Pulmonary cryptococcosis mimicking neoplasm in terms of uptake PET/CT

1. Faculdade de Medicina da Universidade de São Paulo (FMUSP), São Paulo, SP, Brazil. Mailing address: Dra. Karina de Souza Giassi. Faculdade de Medicina da Universidade de São Paulo. Avenida Doutor Arnaldo, 455, Pacaembu. São Paulo, SP, Brazil, 01246-903. E-mail: ksgiassi@gmail.com. All 20 reports were reviewed and defined as follows: “not described”—when features of EFN were overlooked; “described but not suggested”—when EFN features were described but the diagnosis was not suggested in the report; and “suggested”—when EFN findings were described and its diagnosis was suggested. The outcome was surprising. As shown in Figure 2, we found a progressive number of diagnoses over the years, especially after 2013. In 2011, when radiologists were still unaware of the entity, the number of correct diagnoses was zero. In 2014, Figure 2. The 4-year progression of EFN diagnosis data at a single institution. after the educational intervention, there were no more missed diagnoses of EFN at the institution. We can suggest that the dissemination of knowledge at our institution changed the pattern of the diagnosis of a disease. We believe that, in the next few years, EFN will become known worldwide, the labels “rare” and “unknown” therefore no longer being associated with this entity.

All 20 reports were reviewed and defined as follows: "not described"-when features of EFN were overlooked; "described but not suggested"-when EFN features were described but the diagnosis was not suggested in the report; and "suggested"-when EFN findings were described and its diagnosis was suggested.
The outcome was surprising. As shown in Figure 2, we found a progressive number of diagnoses over the years, especially after 2013. In 2011, when radiologists were still unaware of the entity, the number of correct diagnoses was zero. In 2014, after the educational intervention, there were no more missed diagnoses of EFN at the institution.
We can suggest that the dissemination of knowledge at our institution changed the pattern of the diagnosis of a disease. We believe that, in the next few years, EFN will become known worldwide, the labels "rare" and "unknown" therefore no longer being associated with this entity.   Pulmonary cryptococcosis is caused by fungi of the genus Cryptococcus (C. neoformans and C. gattii), which are monomorphic encapsulated yeasts that are found worldwide, particularly in soil contaminated with pigeon droppings and decomposing wood. Although infection occurs through the inhalation of airborne infectious particles, pneumonia is relatively uncommon in infected individuals. In fact, after hematogenous dissemination, infection of the central nervous system is more common than is pneumonia (1-4) .
Pulmonary cryptococcosis has a variety of clinical and pathological presentations. It can manifest in immunocompetent and immunocompromised patients, although the latter account for the majority of cases, with a wide variety of radiological abnormalities. The following are the main characteristics to be identified by CT (2-4) : location and distribution; solitary or multiple nodules that can progress to confluence or cavitation; segmental consolidation or infiltrative masses; hilar or mediastinal lymph node enlargement; pleural effusion; reticular or nodular infiltrate; linear opacities; septal thickening; and endobronchial lesions. The diagnosis of pulmonary cryptococcosis is difficult to make, because the organisms often colonize the upper airways and the symptoms are nonspecific, as are the radiological manifestations (4) .
In the diagnosis of pulmonary cryptococcosis, PET/CT plays a complementary role. In approximately 60% of patients, cryptococcal lesions show 18 F-fluorodeoxyglucose uptake that is greater than is that of the mediastinal blood pool. The SUV, a calculated measure of contrast uptake, is used in order to identify the underlying cause of such lesions, knowledge of their physiological distributions and variants being of fundamental importance for minimizing errors of interpretation (1,5,6) . Typically, low SUVs (≤ 2.5) are associated with benign lesions, whereas high SUVs (> 2.5) are associated with malignant lesions (1,5) . Sharma et al. (1) demonstrated that the SUVs of cryptococcal lesions range from 0.93 to 11.6.
When PET/CT is used in order to differentiate pulmonary nodules and to discriminate between infection and malignancy, its potential pitfalls should be borne in mind, especially in areas where the prevalence of granulomatous infection is high, as well as in immunocompromised patients (7,8) . Inflammatory and infectious lesions can show elevated metabolic rates and can therefore be misidentified as malignant lesions (5,6,8) , thus posing a diagnostic challenge (6) .
In patients with pulmonary cryptococcosis, there is great variability in the SUVs of cryptococcal lesions. Therefore, the clinical correlation, risk factors for cancer development, and geographic location, together with the PET/CT findings, are fundamental for diagnostic clarification, although it is usually necessary to perform a lung biopsy (1,5,8) .