Histologic, viral, and molecular correlates of dengue fever infection of the liver using highly sensitive immunohistochemistry.

The mechanism by which the virus associated with dengue fever can cause a fatal hepatitis is not well understood. The purpose of this study was to examine 9 cases of fatal dengue hemorrhagic fever-associated hepatitis, and to correlate the histologic findings with viral detection and cytokine response. The histologic changes were nonspecific and included massive hepatic necrosis and a pauci-cellular acute hepatitis. Viral cDNA detection by reverse transcriptase in situ polymerase chain reaction demonstrated that the fatal hepatitis was due to infection on average of >90% of hepatocytes and many Kupffer cells. Similar results were obtained using immunohistochemistry for viral protein using an automated highly sensitive system. Immunohistochemical analysis for tumor necrosis factor alpha, and interleukin-2, showed rare positive Kupffer cells. In comparison, fatal cases of hepatitis C associated liver failure demonstrated far fewer infected hepatocytes and a concomitant strong up-regulation of many cytokines, notably tumor necrosis factor alpha and interleukin-2. It is concluded that fatal dengue hemorrhagic fever is associated with acute, severe liver damage due primarily to massive direct infection of hepatocytes and Kupffer cells with minimal cytokine response. The infection can be readily detected in a few hours using an automated system that has a sensitivity equivalent to reverse transcriptase in situ polymerase chain reaction.

Histologic, infectious, and molecular correlates of idiopathic spontaneous abortion and perinatal mortality.

The purpose of this study was to analyze the placental and neonatal tissues in fatal cases for a wide variety of infectious agents and cytokine expression. Placentas and corresponding neonatal tissues in 21 consecutive cases of idiopathic spontaneous abortion or perinatal death, before or within 2 days of birth, were tested for an infectious agent. The controls included 10 consecutive cases of fetal and placental tissues from therapeutic abortions, 5 placentas from unremarkable childbirths, and 11 placentas from cases of spontaneous abortion or perinatal death of known cause (ruptured uterus, placenta abruption, prolapsed cord). An intrauterine infection was noted in 16 of 21 (76%) of the placentas associated with neonatal mortality; in each case, the same infectious agent was found in the neonatal tissues, primarily the spleen. The most common infectious agent was enterovirus/coxsackie virus (10 cases); the histologic findings in the placenta were nonspecific. There was strong expression of TNF-alpha in the placenta and spleen of each of the cases of intrauterine infection and in none of the 26 controls. It is concluded that in utero infection and the associated cytokine up-regulation are responsible for many cases of unexplained fetal and neonatal loss.