Pulmonary embolism in chronic kidney disease: a lethal, overlooked and research orphan disease

Authors

  • C. ZOCCALI,

    1. Renal and Transplantation Unit, Clinical Epidemiology and Physiopathology of Renal Disease and Hypertension, CNR National Research Council (Italy), Reggio Calabria, Italy
    Search for more papers by this author
  • F. MALLAMACI

    1. Renal and Transplantation Unit, Clinical Epidemiology and Physiopathology of Renal Disease and Hypertension, CNR National Research Council (Italy), Reggio Calabria, Italy
    Search for more papers by this author

Carmine Zoccali, Nefrologia e CNR, Ospedali Riuniti, Reggio Calabria 89124, Italy.
Tel.: 39 0965 397010; fax: 39 0965 26879.
E-mail: carmine.zoccali@tin.it

Abstract

See also Ocak G, van Stralen KJ, Rosendaal FR, Verduijn M, Ravani P, Palsson R, Leivestad T, Hoitsma AJ, Ferrer-Alamar M, Finne P, De Meester J, Wanner C, Dekker FW, Jager KJ. Mortality due to pulmonary embolism, myocardial infarction, and stroke among incident dialysis patients. This issue, pp 2484–93.

Deep vein thrombosis (DVT) and pulmonary embolism (PE) are the two components of venous thromboembolism (VTE), a major, growing public health issue [1,2]. PE is the death-trigger in patients with VTE and the most frequent cause of preventable hospital death in the USA [1,2]. In an analysis by the CDC focusing on 2007–2009 data from the National Hospital Discharge Survey [3] about half a million hospitalizations with VTE per year were registered. VTE was equally common in both genders and the VTE hospitalization rate was at least three times more frequent in individuals older than 60 years than in younger age-classes. Overall the VTE-attributable mortality rate amounted to 28 726 deaths per year [3].

Even though DVT and PE events are preventable by appropriate pharmaco-prophylaxis and/or mechanical devices, prophylaxis of VTE is often not applied in patients at high risk for this complication [1,4]. For this reason major efforts are being made by the CDC and other public health agencies [3] to promote implementation of evidence-based guidelines for DVT and PE prevention. From an etiological point of view, VTE is seen as a multifactorial condition secondary to genetic and environmental risk factors, some of which are transient, such as the use of contraceptives or hormone therapy, pregnancy, hospitalization, surgery, trauma or immobilization. Anti-phospholipid antibodies, obesity, cancer and chronic diseases, including hypertension, chronic heart diseases and autoimmune diseases, as well as heavy smoking are all risk factors of paramount importance [5]. In about a half of cases no known environmental risk factor is present and in about 1/5 neither environmental nor genetic risk factors are identified in affected patients, emphasizing incomplete knowledge of the etiology of this condition [1,6]. Hospitalization is of paramount importance for PE because it leads to exposure to several risk factors, including intravenous catheters, surgery and immobilization. Genetic risk factors encompass deficiency of anticoagulants protein C, protein S and antithrombin, factor V Leiden and prothrombin G20210A gene polymorphisms. Mutated factor V and prothrombin genes are rare (5% or less), but such mutations are clinically relevant because they increase the risk of VTE in heterozygous individuals from 3 to 10 times [reviewed in Ref. 7]. Concurrence of gene–gene and gene– environment interactions increases the risk of VTE in affected individuals. Thus women with factor V Leiden on oral contraceptives have a 5-fold higher risk of VTE than women with neither risk factor.

Hemostasis and pulmonary embolism in end-stage kidney disease

Patients with end-stage kidney disease (ESKD) maintained on chronic dialysis have complex alterations in the hemostatic system and may develop both thrombotic and bleeding complications. Decreased sensitivity to agonists, altered adhesiveness [8], low glycoprotein I expression [9], low procoagulant activity [10], reduced thromboxane A2 synthesis [11], disturbed ability to retract fibrin clot [12] and high cyclical adenosine monophosphate (camp) [13] in platelets of patients with ESKD all combine to prolong bleeding time and impair the ability to form effective clots in these patients. On the other hand serum fibrinogen is substantially increased [14] and plasma levels of homocysteine, a factor favoring vascular thrombosis, are more than doubled in this population [15] and both these risk factors predict a high risk of thrombotic events in ESKD patients on dialysis [14,15], a phenomenon also attributable to concomitant, severe endothelial dysfunction and inflammation. The balance between the thrombotic risk and the bleeding risk in ESKD patients is still unclear and there is substantial case to case variability for risk factors impinging upon bleeding and thrombosis among these patients. Furthermore, this balance may be unpredictably altered by superimposed conditions such as sepsis or liver disease or by the use of the several drugs that interfere with vitamin K metabolism, from antibiotics to antiepileptic medications.

