Disparities in renal replacement in lupus nephritis: Current practice and future implications


Health disparities are well known to exist in systemic lupus erythematosus (SLE). The incidence and prevalence of SLE have been reported to be higher among African Americans, Asians, and Hispanics compared to European Americans in the US (1–3). A younger median age of disease incidence and prevalence in African American and Hispanic women has been noted along with lower socioeconomic status (2–4). Significant ethnic disparities have also been reported in disease outcomes, with greater damage and disease severity in African American and Texas Hispanics than in whites and Puerto Rican Hispanics in the LUpus in MInorities, NAture versus nurture cohort along with different genetic associations (4). Among the protean manifestations of SLE, the development of lupus nephritis (LN) predicts poor survival and is more common and severe in African American, Hispanic, and Asian patients in the US (5). In addition, response to treatment has been shown to vary by age, sex, location, and race/ethnicity, with risk for more aggressive disease in specific groups. All of these issues point to an unmet need for identifying modifiable risk factors for outcome.

In this issue of Arthritis Care & Research, Devlin et al (6) examine the clinical factors and comorbidities associated with the initial choice of renal replacement therapies (RRTs) among 11,317 patients with LN end-stage renal disease (ESRD) identified in the US Renal Data System out of all the ESRD patients receiving any RRT during a 12-year span. The subjects were ages 18–80 years, and SLE was given as the cause of ESRD at enrollment into the US Renal Data System. The source of the registry data is a Medical Evidence Report Form completed by an attending nephrologist. As a result of this analysis, Devlin et al reported that hemodialysis (HD) was the predominant mode chosen (85%) and peritoneal dialysis (PD) constituted a smaller group (12.2%). Only 2.8% of the patients proceeded directly to transplant. In addition, over the time span of the study, there was a steady decline from 16.8% to 9.7% in the proportion of patients who were placed on PD, and there was a slight increase of 1.7% to 3.7% of patients receiving preemptive transplant in the same time period. On multivariate analyses, within the PD group, patients were more often found to be women, younger, white, employed, and privately insured. These PD patients also generally had better surrogates for health (higher hemoglobin and albumin levels) and were more hypertensive than the HD group, while comorbidities such as congestive heart failure, peripheral vascular disease, and the inability to ambulate were associated with decreased utilization of PD. Similarly, patients receiving preemptive renal transplantation were significantly younger, with a greater proportion of them being privately insured, employed, white, and non-Hispanic patients, who were less often from the South. There were no data available on the long-term outcomes specific to the renal replacement mode chosen.

Considering the skewed incidence of HD as initial renal replacement therapy over PD or preemptive renal transplant in LN, the obvious question to explore further is whether the preference is unique to LN. A search of the 2010 US Renal Data System database reveals that trends in using renal replacement therapy for all-cause ESRD are strikingly similar to those in LN. In general, patients who received PD as initial renal replacement therapy tend to be healthier, have Medicare as the secondary insurance provider and employer-sponsored private insurance as the primary provider, and be white (7). HD was the dominant modality in African Americans, and Medicare/Medicaid was commonly the primary insurance in patients receiving HD. Geographic differences in renal replacement therapy were also consistent with those observed by Devlin et al within the LN population. There is also an overall trend in the increased utilization of HD and steady decline in use of PD as an initial modality for renal replacement therapy over the years (7).

This suggests that renal replacement therapy choice in LN is not different from nationwide trends in all-cause ESRD. Data comparing renal replacement therapy modalities in LN are sparse and usually from observational studies. While it is clear that transplant has the greatest survival benefit and overall best outcome among renal replacement therapies, data comparing HD and PD are somewhat inconclusive (8). Since the 2 dialysis techniques are inherently different from one another, randomization and blinding is difficult. Hence, there are currently no clear guidelines to suggest superiority of one method over the other. The low level of preemptive transplant use is likely multifactorial. Possible factors that might explain this underutilization are physician concerns regarding LN recurrence in the transplanted kidney, disease activity of SLE, ability and readiness of a patient to commit to comply with lifelong immunosuppressive therapy, physician beliefs regarding the benefits of transplantation in certain populations, and the multistep evaluation process for transplant listing (9).

