Changes in the Incidence of End-stage Renal Disease Due to Lupus Nephritis, 1982-1995 FREE

NEPHRITIS develops in up to 60% of the patients who have systemic lupus erythematosus, and is a major cause of morbidity in these patients.¹ Twenty percent of the patients who have lupus nephritis progress to end-stage renal disease (ESRD) within 10 years of its onset.^2- 7 Numerous treatment regimens to improve the outcomes of patients with severe lupus nephritis have been studied, the most promising of which include cytotoxic medications, particularly cyclophosphamide.^8- 12 Over the decade of the 1980s, regimens for the treatment of severe lupus nephritis that included cytotoxic medications changed from being experimental treatment to accepted practice.^{6- 7,13- 20} However, it is unknown if the availability of these newer treatment regimens has influenced the rate at which patients with lupus nephritis develop ESRD. To address this question, national trends in the incidence of treated ESRD due to lupus nephritis from 1982 to 1995 were examined.

Data used in this study were those of the US Renal Data System. The US Renal Data System, operated by the National Institute of Diabetes and Digestive and Kidney Diseases and the Health Care Financing Administration, has collected information on all patients in the United States who have received Medicare-reimbursed maintenance renal replacement therapy since 1977.²¹ This database is estimated to include approximately 94% of the patients who receive renal replacement therapy in the United States. As a result of changes in reporting procedures, patients with incident ESRD that was not Medicare-reimbursed have also been included in the database since 1994.

To be included in the database, patients must be using long-term dialysis or have undergone renal transplantation. Patients who receive dialysis for acute renal failure, who die of renal failure before receiving dialysis or a renal transplant, or who do not accept renal replacement therapy are excluded. The date on which a patient began long-term dialysis or received a renal transplant was considered the date of onset for determining the incidence of ESRD.

The database includes information on patient demographic characteristics, hospitalizations, and mortality. Information on treatment prior to the development of ESRD was not recorded. Detailed information on the primary renal disease causing ESRD, including lupus nephritis, was first collected in 1982. The diagnosis of the primary renal disease was a clinical attribution by the nephrologist supervising the patient's care, as reported on the Health Care Financing Administration Medical Evidence Report, and was not necessarily based on histological evidence. Information on the primary renal disease was missing for 5.8% of the patients. Patients in the database are followed up from the start of renal replacement therapy until death; the database is updated annually. In this study, data were included on incident patients from 1982 to 1995, the most recent year for which data were complete.

The crude incidence of ESRD due to lupus nephritis was calculated by dividing the number of incident cases in each year by the estimated resident population of the United States in the same year, as reported by the US Bureau of the Census.²² To reduce confounding by changes in the demographic composition of the population over time, standardized incidence rates in each year were also computed. These rates were adjusted to the age, sex, and ethnic composition of the resident US population in 1990, using direct standardization. Adjustment for age was based on 5-year age groups from age 0 to 84 years; patients 85 years and older were considered a single group. Standardized incidence rates included only persons of white, black, Native American, and Asian/Pacific Islander ethnicity, as census data do not include information on the age and sex composition of the groups with "other" or "unknown" ethnicity. Hispanic ethnicity was not recorded in the US Renal Data System.

The number of incident patients receiving renal replacement therapy has increased markedly over the past 20 years, an increase attributed to increases in both the need for and acceptance of such therapy.²¹ Changes over time in the acceptance of renal replacement therapy by patients with end-stage lupus nephritis may also have occurred apart from any changes in the incidence of ESRD due to lupus nephritis. To account for such changes in the acceptance of renal replacement therapy over time, the age-sex-ethnicity standardized incidence of ESRD due to lupus nephritis was compared with that of the age-sex-ethnicity standardized incidence of ESRD due to all primary renal diseases other than lupus nephritis from 1982 to 1995. The 1990 US resident population was also used as the standard population for these incidence rates. Relative risks of incident ESRD due to lupus nephritis, compared with other primary renal diseases, were computed for each year, along with 95% confidence intervals that accounted for standardization.²³ Changes in these relative risk estimates over time indicate the extent to which the incidence of treated ESRD due to lupus nephritis had increased or decreased relative to changes in the incidence of treated ESRD due to other causes. This analysis provides an adjustment for changes in the acceptance of renal replacement therapy over time. These relative risk estimates would be expected to decrease over time if the incidence of ESRD due to lupus nephritis was decreasing. Linear regression analysis was used to examine trends in the relative risk estimates. Analyses were performed using SAS programs (SAS, Cary, NC).

Most of the patients with ESRD due to lupus nephritis were young or middle-aged women, reflecting the epidemiological characteristics of patients with systemic lupus erythematosus (Table 1). The number of incident cases rose from 261 patients in 1982 to 842 patients in 1995, and the crude incidence rose from 1.13 cases per million person-years to 3.20 cases per million person-years over this 14-year study period.

The adjusted incidence rate, standardized to the age, sex, and ethnicity composition of the 1990 US population, increased from 1.16 cases per million person-years in 1982 to 3.08 cases per million person-years in 1995 (Figure 1). Incidence rates were highest among those aged 20 to 44 years and among blacks (Figure 2). The incidence rates increased over time in all age groups, except those younger than 20 years, in whom the incidence rate has remained stable since 1988. Incidence rates also increased over time in all sex-ethnic subgroups, although the increase was small among white females.

The increase in incidence of treated ESRD due to lupus nephritis paralleled that of the incidence of treated ESRD due to other causes, resulting in relative risk estimates that remained stable over time (Figure 3). The relative risk of ESRD due to lupus nephritis ranged from 0.011 to 0.015, indicating that patients with lupus nephritis comprised slightly more than 1% of the incident cases of treated ESRD in each year. There was no evidence of a trend in the relative risk estimates over time (t = −0.46; P = .66).

