The important work of Gilks suggests that the epidemiology and clinical spectrum of pneumococcal disease in HIV-infected patients is strikingly similar in the United States and Kenya. In both locations, HIV-infected patients comprise a significant proportion of patients with pneumococcal bacteremia, and Streptococcus pneumoniae is among the most common causes of bacteremia and pneumonia in patients with HIV. Moreover, rates of recurrent disease (relapse or reinfection) may be high, whereas mortality is often lower than among bacteremic, HIV-seronegative patients. Delineation of associated risk factors in Africa such as sickle cell disease, malnutrition, or malaria would be of interest. More information from developing countries is needed about the clinical presentation, etiology, and response to therapy for pneumonia in young children infected with HIV. Studies of childhood pneumonia should be conducted in settings of high perinatal HIV seroprevalence to ensure that global strategies for control of pneumonia are adapted according to the rising prevalence of HIV infection. As Gilks notes, invasive pneumococcal disease is one of the few serious HIV-related infections for which therapy is available, affordable, and effective if the infection is treated quickly. Prevention of pneumococcal disease with vaccine in children and adults both with and without HIV infection remains an elusive and critical goal worldwide [1].

Klenerman and colleagues provide intriguing preliminary data on the impact of advanced HIV disease on the acute inflammatory response to bacterial infections. C-reactive protein is an evolutionarily conserved acute-phase reactant that binds to C polysaccharide in pneumococcal cell walls and activates complement binding to and phagocytosis of the organism. Hepatic synthesis of CRP is regulated by cytokines, such as interleukin-6; cytokine production may be abnormal in patients with HIV infection [2]. A consistent inability to produce CRP in response to acute bacteremia would be a distinctive sequela of advanced HIV disease. Larger studies are needed to determine whether the acute phase response to S. pneumoniae in bacteremic, HIV-infected patients is blunted or normal, as is suggested by the presence of fever and relatively low rates of mortality in most bacteremic patients [3]. In this setting, S. pneumoniae serves as a clinically relevant probe of the complex effects of HIV on nonimmune and immune host defenses to antigenic challenge.