In the absence of solid data on the risk of PE, until a decade ago the prevailing view was that the bleeding propensity may protect dialysis patients from this complication [16]. However, in 2002 a study based on the United States Renal Data System registry showed that it was exactly the opposite [17]. Indeed, in a cohort of 76 718 patients starting dialysis in 1996, the incident rate of PE (150 cases per 100 000 persons × years) was 2.34 times higher than that in the parent general population. In this study older dialysis patients had a higher incidence of PE than young patients, but when compared with coeval cohorts in the general population, the risk excess for PE was higher in the young than in the old dialysis patients. Furthermore, in analyses excluding patients without known risk factors for VTE, including heart failure, ischemic heart disease and myocardial infarction, arrhythmia, stroke, peripheral vascular disease, tobacco use, cancer, inability to ambulate, or use of erythropoietin, the PE risk remained substantially unmodified. Thus, ESKD per se rather than the associated co-morbidities seems to be the main driver of PE in dialysis patients. This seminal USRDS-based study, the first to focus on the epidemiology of this complication in a sizeable cohort, did not provide information on the risk of death attributable to PE in ESKD, which is a problem of obvious relevance in this population.

The risk of death due to pulmonary embolism in ESKD

In this issue of the Journal, Gurbey Ocak et al. [18] report observations in an even larger cohort [over 130 000 hemodialysis patients included in the European Renal Association, European Dialysis and Transplant (ERA-EDTA) Registry] than the USRDS-based study. Herein these authors nicely show that the proportion of deaths attributable to PE among dialysis patients (0.7% over 2 years, or 140 deaths per 100 000 persons × years) is only marginally higher than that in the background general population (0.5%) but entails a 12.2 times higher death risk of death. Of note, this risk excess was of a magnitude comparable to that portended by myocardial infarction and stroke (11.0 and 8.4 times higher in dialysis patients, respectively). Even though the incidence rate of PE in the ERA-EDTA registry was not reported, it should be noted that the mortality rate due to PE in this Registry (140 cases per 100 000 persons × years) closely approximates the incidence rate of PE in the USRDS (150 per 100 000 persons × years), suggesting that PE is fatal in the vast majority of cases. PE in the general population represents the third most common cause of vascular death after myocardial infarction and stroke. Therefore findings by Ocak et al. are a strong call to action for nephrologists.

Specific research in ESKD patients focusing on this largely overlooked complication is a priority. Clinicians often face quite difficult therapeutic decisions in these high-risk patients and issues related to thromboprophylaxis are among the most challenging ones. Warfarin increases the risk of hemorrhagic stroke [19], accelerates cardiovascular calcification [20] and may aggravate bone disease [21] in ESKD patients. The safety of low-molecular-weight heparins as compared with unfractionated heparins still remains unknown in this population [22]. Indirect Xa inhibitors, such as fondaparinux, have been only sparsely investigated in ESKD [23] and no solid recommendation on the use of these agents can be formulated. Direct thrombin inhibitors (such as dabigatran) and direct Xa inhibitors (such as as rivaroxoban and apixaban) are promising new drugs but we have virtually no data in patients with ESKD. By the same token there is no study on rivaroxaban and apixaban. Due to the age-related decline in renal function, dabigatran, a drug disposed by the renal route, may be unsafe in elderly patients [24], a population where estimates of the GFR based on serum creatinine may be misleading [25].

In theory high-risk dialysis patients would appear prima facie suitable candidates for treatments aimed at preventing thrombosis. Yet, the many uncertainties about the safety and effectiveness of current drugs prevent formulating strong recommendations for clinical practice in these patients. The elegant, ERA-EDTA registry-based study published in this issue of the Journal [18] extends findings in stage 4-5 CKD patients [26] to those undergoing dialysis. It is now undisputable that severe CKD is a strong risk factor for death in patients with PE. To gather evidence-based information about thromboprophylaxis in the CKD population, these patients can be actively recruited in current and future thromboprophylaxis trials. Furthermore, comparative effectiveness studies in large clinical databases represent an additional pragmatic solution for gathering sensible data on thrombosis prevention and treatment in this population. Hopefully these novel observations by Ocak et al. should serve as a further stimulus to spur on clinical research on thromboprophylaxis in ESKD, a truly research-orphan issue with major public health implications.

Disclosure of Conflicts of Interest

The authors state that they have no conflicts of interest.

Ancillary