The occurrence of anticoagulant bleeding for atrial fibrillation quadruple the risk of cancer at the hemorrhagic site. However, explorations are insufficient, depriving patients with an early diagnosis that would improve the prognosis.
In the event of atrial fibrillation (FA), the prevention of ischemic stroke constitutes a major therapeutic issue. The long -term prescribed anticoagulants in this indication nevertheless expose a significant hemorrhagic risk. The occurrence of a hemorrhagic accident under anticoagulant treatment should systematically trigger an in -depth etiological assessment. However, in clinical practice, this alarm signal is not always taken into account. The hemorrhagic event often tends to be trivialized and attributed exclusively to anticoagulant treatment, without search for an underlying pathology. This minimalist attitude is reinforced by the absence of specific recommendations to manage these complex situations, particularly in asymptomatic patients reluctant to additional, sometimes invasive explorations. A recent Canadian prospective cohort study highlights the dangerous limits of the expectant approach. Diagnostic abstention can be fraught with consequences for the patient, also exposing the practitioner to significant forensic risks.
More than 119,000 patients under anticoagulant and more than 26,000 hemorrhagic accidents
The cohort studied, made up between 2008 and 2022, has 119,480 patients (average age, 77.4 years; men: 52 %) who all started anticoagulant treatment such as AVK (Warfarin) or AOD (direct oral anticoagulants) due to FA. More than one in five participants (n = 26,037; 21.8 %) was the victim of a more or less severe hemorrhagic accident, in all cases perfectly documented by the attending physician or a hospital service. After 2 years monitoring, cancer was diagnosed in 5,800 patients (4.9 %). The data was processed using the proportional risk model of COX with multiple adjustments.
Hemorrhagic accidents have been associated with a risk of quadruple cancer, the corresponding Hazard Ratio (HR) being estimated at 4.0 (95 %CI, 3.8-4.3). This ripper proved to be dependent on the site of the malignant lesion: gastrointestinal tract (HR = 5.0), genitorerinary (HR = 5.0), respiratory (HR = 4.0), brain/meninges (HR = 1.8) and Nasopharynx (HR = 1.5).
A signal to take into account
The HRs were significantly higher for cancers concordant with the bleeding site (gastrointestinal, 15.4; genito-cross, 11.8; respiratory, 10,1). Cancers were diagnosed at an earlier stage after bleeding (27.6 % of stadium cancer after bleeding versus 31.3 % stadium 4 in the absence of bleeding; p = 0.029).
The endoscopic assessment justified in the month following a digestive hemorrhage was only practiced less than once in four, proof of a certain lightness in the management of a serious complication.
The occurrence of a hemorrhagic accident under anticoagulant in a patient with FA deserves to be taken seriously. This event can reveal latent cancer, notably digestive, genitarian or bronchopulmonary. Consequently, it justifies additional investigations oriented by the site of the hemorrhagic accident. The message is primarily intended for the doctor, but the patient must also be informed, so as not to neglect a symptom which, instead of being trivialized and put on the drug account, must be considered as a warning signal.