Because standardized relative risk estimates can be skewed by marked differences in the compositions of the groups being compared, and because few young or elderly patients had ESRD due to lupus nephritis, relative risks were also computed using incidences based only on persons aged 20 to 64 years. In this subgroup, the estimated relative risks ranged from 0.015 to 0.022, indicating that patients with lupus nephritis comprised about 2% of the incident cases of ESRD in each year in this age group (Figure 3). In this subgroup, there was a slight increasing trend in the relative risk estimates over time, indicating that relative to other primary renal diseases, lupus nephritis was more commonly the cause of ESRD later (t = 2.13; P = .06).

This study indicates that the incidence of treated ESRD due to lupus nephritis has increased steadily from 1982 to 1995. This increase was observed in all demographic subgroups of patients, except those younger than 20 years. The rate of increase kept pace with the increasing incidence of renal failure due to all other causes over time, resulting in relative risk estimates that were stable over the 14-year study period.

The incidence of treated ESRD due to lupus nephritis increased despite the introduction and recognition of efficacious new treatment regimens for lupus nephritis since 1982, particularly regimens that include cytotoxic medications.^8- 12 The findings of this study may indicate that these treatments have not been widely used, or that the benefits observed in clinical trials are not commonly replicated when these treatment regimens are used in clinical practice. Use of cytotoxic-based regimens may be limited because some physicians may be reluctant to use these medications, or patients may judge the risk-benefit balance of these treatment regimens to be unfavorable. A recent survey of European rheumatologists and nephrologists reported that more than 70% of physicians recommended that patients with severe lupus nephritis be treated with regimens that included cytotoxic medications,²⁰ suggesting that these medications may be in wide use. Unfortunately, there are no population-based or community-based studies of the treatments received by unselected patients with lupus nephritis, so it is difficult to judge the extent to which limited use of these regimens may have contributed to the increasing incidence of ESRD due to lupus nephritis.

Some observational studies of cytotoxic-based treatment regimens in patients with lupus nephritis have reported results comparable to those of clinical trials, with preservation of renal function and improved survival.^16- 17 However, other studies have noted limited effectiveness of these treatments, with 40% to 50% of the patients progressing to ESRD within 4 years.^7,24 Differences in the types of patients included in the clinical trials and those treated in clinical practice may explain some of this difference in outcomes, as may modifications of the dose or duration of treatment. Frequent relapses after successful treatment may also limit the long-term effectiveness of these regimens.²⁵

This study may have underestimated the influence of newer treatment regimens on the development of ESRD in patients with lupus nephritis if the outcomes of treatment are substantially delayed. Thus, patients who developed ESRD in 1995 may represent the failure of treatment practices in common usage in 1990 or earlier. More widespread use of cytotoxic-based treatment regimens during the 1990s may portend a decrease in the incidence of ESRD due to lupus nephritis in the future. It is also possible that in more recent years, more patients with lupus nephritis have presented with advanced renal insufficiency, and were less likely to benefit from cytotoxic-based treatment regimens, but data to support this hypothesis are lacking.

The strengths of this study include the use of multiyear data on a national sample, and comparisons of the incidence of end-stage lupus nephritis with changes in the incidence of ESRD of other causes. However, the study also has some limitations. The US resident population was used as the population at risk for ESRD due to lupus nephritis, rather than the subgroup of persons with lupus nephritis. This choice allowed comparisons between trends in the incidence of ESRD due to lupus nephritis and ESRD due to all other causes, and was necessary because population-based estimates of the prevalence of lupus nephritis are unavailable. However, this choice enlists the assumption that changes in the size of these groups over time have been similar. If the number of patients who have lupus nephritis has increased more rapidly than the size of the general population, this study would have overestimated the increase in incidence of ESRD due to lupus nephritis. However, the number of patients who have lupus nephritis would have had to increase at a rate 2.6 times faster than the general population over the period from 1982 to 1995 for the incidence of ESRD due to lupus nephritis to have remained at the 1982 rate of 1.1 cases per million, and would have had to increase at a rate 4.5 times faster than the general population to obscure a 50% decrease in the incidence of ESRD among patients who have lupus nephritis over this time. Increases of these magnitudes in the prevalence of lupus nephritis, due to an increased incidence, longer survival of patients, or more correct assignment of the cause of ESRD by nephrologists, have not been noted.²⁶ One report of an increasing prevalence of systemic lupus erythematosus noted a lower prevalence of nephritis among patients in recent years, suggesting that much of the increase was due to the recognition of milder cases of systemic lupus erythematosus rather than to substantial changes in the prevalence of lupus nephritis.²⁷

This study suggests that the availability of cytotoxic-based treatment regimens has not yet influenced the rate at which patients with lupus nephritis develop ESRD. Learning whether this may be due to limited use or limited effectiveness of these treatments will be important in efforts to improve the outcomes of patients with lupus nephritis.

Accepted for publication January 28, 2000.

This study was supported in part by the Maudie Long Gift Fund at the Veterans Affairs Palo Alto Health Care System, Palo Alto, Calif.

Data reported herein were supplied by the US Renal Data System. The interpretation and reporting of these data are the responsibility of the author and in no way should be seen as an official policy or interpretation of the US government.

Corresponding author: Michael M. Ward, MD, MPH, Veterans Affairs Palo Alto Health Care System 111G, 3801 Miranda Ave, Palo Alto, CA 94304 (e-mail: mward@leland.stanford.edu).