Multiple major morbidities and increased mortality during long-term follow-up after recovery from thrombotic thrombocytopenic purpura

Abstract

Recovery from acute episodes of thrombotic thrombocytopenic purpura (TTP) appears complete except for minor cognitive abnormalities and risk for relapse. The Oklahoma TTP-HUS (hemolytic uremic syndrome) Registry enrolled 70 consecutive patients from 1995 to 2011 with ADAMTS13 activity <10% at their initial episode; 57 survived, with follow-up through 2012. The prevalence of body mass index (BMI), glomerular filtration rate (GFR), urine albumin/creatinine ratio (ACR), hypertension, major depression, systemic lupus erythematosus (SLE), and risk of death were compared with expected values based on the US reference population. At initial diagnosis, 57 survivors had a median age of 39 years; 45 (79%) were women; 21 (37%) were black; BMI and prevalence of SLE (7%) were greater (P < .001) than expected; prevalence of hypertension (19%; P = .463) was not different. GFR (P = .397) and ACR (P = .793) were not different from expected values. In 2011-2012, prevalence of hypertension (40% vs 23%; P = .013) and major depression (19% vs 6%; P = .005) was greater than expected values. Eleven patients (19%) have died, a proportion greater than expected compared with US and Oklahoma reference populations (P < .05). TTP survivors may have greater risk for poor health and premature death.

Figure 1

BMI of the 57 patients who survived their initial episode of TTP. Data are from the time of their initial diagnosis of TTP. Black bars represent patients; gray bars represent the expected prevalence based on age, race, and gender-specific rates from the NHANES US reference population. Patients’ BMI during follow-up was not different from initial values (n = 52; median follow-up, 5.7 years; range, 0.2 to 16.5 years; P = .922).

Figure 2

Kidney function following recovery from TTP. GFR, estimated by the CKD-EPI equation^, in all 57 patients at the time of their most recent serum creatinine measurement (median follow-up, 5.3 years; range, 0.1 to 15.7 years), was not different from the expected prevalence based on age, race, gender, and BMI-specific rates from the NHANES US reference population (P = .397). The albumin/creatinine ratio was measured in 2011 on 37 (80%) of 46 surviving patients (median follow-up, 5.9 years; range, 1.1 to 15.7 years); it was also not different from the expected prevalence based on age, race, gender, and BMI-specific rates from the NHANES US reference population (P = .793).

Figure 3

Prevalence of hypertension, diabetes, SLE, and major depression following recovery from TTP. Black bars for pre-TTP data represent point prevalence for all 57 patients at the time of their initial diagnosis of TTP. Black bars for post-TTP data for hypertension, diabetes, and SLE represent point prevalence data for 43 (93%) of the 46 surviving patients who were assessed in 2011-2012, comparable to the 2-year cycles of NHANES point prevalence data. The black bar for major depression represents point prevalence data for 37 (80%) of the 46 surviving patients who were assessed in 2012. Gray bars represent expected prevalence based on the NHANES US reference population. Hatched bars represent lifetime prevalence from assessments over the entire follow-up period for 55 (96%) of the 57 surviving patients for whom post-TTP data were available. Lifetime prevalence proportions were hypertension (45%), diabetes (24%), and SLE (14%). Pre-TTP point prevalence of hypertension among patients (19%) was not different from the expected prevalence from the NHANES US reference population (16%; P = .442); post-TTP point prevalence among patients (40%) was greater than the expected prevalence of the NHANES US reference population (23%; P = .011). Both pre-TTP (9%) and post-TTP (14%) point prevalences of diabetes were not different from the expected prevalence of the NHANES US reference population (7%; P = .799; 10%; P = .439). Both pre-TTP (7%) and post-TTP (12%) point prevalences of SLE were greater than the expected prevalence of the NHANES US reference population (0.3%, indicated by the asterisk; P < .001). The post-TTP point prevalence of major depression (19%) was greater than the expected prevalence from the NHANES US reference population (6%; P = .005).

Figure 4

Mortality of the 57 patients following recovery from TTP. The probability of patient death (solid circles) is compared with the expected probability based on age, race, and gender-specific rates from the reference United States (US; +) and Oklahoma (OK; open circles) populations obtained from the CDC. Broken lines indicate the 95% confidence intervals (CIs) around the patient mortality. The probability of patient death and 95% CIs as well as the population probability of death for the US and Oklahoma were calculated by using the Kaplan-Meier methods with point-wise limits. The data for the probability of patient death at 1, 5, 10, and 15 years after their initial episode of TTP, with 95% CIs and expected probability of death based on the US and Oklahoma reference populations are presented in the box.