Ludovic Nguyen-Them1, Myrto Costopoulos2, Marie-Laure Tanguy3, Caroline Houillier1, Sylvain Choquet4, Hind Benanni5, Rwaida Elias-Shamieh6, Marine Armand7, Geraldine Faivre1, Sylvie Glaisner6, Sandra Malak6, Jacques Vargaftig6, Khê Hoang-Xuan8, Guido Ahle9, Valérie Touitou10, Nathalie Cassoux11, Frédéric Davi7, Hélène Merle-Béral7, Magali Le Garff-Tavernier7, Carole Soussain12. 1. Groupe Hospitalier Pitié-Salpétrière, Neuro-oncology, Paris, France. 2. Groupe Hospitalier Pitié-Salpétrière, Biological Hematology, Paris, France; Paris University Sorbonne UPMC, INSERM UMRS 1138, France. 3. Groupe Hospitalier Pitié-Salpétrière, Biostatistics, Paris, France. 4. Groupe Hospitalier Pitié-Salpétrière, Hematology, Paris, France. 5. Hôpital René Huguenin, Institut Curie, Biological Hematology, Saint-Cloud, France. 6. Hôpital René Huguenin, Institut Curie, Hematology, Saint-Cloud, France. 7. Groupe Hospitalier Pitié-Salpétrière, Biological Hematology, Paris, France. 8. Groupe Hospitalier Pitié-Salpétrière, Neuro-oncology, Paris, France; Paris University Sorbonne UPMC, INSERM U1127, CNRS UMR 7225, IHU, ICM, France. 9. Hôpitaux Civils de Colmar, Neurology, Colmar, France. 10. Groupe Hospitalier Pitié-Salpétrière, Ophthalmology, Paris, France. 11. Hôpital Claudius Regaud, Institut Curie, Ophthalmology, Paris, France. 12. Hôpital René Huguenin, Institut Curie, Hematology, Saint-Cloud, France; Collège de France, INSERM U1050, CNRS UMR 7241, Paris, France. Electronic address: carole.soussain@curie.fr.
Abstract
INTRODUCTION: We aimed to confirm the diagnostic value and to evaluate the pre- and post-therapeutic prognostic value of cerebrospinal fluid (CSF) concentrations of interleukin (IL)-10 and IL-6 in patients with diffuse large B-cell primary central nervous system lymphoma (PCNSL). PATIENTS AND METHODS: IL-10 and IL-6 concentrations were measured in 79 patients with PCNSL at diagnosis and in 40 control individuals. Fifty-four PCNSL patients underwent repeat assessments starting at diagnosis. RESULTS: The IL-10 concentration distinguished PCNSL from other neurologic diseases with a sensitivity of 88.6% and a specificity of 88.9% with a cutoff of 4 pg/ml. In a multivariate analysis of PCNSL patients, CSF involvement was associated with a higher IL-10 concentration (mean log (IL-10) of 4.4 versus 2.5 pg/ml, respectively, p = 0.0004). The pre-therapeutic IL-10 concentration had no prognostic impact on outcome. The IL-10 concentration decreased after treatment for most patients tested. Among patients with complete remission or partial remission, as evaluated by magnetic resonance imaging (MRI), a persistent detectable IL-10 level in the CSF at the end of treatment was associated with a negative impact on progression-free survival (PFS) (1-year PFS: 15%, 95% confidence interval [CI]: 2.5-38% versus 59%, 95% CI: 32-78%, respectively, p = 0.0004). CONCLUSION: Our study confirmed that IL-10 is a useful biomarker for the diagnosis of PCNSL. We highlight new findings showing that the IL-10 level in the CSF could be used as a surrogate marker for CSF involvement and that the post-treatment IL-10 concentration could complement standard MRI for therapeutic response assessment in PCNSL.
INTRODUCTION: We aimed to confirm the diagnostic value and to evaluate the pre- and post-therapeutic prognostic value of cerebrospinal fluid (CSF) concentrations of interleukin (IL)-10 and IL-6 in patients with diffuse large B-cell primary central nervous system lymphoma (PCNSL). PATIENTS AND METHODS: IL-10 and IL-6 concentrations were measured in 79 patients with PCNSL at diagnosis and in 40 control individuals. Fifty-four PCNSLpatients underwent repeat assessments starting at diagnosis. RESULTS: The IL-10 concentration distinguished PCNSL from other neurologic diseases with a sensitivity of 88.6% and a specificity of 88.9% with a cutoff of 4 pg/ml. In a multivariate analysis of PCNSLpatients, CSF involvement was associated with a higher IL-10 concentration (mean log (IL-10) of 4.4 versus 2.5 pg/ml, respectively, p = 0.0004). The pre-therapeutic IL-10 concentration had no prognostic impact on outcome. The IL-10 concentration decreased after treatment for most patients tested. Among patients with complete remission or partial remission, as evaluated by magnetic resonance imaging (MRI), a persistent detectable IL-10 level in the CSF at the end of treatment was associated with a negative impact on progression-free survival (PFS) (1-year PFS: 15%, 95% confidence interval [CI]: 2.5-38% versus 59%, 95% CI: 32-78%, respectively, p = 0.0004). CONCLUSION: Our study confirmed that IL-10 is a useful biomarker for the diagnosis of PCNSL. We highlight new findings showing that the IL-10 level in the CSF could be used as a surrogate marker for CSF involvement and that the post-treatment IL-10 concentration could complement standard MRI for therapeutic response assessment in PCNSL.