutes, but in patients with advanced renal impairment, the elimination half-life can be prolonged to more than 30 hours [113]. Although, there is no evidence that hemodialysis immediately after gadolinium exposure lowers the risk or severity of NSF, studies have reported that gadolinium removal was up to 99% after three dialysis sessions [114–116]. Restoration of kidney function results in Gd excretion and detection in the urine, even if the contrast agent was administered up to three years prior to transplantation [54, 117], suggesting that mobilization from tissue and bone deposits [115, 118] can occur after transplantation. This weakens the evidence regarding intensive hemodialysis after exposure. If Gd was indeed the trigger for NSF, the removal of the offending agent should, in theory, ameliorate the inflammation.
Topical, intralesional, or oral glucocorticoid therapy and cyclophosphamide have, in general, shown no benefit in some studies [4, 8]. To the contrary, our observation suggests that immunosuppressive agents, including steroids, may play an important role in treating patients with NSF as the results favor a higher clinical response rate to transplantation. It has been demonstrated that NSF patients express phospho-70-s6 kinase and transforming growth factor-beta (TGF-β), which are targets to rapamycin and mycophenolate (MMF) therapy, respectively. It is possible that there is a downregulation of these proteins in the post-transplantation period, halting the progression of inflammation. However, MMF therapy alone may not confer complete benefit unless used in conjunction with renal transplantation [52].
Renal transplantation may offer a hope for a successful treatment in suitable NSF patients, although efficacy is unproven in some reports [119]. A single center-based study by Leung et al. [28] failed to show any benefit with renal transplantation when compared to those patients who remained on dialysis. This observation is contrary to previously published reports [9, 32, 105]. We found that patients who underwent renal transplantation had a higher chance of a positive response, documented as a clinical improvement, when compared to patients who either continued on maintenance dialysis or those who received alternative forms of treatment.
It has been previously reported by Swartz et al. [23] that there is no association between the risk of development of NSF and patients’ age, sex, race, or etiology of renal disease. Our analysis of the literature found that age plays an important role in determining which patients were likely to respond to treatment and have a favorable outcome. Those patients who responded to any treatment were younger and had a shorter dialysis vintage (). Interestingly, patients who improved had a longer time to treat (). This could be explained by the high percentage of response in the renal transplant group that had the longest time to treat. A possible explanation is that ESRD and dialysis are proinflammatory states [120]. Many factors play a role in the inflammatory state including membrane bio-incompatibility, the use of unpure or nonsterile dialysate, and poor dialysis adequacy. Frequency and duration of dialysis was also found to contribute to the inflammatory state in these patients [121–123]. C-reactive protein (CRP) and IL-6 are the most commonly used biomarkers to determine the uremic inflammatory state in ESRD patients. Panichi et al. estimated CRP levels at baseline and at 6 months in long-term HD patients on various modalities using different dialyzer membranes. This analysis showed that baseline CRP values were high (>5 mg/L) in 47% of the patients and remained elevated even up to 6 months, which was dependent on the type of the extracorporeal modality (hemodiafiltration versus hemodialysis) and membrane used [124]. Another study involving pediatric patients on maintenance HD demonstrated that 30 minutes after HD treatment, the levels of TNF-alpha and IL-1-beta were significantly higher when compared to their pre-HD levels (). Likewise, a positive correlation between dialysis duration and serum levels of CRP (), IL6 (), and IL-10 () was also noticed [125]. This was thought to be due to a persistent uremic state. Gadolinium, either in the free ionic state, Gd3+ or Gd-chelate complex form, stimulates the release of cytokines from activated macrophages in skin and peripheral monocytes in blood, respectively. The released cytokine shower, in turn, triggers the bone marrow derived circulating fibrocytes to induce fibrosis in NSF. Collectively, this suggests that dialysis vintage and an inflammatory state are complimentary to one another. In relation to NSF, it is possible that dialysis vintage in the presence of gadolinium maintains a proinflammatory state which promotes a continuous supply of cytokines in the circulation leading to the prolongation of the inflammatory lesions. We found a statistically significant relationship between dialysis vintage and response to treatment (). Analysis of 23 patients (3 patients with missing data) who received renal transplantation showed that a better response was observed in those who were dialyzed for less than four years ().
The major weakness in our observational study is publication bias. Since no clinical trial establishing the link between Gd and NSF can ever be conducted in human subjects, we are limited to the existing database of published cases to draw conclusions and render treatment. In fact, most of the literature on NSF is in these forms of publications. Furthermore, journals tend to publish reports showing predominantly positive results. The cases that have been reported in the literature provide limited details with respect to the immunosuppressants used or the type of transplant (i.e., live donor versus cadaveric) received by the patient. There may be other confounding factors that may not have been accounted for. The follow-up period following the success of a treatment modality has not been well delineated in most of the case reports. Therefore, there exists a possibility of the recurrence of NSF that may not have been accounted for. We recognized these issues and hence included all cases of NSF including the reports showing both positive and negative results regarding each treatment modality to minimize publication bias. Every attempt was made to include as many unique patients as we could, resulting in 298 patients, which is approximately 80% of the number of patients’ registry at Yale University.
5. Conclusion
Nephrogenic systemic fibrosis is a debilitating disease in patients with impaired renal function. Currently, there is no standard treatment, but improvement has been reported after restoration of kidney function. We compared the clinical response rates to various treatments and found that the renal transplantation group had a higher chance of clinical response to treatment when compared to dialysis alone or to any other treatment. Henceforth, based on our analysis, it appears that the renal transplantation is likely a beneficiary treatment modality that can be offered to NSF patients.
Conflict of Interests
The authors declare that they have no conflict of interests.