was first reported as a complication of chronic HIV infection in 1985 (4
). Since that time, approximately 300 cases of opportunistic nocardiosis in the setting of HIV infection have been described worldwide. We were able to identify 161 cases published in peer-reviewed journals before the era of highly active antiretroviral therapy (HAART), 115 cases in the post-HAART era and 60 cases at the intersection of these two eras.
The prevalence of Nocardia
in persons with AIDS in North America was 0.3% during the early 1980s (4
). A Spanish series (5
) described a similar prevalence (0.38%) in persons with AIDS between 1981 and 2000.
This likely represents an underestimate due to challenges in microbiological identification and under-reporting of Nocardia
isolation in respiratory secretions is hampered by its slow growth. Without selective media, it will usually be obscured by other more rapidly growing flora. It is also destroyed by the methods traditionally used to decontaminate sputum before mycobacteriological culture (3
The prevalence of Nocardia
infection is known to undergo seasonal and geographical variation (6
). Miami has the highest reported prevalence of Nocardia
and AIDS (1.8%) in North America – 30 cases were identified at Jackson Memorial Hospital (USA) between 1985 and 1989 (7
). A recent study (8
) from the same centre between 1999 and 2004 identified 19 more cases, with a prevalence of 0.1% in persons with AIDS. This provides further support for the observation that Nocardia,
similar to other opportunistic infections in AIDS, is declining in the developed world since the introduction of HAART.
complicating AIDS appears to be most common in the developing world, where it has a prevalence of approximately 4% (3
The predominant route of inoculation by Nocardia
is by inhalation, or less commonly by direct infection of soft tissue through traumatic injury. Reflecting these modes of acquisition, Nocardia
usually presents in patients with HIV as a chronic pneumonia syndrome (75% to 90%) or soft tissue infection (5% to 15%) (3
). The presentation of Nocardia
in HIV-infected patients is similar to patients without HIV, but it is more likely to be associated with extrapulmonary dissemination (7
). If defined as an established infection occurring at a noncutaneous extrapulmonary site, disseminated disease was seen in 84% of persons in the pre-HAART era, but appears to be less prevalent in more recent series (13
). Common sites of extrapulmonary spread include central nervous system, bone and joint, liver and pericardium.
Presenting symptoms are nonspecific and usually consist of cough and shortness of breath for several weeks duration, usually associated with fever and progressive weight loss. Cutaneous abscess formation or pericarditis appears to be more common in HIV patients with active injection drug use, suggesting that direct inoculation may be a significant risk factor in this population (6
The radiographic appearance of pulmonary nocardiosis is nonspecific, and most commonly presents with upper lobe consolidation or reticulonodular infiltrates (14
). Progression toward cavitation is more common in patients with HIV infection, and is seen in 60% to 80% of patients (7
infection in the HIV population tends to be associated with advanced immunodeficiency. The CD4 cell count is less than 200 cells/μL in the vast majority of patients and less than 50 cells/μL in 50% to 85% (8
appears to be more common in patients not receiving active treatment for their HIV, and in one recent series (8
), 37% of patients with Nocardia
infection presented with previously undiagnosed HIV infection. Intravenous drug use also appears to be a risk factor for nocardiosis in the developed world (6
infection in patients with HIV is often complicated by other coinfections such as Mycobacterium tuberculosis
), Mycobacterium avium
), P aeruginosa
) and P jiroveci
First-line therapy for all strains of Nocardia
infection has traditionally been sulfonamides; however, cotrimoxazole therapy has been found to be more effective than sulfonamide monotherapy (22
). This complicates treatment in the HIV-positive population, in which the prevalence of hypersensitivity or intolerance to cotrimoxazole ranges from 20% to 50%. (13
). A dose of 15 mg/kg/day (trimethoprim component) is suggested for the immunocompromised host, often in combination with a second active antimicrobial agent. Second-line agents against Nocardia
species include ceftriaxone, minocycline, imipenem and amikacin; sensitivity is seen in greater than 70% of isolates of N asteroides
, Nocardia nova
and Nocardia brasiliensis
). N farcinica
is typically more drug-resistant, but usually retains sensitivity to amikacin, ciprofloxacin and imipenem. (3
). An alternative to cotrimoxazole must be utilized in the setting of sulfa resistance (common with N otitidiscaviarum
), sulfa intolerance and for those patients failing sulfonamide therapy. The combination of imipenem and amikacin has demonstrated good activity clinically, as well as in animal models and in vitro (12
). Parenteral therapy can generally be stepped down to oral therapy after three to six weeks, once a clinical response has been achieved (25
). The best-studied oral alternatives to cotrimoxazole are minocycline and amoxicillin-clavulanate (3
). Long-term oral therapy should be guided by susceptibility data.
