Candida pneumonia is one of the most challenging of all the Candida infections. Rules for a practical and accurate diagnostic approach are elusive. As Candida spp. are frequent asymptomatic colonizers of the upper respiratory tract, especially among hospitalized patients, the predictive value of sputum and even bronchoalveolar cultures for actual candidal lung invasion is very low. In addition, physicians would ideally like to distinguish hematogenous Candida pneumonia related to invasive candidiasis from isolated lung infection. However, in the clinical practice such discrimination is very difficult.

Nevertheless, in order to make this discussion meaningful, some standards must be applied. The following definitions have been proposed by different authors and we will use them here:

  • Primary Candida pneumonia refers to invasive infections limited to the lungs [989]. Aspiration pneumonia and bronchopulmonary pneumonia have been also used to imply the same process, but these terms stress the pathophysiologic mechanisms involved [1160, 1353, 1966]. This entity, as expected, is associated with conditions that favor aspiration (e.g., altered mental status) and colonization of the upper respiratory tract [1160, 1463].
  • Secondary Candida pneumonia refers to lung involvement due to hematogenous dissemination in the setting of one of the forms of invasive candidiasis. This entity, while often documented at autopsy [1946] is usually completely masked by the other manifestations of invasive candidiasis.


The true incidence of primary Candida pneumonia is unknown. However, two comprehensive studies looking at primary Candida pneumonia reported incidence of 0.2 and 0.4% in autopsied cancer patients [989, 1463]. Only 55 cases of unequivocal evidence of primary Candida pneumonia had been reported in the English literature up to 1993 [989]. Searching through the more recent literature, only one additional case was found [1205].

In contrast, the lung is always among the three organs most frequently involved in patients who die with invasive candidiasis [233, 266, 270, 991, 993, 1082, 1218, 1376, 1419].

Frequency Of At Least Some Degree Of Lung Involvement In Patients Dying With Invasive Candidiasis
Reference N Rank Frequency(%)
Bodey et al. [266] 265 1st 42
Hughes et al. [1082] 109 1st 81
Maksymiuk et al. [1419] 46 1st 50
Knox et al. [1218] 25 1st 72
Bernhardt et al. [233] 14 3rd 21
Gaines et al. [792] 42 2nd 29
Louria et al. [1376] 19 1st 47
Myerowitz et al. [1602] 39 1st 62
Parker et al. [1724] 25 2nd 48

Candida Pneumonia and Candida species

Based on the data from Masur et al. and Haron et al. [989, 1463], C. albicans explains between 40 to 70% of cases of primary pneumonia . C. tropicalis and C. parapsilosis are the most frequent non-albicans species implicated in cases of primary Candida pneumonia. The species distribution for secondary pneumonia will follow that of (A):candidemia.

Clinical Manifestations

Again, we refer to the series by Masur et al. and Haron et al. Patients with primary or aspiration pneumonia are usually severely ill, with multiple organ failure, and some degree of altered mental status [989, 1463]. Clinically, the most common symptoms are fever, tachypnea, dyspnea, and chest pain. Secondary Candida pneumonia will present as part of the complex of symptoms related to the concomitant episode of invasive candidiasis. Pulmonary involvement is usually inapparent.

Specific Diagnostic Strategies

Candidal pneumonia is exceedingly difficult to diagnose antemortem. Cultures from either the sputum or bronchoscopic samples (whether quantified or not) are both poor predictors of tissue invasion [653, 1230]. For example, Kontoyiannis et al. estimated the positive predictive value of a positive sputum and/or BAL culture at only 42% [1230]. The negative predictive value of a negative culture in this autopsy-controlled study was, however, 93%, suggesting that a negative culture should lessen the suspicion of pulmonary candidiasis.

There is no clear radiologic pattern that is indicative of either primary or secondary candidal pneumonia. Many of the patients will have normal chest radiographs [270, 1463]. When the film is abnormal, reported findings have most often been bilateral interstitial or alveolar patchy infiltrates [337, 989, 1160, 1463] However, cystic lesions [1160], cavitating masses [2384], and exudative pleural effusions [337] have also been described.


The diagnosis of Candida pneumonia can only be solidly established with a histopathologic sample. Expected findings on microscopic examination of cases of primary Candida pneumonia are:

  1. evidence of aspiration: presence of oropharyngeal elements like food particles and squamous cells lying freely within the bronchi lumen and the alveoli and
  2. evidence of Candida invasion into the bronchial wall, with or without detached respiratory epithelium.

Classic findings of secondary Candida pneumonia are:

  1. microabscesses with pseudohyphae penetrating blood vessels (capillaries, arterioles and small arteries), PLUS
  2. evidence of Candida pseudohyphae invading lung interstitium and airways.


There are no data on the use of any particular antifungal drug or any particular regimen for the treatment of primary or secondary Candida pneumonia. The concepts used for the treatment of invasive candidiasis are presumably valid for lung infections.