Devlin et al also report the fact that nephrologists acknowledge the underutilization of PD and home HD and the feasibility to place up to one-third of ESRD patients on PD. They also report that patient satisfaction is better with PD, which has a clear cost advantage over HD. So what are the possible reasons for underutilization of PD and home HD in LN and all other ESRD patients? In the LN population, there may be some perceptions among nephrologists that SLE disease activity is better controlled on HD with fewer relapses (8). There have been reports to suggest that HD may decrease disease activity in SLE because it may decrease T helper lymphocytes, remove middle molecules, and have an antiinflammatory effect (10–12). However, no data conclusively prove any of these theories. Another concern may be the increased chance of infections with PD in an already immunocompromised SLE patient (8).

A study by Mehrotra et al suggests that predialysis counseling for the patient plays a significant role in therapy choice, and it is speculated that up to one-third of ESRD patients are never offered PD as their initial treatment modality for renal replacement therapy (13). Therefore, along with other medical factors, this failure on the part of a health care provider to educate patients about these options may be an important reason for underutilization of PD and home HD. There have been speculations about the profitability of in-center HD as compared with PD inherent to the traditional payment structure by the Centers for Medicare and Medicaid Services (CMS), which paid separately for dialysis, injectable and oral medicines, and diagnostic tests (14). However, CMS has proposed a new composite bundle system for ESRD payment. Components of this system include reimbursements for predialysis educational sessions and for covering the costs of dialysis, medications for dialysis, and related diagnostic tests (15). This may encourage increased utilization of PD. This new system would also encourage home HD. Therefore, it will be noteworthy to see if the trends change over the next few years.

Racial disparities in transplant have often been attributed to lack of organ donation practices in minority communities. However, an interesting statistic to note from the 2010 US Renal Data System is that the living and deceased kidney donation rate was highest in African Americans, and yet the projected wait list time for transplant was also longest in this group (7). While this might just reflect the disparities between incidence of all-cause ESRD and organ donation among African Americans, it still highlights the need to deliver better health care access to ethnic minorities, especially since ethnic minorities are often subject to delayed referral to nephrologists (16). This may be of special importance in LN populations considering the fact that these patients are good candidates for transplant and recurrence rate of LN in allograft is low (8).

Ultimately, the ideal future scenario would be the development of a treatment for LN that would prevent progression to ESRD with minimal toxicity or even prevent the development of LN altogether. A scenario such as this would require a way to identify patients at risk for nephritis early in the disease course and a way to predict those patients who may have the worst outcomes. The disparities in development of LN in predominantly US minority populations cannot be completely explained by differences in access to care alone. Poorer response to traditional therapy with intravenous cyclophosphamide has been reported in African American patients (17, 18). Once nephritis is present, studies have shown a greater likelihood of progression to ESRD among African American and Hispanic patients than whites (19). For LN in Asians, better outcomes and responses to cytotoxic therapy have been observed (20). These ethnic differences may be associated with a predisposition toward kidney failure following an insult to the kidney, which may involve genetic factors (19). Ongoing work in the discovery and understanding of genetic associations in LN may help us to identify SLE patients at risk for nephritis, progression to ESRD, and early mortality. It may be possible to potentially leverage these technologies for prompt recognition of high-risk patients and subsequent implementation of early therapy, as well as the ability to tailor therapy based on predicted disease severity. This is a particularly important therapeutic goal as many patients who may receive cytotoxic therapy are young and of reproductive age.

In summary, this clever epidemiologic study does confirm many of the disparities in choice of initial renal replacement therapy with LN not being different from all-cause ESRD. Renal replacement therapy, in particular HD, is a treatment modality with a high cost burden to the US health care system. The reasons for initial choice of renal replacement therapy are always complex and sometimes involve uninformed decisions, fears, biases, and real disease-related issues. Multivariate analyses cannot take into account all of the factors that are involved in decision making regarding choice of initial renal replacement therapy. Hence, additional types of studies including well-designed qualitative studies need to be conducted to understand these factors better. Ultimately, it does raise challenges to the lupus community and to those individuals who determine priorities for funding research support. How can we prevent ESRD in an already intensively studied condition such as SLE? What other predictors of ESRD have not yet been identified? Are they genetic or environmental? Further studies will need to take these issues into consideration, given the cost implications to the US health care system.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication.


We are grateful to Dr. Daniel J. Wallace for his advice and guidance during the preparation of this editorial.