Antje Giede-Jeppe1, Tobias Bobinger2, Stefan T Gerner2, Dominik Madžar2, Jochen Sembill2, Hannes Lücking2, Stephan P Kloska2, Toni Keil2, Joji B Kuramatsu2, Hagen B Huttner2. 1. From the Departments of Neurology (A.G.-J., T.B., S.T.G., D.M., J.S., J.B.K., H.B.H.), Neuroradiology (H.L., S.P.K.), and Anaesthesiology (T.K.), University of Erlangen-Nuremberg, Erlangen, Germany. Antje.Giede-Jeppe@uk-erlangen.de. 2. From the Departments of Neurology (A.G.-J., T.B., S.T.G., D.M., J.S., J.B.K., H.B.H.), Neuroradiology (H.L., S.P.K.), and Anaesthesiology (T.K.), University of Erlangen-Nuremberg, Erlangen, Germany.
Abstract
BACKGROUND AND PURPOSE: Stroke-associated immunosuppression is an increasingly recognized factor triggering infections and thus potentially influencing outcome after stroke. Specifically, lymphocytopenia after intracerebral hemorrhage (ICH) has only been addressed in small-sized retrospective studies of mixed intracranial bleedings. This cohort study investigated the natural course of lymphocytopenia, parameters associated with lymphocytopenia on admission (LOA) and during stay, and evaluated the clinical impact of lymphocytopenia in solely ICH patients. METHODS: This observational study included 855 consecutive patients with ICH. Patient demographics, clinical and neuroradiological data as well as laboratory and in-hospital measures were retrieved from institutional prospective databases. Functional 3-month outcome was assessed by mailed questionnaires. Lymphocytopenia was defined as <1.0 (10(9)/L) and was correlated with patient's characteristics and outcome. RESULTS: Prevalence of LOA was 27.3%. Patients with LOA showed significant associations with poorer neurological status (18 [10-32] versus 13 [5-24]; P<0.001), larger hematoma volume (18.5 [6.2-46.2] versus 12.8 [4.4-37.8]; P=0.006), and unfavorable outcome (74.7% versus 63.3%; P=0.0018). Natural course of lymphocyte count during hospital stay revealed a lymphocyte nadir of 1.1 (0.80-1.53 [10(9)/L]) at day 5. Focusing on patients with day-5-lymphocytopenia, compared with patients with LOA, revealed increased rates of infections (63 [71.6] versus 113 [48.5]; P<0.001) and poorer functional outcome at 3 months (76 [86.4] versus 175 [75.1); P=0.029). Adjusting for baseline confounders, multivariable logistic and receiver operating characteristics analyses documented independent associations of day-5-lymphocytopenia with unfavorable outcome (day-5-lymphocytopenia: odds ratio, 2.017 [95% confidence interval, 1.029-3.955], P=0.041; LOA: odds ratio, 1.391 [0.795-2.432], P=0.248; receiver operating characteristics: day-5-lymphocytopenia: area under the curve=0.673, P<0.0001, Youden's index=0.290; LOA: area under the curve=0.513, P=0.676, Youden's index=0.084), whereas receiver operating characteristics analyses revealed no association of age or hematoma volume with day-5-lymphocytopenia (age: area under the curve=0.540, P=0.198, Youden's index=0.106; volume: area under the curve=0.550, P=0.0898, Youden's index=0.1224). CONCLUSIONS: Lymphocytopenia is frequently present in patients with ICH and may represent an independent parameter associated with unfavorable functional outcome. Developing lymphocytopenia affected outcome even stronger than LOA, a finding that may open up new therapeutic avenues in specific subsets of patients with ICH.
BACKGROUND AND PURPOSE:Stroke-associated immunosuppression is an increasingly recognized factor triggering infections and thus potentially influencing outcome after stroke. Specifically, lymphocytopenia after intracerebral hemorrhage (ICH) has only been addressed in small-sized retrospective studies of mixed intracranial bleedings. This cohort study investigated the natural course of lymphocytopenia, parameters associated with lymphocytopenia on admission (LOA) and during stay, and evaluated the clinical impact of lymphocytopenia in solely ICHpatients. METHODS: This observational study included 855 consecutive patients with ICH. Patient demographics, clinical and neuroradiological data as well as laboratory and in-hospital measures were retrieved from institutional prospective databases. Functional 3-month outcome was assessed by mailed questionnaires. Lymphocytopenia was defined as <1.0 (10(9)/L) and was correlated with patient's characteristics and outcome. RESULTS: Prevalence of LOA was 27.3%. Patients with LOA showed significant associations with poorer neurological status (18 [10-32] versus 13 [5-24]; P<0.001), larger hematoma volume (18.5 [6.2-46.2] versus 12.8 [4.4-37.8]; P=0.006), and unfavorable outcome (74.7% versus 63.3%; P=0.0018). Natural course of lymphocyte count during hospital stay revealed a lymphocyte nadir of 1.1 (0.80-1.53 [10(9)/L]) at day 5. Focusing on patients with day-5-lymphocytopenia, compared with patients with LOA, revealed increased rates of infections (63 [71.6] versus 113 [48.5]; P<0.001) and poorer functional outcome at 3 months (76 [86.4] versus 175 [75.1); P=0.029). Adjusting for baseline confounders, multivariable logistic and receiver operating characteristics analyses documented independent associations of day-5-lymphocytopenia with unfavorable outcome (day-5-lymphocytopenia: odds ratio, 2.017 [95% confidence interval, 1.029-3.955], P=0.041; LOA: odds ratio, 1.391 [0.795-2.432], P=0.248; receiver operating characteristics: day-5-lymphocytopenia: area under the curve=0.673, P<0.0001, Youden's index=0.290; LOA: area under the curve=0.513, P=0.676, Youden's index=0.084), whereas receiver operating characteristics analyses revealed no association of age or hematoma volume with day-5-lymphocytopenia (age: area under the curve=0.540, P=0.198, Youden's index=0.106; volume: area under the curve=0.550, P=0.0898, Youden's index=0.1224). CONCLUSIONS:Lymphocytopenia is frequently present in patients with ICH and may represent an independent parameter associated with unfavorable functional outcome. Developing lymphocytopenia affected outcome even stronger than LOA, a finding that may open up new therapeutic avenues in specific subsets of patients with ICH.
Authors: Andrea Morotti; Sandro Marini; Michael J Jessel; Kristin Schwab; Christina Kourkoulis; Alison M Ayres; M Edip Gurol; Anand Viswanathan; Steven M Greenberg; Christopher D Anderson; Joshua N Goldstein; Jonathan Rosand Journal: Neurocrit Care Date: 2017-04 Impact factor: 3.210