Bacteria are causing approximately 90 % of infections during the early phase of neutropenia [8]. The radiological appearance of bacterial pneumonia includes consolidation, especially bronchopneumonia and positive pneumobronchogram (Fig. 7.2) [33, 34]. In contrast to immunocompetent patients, ground-glass opacification is found more often and remains nonspecific.

Severe neutropenia lasting for more than 10–14 days is associated with an increasing risk of invasive fungal infection [3], with Aspergillus species being the primary pathogen, while Candida species very rarely cause primary pneumonia (Fig. 7.8) [5]. Typical radiological findings of fungal and non-fungal pneumonia as well as of infiltrates from noninfectious diseases have been reviewed in detail [35]. The typical appearance of pulmonary infiltrates from fungal origin are as follows:

For use in the context of clinical and epidemiological research in neutropenic patients, standards for the interpretation of radiological findings in invasive fungal infections have been elaborated [10, 51]; newly emerged “typical” CT patterns (dense, well-circumscribed lesions with or without a halo sign, air-crescent sign) are classified as a clinical criterion for fungal pneumonia Figs. 7.5 and 7.8. The halo sign, first described in 1984 [49, 50], is nonspecific [33] and not a necessary part of the updated definitions [51]. A nonspecific infiltrate, rated as a minor criterion in the first version [10], was abandoned in the update [51]. In a later workup of a pharmaceutical trial investigating response rates to antifungal treatment, the evidence of the halo sign was associated with an improved response rate (52 % vs. 29 %; p < 0.001), as well as a higher 3-month survival rate (71 % vs. 53 %; p < 0.01) [52]. This large (n = 235) antemortem trial suffers, however, from systemic limitations like investigation of halo which was part of inclusion criteria and technical insufficiencies like usage of thick-section CT instead of appropriate thin-section CT and evaluation of hardcopies instead of monitor reading [52]. Histopathological workup of lung biopsies verified fungal pneumonia in 56 % of cases in another study [37]. Relevant differential diagnoses for the halo sign, such as cryptogenic organizing pneumonia (COP, formerly known as bronchiolitis obliterans organizing pneumonia, BOOP), pulmonary hemorrhage, pulmonary manifestation of the underlying malignancy, lung cancer, and non-fungal infections (CMV, tuberculosis, abscesses (Fig. 7.7), etc.) or Candida (Fig. 7.8), have to be considered [37].

Thus, diagnostic clarification will frequently be necessary, particularly when antifungal treatment is not successful.

Air-crescent sign and cavitation occur with hematopoietic reconstitution during the late phase of fungal infection (Fig. 7.8) [36]. Both radiological signs are known to be associated with a favorable prognosis. However, the specificities of these findings are limited, and relevant differential diagnoses have to be considered (Fig. 7.7) [37]. There are other useful patterns for the identification of fungal pneumonia, e.g., distribution along the bronchovascular bundle resulting in the feeding vessel sign with an angiotrophic location.

The ongoing development of antifungal therapy may have an important impact on the radiological appearance of fungal pneumonia. Thus, in the near future radiologists will not only be confronted with the question for “breakthrough fungal pneumonia,” but also for fungal pneumonia caused by non-Aspergillus pathogens.

Pneumocystis jiroveci pneumonia (PcP) [38] is a typical finding in hematological patients affected by severe cellular (T-cell) immunosuppression and those with graft-versus-host disease after aSCT, if they are not protected by effective chemoprophylaxis [8]. Despite standard trimethoprim-sulfamethoxazole prophylaxis, 8 % of the patients develop PcP, while among patients without prophylaxis, the incidence may reach 29 % [8]. Up to 15 % of these patients will have a fatal outcome [8].

CT provides a valuable characterization for PcP [6, 7, 18, 33] and is a reliable method for discriminating it from other infectious processes [33, 39]. A combination of ground-glass opacities and intralobular septae sparing out the subpleural space (i.e., perihilar distribution) is very typical for PcP (Fig. 7.9) [33, 39, 40].

Tuberculosis (TB) has to be considered as a rare but relevant differential diagnosis. In immunocompromised hosts, TB appears different compared to immunocompetent hosts (e.g., gangliopulmonary (primary) forms) [41]. More widespread lymphogenic and hematogenous dissemination can occur, and therefore, the clinical course might be fulminant [41, 42]. On the other hand, TB might mimic or come along with other infections like pulmonary aspergillosis or systemic candidiasis [42].