People who survive the initial hyperinflammatory "cytokine storm" of severe sepsis regain their blood pressure, but are at high risk for secondary infection and viral reactivation. Animal models strongly suggest sepsis-induced immune suppression occurring later in the course of sepsis is to blame, but that's never been proven in humans.
Jonathan Boomer, Kathleen To, Richard Hotchkiss et al harvested tissue from the lungs and spleens of 40 people dying with severe sepsis after ~8 days. They then established a control tissue bank consisting of lungs and spleens of people without severe sepsis (harvested during resections for cancer; transplant donors; trauma and/or brain death).
Results: On cytokine secretion assays, white blood cells from the spleen of severe sepsis patients secreted 5-10% of the amount of cytokines (TNF, INF-gamma, IL-6, IL-10, including both pro- and anti-inflammatory cytokines) compared to spleen WBCs from non-septic patients. The septic patients also had fewer white blood cells in their spleens.
Why does it matter? Authors argue that the continued emphasis on developing and deploying immune-suppressing therapies for sepsis may be counterproductive if physicians administer these agents throughout the disease history -- after the initial cytokine storm has ebbed and potential immune suppression has begun to flow.
Limitations: There was a 20-year age gap between the subjects with severe sepsis (age ~71) and the controls (~52), and the two groups were markedly different in many other ways (length of ICU stay, extent of pre-existing chronic illness, nutritional status), limiting the strength of the conclusions somewhat.
Boomer JS et al. Immunosuppression in Patients Who Die of Sepsis and Multiple Organ Failure. JAMA 2011;306(23):2594-2605.