Is implantable loop recorder the answer to reduce the increased risk of stroke in cancer patients?

Abstract

Introduction

Recent epidemiologic data suggests increased risk of ischaemic stroke in cancer patients. The etiology of increased ischaemic stroke is unknown. Atrial fibrillation (AF) is among the potential etiologies. The risk of AF has not been studied among cancer patients in the United States.

Purpose

Ascertain the association of AF in cancer patients in the USA by using the largest database i.e. National Inpatient Sample (NIS).

Methods

Patients ≥18 years old were selected in the NIS database for years 2010 to 2014 and stratified based on presence or absence of any of four cancers (lung, colon, breast and prostate; 4CA) using ICD 9 codes. Atrial fibrillation and stroke/TIA were also identified using ICD 9 codes. Components of CHADS2 score (CHF, hypertension, Age>75, diabetes and stroke/TIA) were identified using ICD 9 codes. χ² tests performed for prevalence of AF in patients with or without these cancers stratified by CHADS2 score. Binary logistic regression was used to analyze individual components of CHADS2 score.

Results

AF and stroke/TIA were significantly higher among 4CA than non-4CA group (18.7% vs 12.0%, P<0.001 and 5.4% vs 4.8%, P<0.001 respectively). AF prevalence increased with CHADS2 and was significantly higher in 4CA group with CHADS2 score 0 to 4 (Table 1 and Figure 1). Logistic regression for the outcome of AF showed “Age >75” OR (3.0), CHF (2.8), CVA (1.2), HTN (1.3) and DM (1.1).

Conclusion

This is the first study using a national database of USA patients to estimate prevalence of AF in cancer patients compared to non-cancer patients and reaffirms the higher burden of AF in cancer patients. Prevalence of both AF and stroke were greater in cancer patients when stratified by CHADS2 score. This may indicate not just an increased risk of AF but an increased risk of stroke/TIA for the same CHADS2 score. Stroke incidence was also higher in the 4CA group (5.4% vs. 4.8% P<0.001). Cancer patients with CHADS2 score >1 may benefit from screening with loop recorder to identify previously undetected AF and initiate anticoagulation therapy. Prospective longitudinal studies are needed to validate this retrospective study.

Table 1. Atrial fibrillation (AF) prevalence in patients with one of four cancers (4CA) and without (non-4CA) stratified by CHADS2 score

		CHADS2 score
		0	1	2	3	4	5	6
Non-4CA	n	11,639,659	7,385,013	6,435,610	2,504,497	858,328	222,838	31,716
	AF	269,671 (2.3%)	859,514 (11.6%)	1,305,558 (20.3%)	716,916 (28.6%)	254,796 (29.7%)	75,335 (33.8%)	14,743 (46.5%)
4CA	n	497,046	719,404	742,901	307,648	94,301	24,558	3,523
	AF	29,668 (6.0%)	105,887 (23.5%)	174,824 (23.5%)	95,057 (30.9%)	31,533 (33.4%)	8,504 (34.6%)	1,603 (45.5%)

Funding Acknowledgement

Type of funding source: None

Prevalence and Incidence of Atrial Fibrillation