Debate remains in the literature about an appropriate length of treatment for nocardiosis in the setting of HIV infection. While 12 weeks of therapy appears to be adequate for minor cutaneous infections, most authors recommend at least one year of therapy for a primary course of treatment for patients with pulmonary or disseminated disease (6
). Relapse is common in patients not receiving antiretroviral therapy. Uttamchandani et al (7
) found that 22 of 27 AIDS patients showed clinical improvement on antinocardial therapy; however, 64% of patients had relapsing infections once antibiotics were stopped. Relapse was associated with shorter treatment duration and a mortality of approximately 80%. This has prompted the recommendation that low-dose maintenance therapy should be considered indefinitely (13
). Our case suggests that this may not be necessary in the HAART era; there is currently no literature indicating at what point maintenance therapy can be safely discontinued.
is associated with a high mortality in the HIV-positive population. Poor outcome is associated with disseminated infection, delayed diagnosis and early discontinuation of treatment (7
). A retrospective study (27
) of the English-language literature in the pre-HARRT era highlighted a 74% mortality. This seems high in light of current, improved antiretroviral therapies. A more recent review (28
) cited a 33% mortality in Nocardia
-infected HIV-positive patients.
Our patient did not receive first-line antimicrobial therapy due to a delay in laboratory reporting. His dose of cotrimoxazole, while suitable for P jiroveci
prophylaxis, would have been subtherapeutic for N farcinica
infection. While cotrimoxazole prophylaxis probably does reduce the incidence of nocardiosis in AIDS, there have been numerous reports of Nocardia
infection developing while on prophylaxis with this agent (3
). His response could have been due, in part, to the 28-day parenteral course of antibiotics he received for P aeruginosa
. Susceptibility testing showed sensitivity to beta-lactam agents. His isolate of N farcinica
showed only intermediate susceptibility to the macrolide clarithromycin. Thus, his oral course of azithromycin and rifabutin would have been of limited benefit. Rapid institution of antiretroviral therapy and subsequent immune reconstitution probably played a pivotal role in his rapid recovery, particularly in light of the grim prognosis of this infection seen in the pre-HAART era. There are no other reports in the literature of such a dramatic clinical resolution of Nocardia
pneumonia with antiretroviral therapy.
Although it is a recognized opportunistic pathogen, infection with Nocardia
species is not considered to be an AIDS-defining illness due to its rarity, and the fact that a substantial proportion of Nocardia
infections are not associated with any known underlying immune deficit (4
). Our patient was diagnosed with AIDS based on AIDS-wasting syndrome.
The surprising rarity of Nocardia infection in patients with AIDS is likely to be multifactorial. Nocardia is probably underdiagnosed due to its nonspecific clinical and radiological presentation, the frequency of other coinfections and difficult laboratory isolation. Because it is a soil saprophyte, the degree of environmental exposure is probably a significant determinant of its relative incidence. Finally, the frequent use of cotrimoxazole prophylaxis and effective antiretroviral therapies in Europe and North America will potentially contribute to its ongoing rarity in those populations infected with HIV. For these same reasons, its relatively high incidence in some parts of Sub-Saharan Africa may continue. Nocardia infection is frequently overlooked as a potential pathogen and should always be part of the differential diagnosis in HIV patients with advanced CD4 depletion and cavitary pneumonia.