Treatment delay and outcomes in elderly thrombotic thrombocytopenic purpura patients: A nationwide analysis

Abstract

Thrombotic thrombocytopenic purpura (TTP) is a rare life-threatening thrombotic microangiopathy that is characterized by the deficiency of the ADAMTS13 enzyme. According to large TTP registries, the median age at diagnosis is 42 years for immune TTP but a wide range of age groups are susceptible (9–78 years).1 Recent studies have described atypical presentations of immune TTP in elderly patients, with a propensity for more severe organ damage and less profound thrombocytopenia.2 In addition, a study that attempted to validate the PLASMIC score for pretest probability prediction in various age groups found reduced sensitivity in the elderly (>60 years) population.3 Hence, TTP diagnosis in elderly patients can prove challenging due to overlapping clinical features with other conditions of the elderly, leading to delay in initiation of life-saving treatments like therapeutic plasma exchange (TPE). To assess the effect of age on treatment delay and other major hospital outcomes, we retrospectively analyzed the National Inpatient Sample (NIS) database (2012–2019) and identified all patients aged ≥18 years with a diagnosis of TTP who underwent TPE during their hospital stay. We used ICD-9 and ICD-10 codes to identify TTP, TPE, comorbidities, and adverse hospital outcomes (Data S1). Age was subclassified into three groups (18–39, 40–59, and ≥60), similar to those used by Liu et al in their study, and time to initiation of TPE and other adverse hospitalization outcomes were compared.3 The primary outcome of this study was a delay in initiation of TPE for 2 or more days as a surrogate marker for delay in diagnosis and treatment. Categorical variables were compared in the unadjusted analysis using χ2 test. Means were compared using Welch's analysis of variance (ANOVA). Post hoc testing for in-between group analysis was done with Bonferroni correction as deemed appropriate to ascertain statistical significance. Multivariable logistic regression was performed for each primary and secondary outcomes of interest for adjusted analysis between age groups and to assess other predictors of mortality and delay in initiation of TPE in the study population. All statistical analyses were performed using IBM SPSS statistics version 26. Between 2012 and 2019, the NIS database included 3103 TTP patients aged 18 or above who underwent TPE during their hospital stay. One hundred eleven patients (3.6%) did not have data on the day of initiation of TPE and were excluded from the study. The remaining 2992 patients were subclassified by age groups and were included in our final analysis. The median age of the population was 49 years (range 18–90) and there was higher prevalence of females compared to males (66.3% vs. 33.7%). There was 999, 1142, and 851 patients in the 18–39, 40–59, and ≥60 age groups, respectively. Baseline patient and hospital characteristics are delineated in Table S1. There was a trend toward higher prevalence of males (30.5% vs. 33.1% vs. 38.1%, p-value of .003) and white race (36.9% vs. 39.3% vs. 59.3%, p-value of <.001) with increasing age. Median household income was generally higher in the ≥60 age group (p-value of .038). Most co-morbidities were more prevalent in the older age groups including diabetes, chronic kidney disease (CKD)/end-stage renal disease (ESRD), congestive heart failure (CHF), solid tumors, lymphoma and liver disease (p-values of <.001). Obesity (21.0% vs. 20.0% vs. 11.3%, p-value of <.001) and autoimmune disease (12.4% vs. 11.6% vs. 8.5%, p-value of .017) were more prevalent in the younger age groups. Overall, a ≥2 days delay from admission to starting TPE was more prevalent in the older age groups (31.9% vs. 34.4% vs. 47.6%, p-value of <.001; Table 1A). When groups were compared separately in a post hoc analysis, there was statistical significance only when the ≥60 age group was compared to both the 18–39 (p-value of <.001) and the 40–59 (p-value of <.001) age groups, but not the latter two groups together (p-value of .224). Mean (in days) time from admission to TPE initiation was highest in the ≥60 age group (2.2 vs. 2.3 vs. 3.4, p-value of <.001). A multivariate analysis was done to detect the true effect of age on delay of treatment, while adjusting for sex, race, median household income, hospital status, weekend admission, diabetes, obesity, CHF, CKD/ESRD, autoimmune diseases, liver diseases, lymphoma, and solid tumors. When compared to the 18–39 age group, the ≥60 age group had a 1.65-fold odds of having a delay of ≥2 days from admission to starting TPE (adjusted odds ratio [OR], 1.649; 95% CI, 1.321–2.058, p-value of <.001; Table 1B). There was no statistically significant difference present when the 40–59 and the 18–39 age group were compared together (adjusted OR, 1.149; 95% CI, 0.947–1.395, p-value of .159). Other independent predictors of delay in TPE initiation are delineated in Figure S1. In general, there was a rising trend of in-hospital adverse events in older age groups including mortality (3.4% vs. 7.5% vs. 16.6%, p-value of <.001), mean (days) length of stay (13.5 vs. 14.7 vs. 17.0, p-value of <.001), VTE (4.5% vs. 5.7% vs. 7.6%, p-value of .016), ATE (7.3% vs. 19.6% vs. 23.6%, p-value of <.001), major bleeding (4.5% vs. 6.7% vs. 8.0%, p-value of .007), and mechanical ventilation (9.7% vs. 15.3% vs. 22.2%, p-value of <.001; Table 1). On adjusted analysis, the ≥60 age group was associated with higher mortality (adjusted OR, 3.202; 95% CI, 2.087–4.912, p-value of <.001), longer length of stay (>75th percentile [18 days]; adjusted OR, 1.289; 95% CI, 1.002–1.659, p-value of .048), and higher prevalence of VTE (adjusted OR, 1.708; 95% CI, 1.046–2.789, p-value of .032), ATE (adjusted OR, 3.390; 95% CI, 2.461–4.670, p-value of <.001), major bleeding (adjusted OR, 1.706; 95% CI, 1.084–2.684, p-value of .021), and mechanical ventilation (adjusted OR, 2.103; 95% CI, 1.546–2.860, p-value of <.001), when compared to the 18–39 age group. A ≥2 days delay from admission to starting TPE was independently associated with higher in-hospital mortality (adjusted OR, 1.615; 95% CI, 1.220–2.138, p-value of .001; Figure 1). In this nationally representative study, we found that elderly TTP patients (≥ 60) were independently more likely to experience a delay in time to initiation of TPE and had higher associated mortality as well as length of hospital stay, VTE, ATE, major bleeding, and need for mechanical ventilation. A few studies have sought to demonstrate the relationship between increasing age and worsened outcomes in TTP but, to our knowledge, our study is the largest and most recent study to directly evaluate this relationship as well as the relationship between delay in initiation of TPE and mortality.4-6 Only the study by Prevel et al. has commented on the direct effect of age on the delay from admission to diagnosis and treatment, albeit with a smaller sample size (71 in the ≥60 age group). They found a delay from admission to diagnosis of 3 days in the ≥60 age group compared to 1 day in the <60 age group (p-value of .001), with the main reason being atypical neurological manifestations.2 Patients ≥65 years are more likely to present with neurological and renal involvement and are less likely to have bleeding or anemia compared to younger patients. They also have atypical neurological features at diagnosis and frequently present with delirium and behavioral disturbance in contrast to younger patients. In fact, the well-validated PLASMIC score was found to be less sensitive in the elderly population.3 Older patients are less likely to exhibit significant cytopenias, more likely to have more severe renal impairment (creatinine >2 mg/dL), and more likely to have co-morbid malignancies; all are factors that would alter the PLASMIC scores and limit its predictive value. Furthermore, TTP inherently shares some clinical features with other conditions and hence, misdiagnosis is common. The above factors, coupled with the rarity of TTP, may lead to a low suspicion from providers and a delay in diagnosis and initiation of treatment. TTP is an extremely rare disease and therefore retrospective studies represent the most feasible way to assess outcomes. However, with retrospective studies come certain limitations including misclassification bias as we could not definitively confirm TTP diagnosis with no access to laboratory data such as ADAMTS13 levels. Other limitations include unaccounted confounders, lack of data on other therapies, and mortality bias where patients who died early may not have been included in the study due to death before diagnosis of TTP. In summary, this study sheds light on the clinical challenges in TTP management in the elderly population. Older patients with TTP have a higher likelihood of treatment delay and have worse in-hospital outcomes. Treatment delay was also independently associated with higher mortality and may mitigate the benefit of new emerging therapies like caplacizumab. Clinicians should be aware of the peculiar nature of TTP in the elderly population to increase suspicion of the disease and shorten time to diagnosis and treatment initiation. Future research is warranted to identify sensitive diagnostic criteria and specialized treatment plans for elderly TTP patients. No third-party financial funding or materials were accepted or necessary for the execution of this research project. The authors declare no conflicts of interest. The data used in this study are available in the national inpatient sample database on request from www.hcup-us.ahrq.gov. All data generated or analyzed during this study are included in this published article. Data S1. Supporting Information Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts