By Carol A. Kemper, MD, FACP

Clinical Associate Professor of Medicine, Stanford University, Division of Infectious Diseases, Santa Clara Valley Medical Center

Outcome of Coccidioidomycosis in the Pre-Antifungal Era

SOURCE: Bays DJ, Thompson GR, Reef S, et al. Natural history of disseminated coccidioidomycosis: Examination of the VA–Armed Forces database. Clin Infect Dis Aug 2020; Aug. 11; ciaa1154. [Online ahead of print].

These authors describe the natural history of coccidioidomycosis before the availability of antifungal treatment in a large, healthy cohort of military veterans. The fate of non-disseminated (non-DCM), disseminated (DCM), and central nervous system (CNS) disease was observed, both with regard to the presentation of initial infection and mortality.

A total of 669 cases of cocci infection were identified in veterans diagnosed from 1955-1958 and followed through at least 1966, and in many cases for up to 10 to 30 years. Insufficient data were available for 104 patients, and another 64 were excluded for receipt of topical or systemic treatment with amphotericin during the initial period of followup through 1966. Of the remaining 531 cases, 87% were classified as non-disseminated (non-DCM, no evidence of disease outside the lungs), 8.3% had disseminated disease, and 4.7% involved the CNS (meninges/central nervous system, with or without other areas of dissemination). Overall, this was a fairly young and healthy cohort. The median age for each of the three groups was mid-30s, and most had no comorbidities or underlying pulmonary disease. Diabetes was present in 4.5%, with similar prevalence for the three groups. As expected, Black Americans and Filipinos were more likely to have disseminated infection; close to one-third of Black Americans and Filipinos in the cohort had dissemination and/or CNS involvement, and there was a trend toward increased disseminated infection in Latinos.

The outcomes of the three groups were very different. Regarding the outcome of primary infection, pulmonary nodules occurred or developed in 39.6%, 13.6%, and 20% of those with non-DCM, DCM, and CNS disease (P < 0.001) and cavitation developed in 34.2%, 9%, and 8% of those with non-DCM, DCM, and CNS disease, respectively (P < 0.001). This was construed as evidence of differing host immune responses between the three groups at the time of initial exposure. Note that these data were collected prior to the availability of computed tomography (the availability of which may have altered these observations).

Dissemination occurred early in the course of disease in the majority of patients, and was present at the time of initial infection or as an initial manifestation in 41% of those with DCM and 56% of those with CNS infection. Most cases of dissemination were diagnosed within two to six months of presentation. Only 28% and 36% of patients with DCM and CNS involvement, respectively, developed signs or symptoms of disseminated infection more than six months after initial diagnosis, consistent with other data suggesting antifungal therapy does not prevent the dissemination of infection.

All-cause mortality was substantial for patients with CNS disease (96%) compared with DCM patients (29.6%) and non-DCM patients (5.4%) (P < 0.00001). At a time before the availability of antifungal therapy, only 3/459 patients (0.65%) with non-DCM disease died as the result of their cocci infection. Two of these patients had cavitary lung disease and a third had marked elevation of cocci serology suggestive of occult disseminated infection.

PET Imaging for Fever of Unknown Origin

SOURCE: Wright WF, Auwaerter PG, Dibble EH, et al. Imaging a fever — Redefining the role of 18FDG-PET/CT in FUO investigations. Clin Infect Dis 2020; Aug 23;ciaa1220. [Online ahead of print].

Infectious disease physicians are at the front line for investigating fever of unknown origin (FUO), although historically only about one-third of such cases are the result of an infection. Historically, about one-third of FUO cases are secondary to malignancy, somewhere around 16% to 55% are due to noninfectious inflammatory disorders, while about 5% elude diagnosis. I liken FUO workups to peeling an onion, with an initial “layer” of laboratory and radiographic studies, followed by “layers” of subsequent studies. Depending on how quickly you peel through these layers and are able to schedule and secure results, FUO workups may take weeks before arriving at a solution.

Increasingly, fluoro-D-glucose-positron emission tomography/computed tomography (FDG-PET/CT) imaging has found a role in the investigation of FUO. Using PET/CT, my partners have diagnosed an elusive case of malignancy and another case of giant cell arteritis involving the arterial system extending from the carotids to the iliac arteries. In contrast to these successes, FDG-PET/CT was not helpful in the diagnosis of a rare case of Cogan’s syndrome (a variant of giant cell arteritis), which resulted in two months of intermittent fever and hearing loss in a young woman, and another case of drug hypersensitivity syndrome with eosinophilia and systemic symptoms (DRESS), both of which remained diagnostic challenges.

There are increasing data to support the use of FDG-PET and FDG-PET/CT in FUO workup. These authors provide a review of 50 clinical studies, published between 2011-2019 (44 retrospective and six prospective studies), and examine the contribution of PET/CT imaging to the investigation of FUO. Although some studies couch their results in terms of specificity and sensitivity, the diagnostic yield of PET/CT scanning is more appropriately presented in terms of “agreement” or “non-agreement,” since often there is no gold standard or reference standard for comparison. Examining pooled data, PET/CT imaging showed a percentage positive agreement in diagnostic yield of 79% to 94%, and a percentage negative agreement of 44% to 97%. The overall percent agreement varied from 54% to 94%. Further, several studies observed that the use of PET/CT achieved a speedier diagnosis and decreased the overall cost of FUO workup by at least 35% to 60% by reducing the number of diagnostic evaluations and the number of invasive evaluations. Tests that commonly were no longer required included various viral and serologic studies, bone marrow and other biopsies, endoscopies, and various ultrasound and magnetic resonance imaging (MRI) studies. Several studies also demonstrated reductions in hospital days. Overall, PET/CT was 4.6 times more useful than labeled leukocyte scanning.

More than 90% of FDG-PET/CT imaging is currently used for diagnosis, staging, and monitoring in oncology. As recently as Jan. 9, 2021, Medicare issued a decision memorandum stating inadequate supporting data to justify the cost of PET/CT in the diagnosis of various infections and inflammatory disorders, including investigations of FUO, chronic osteomyelitis, chronic hip arthroplasty infection, sarcoid, etc. This prompted certain insurance companies (e.g., Aetna) to follow suit with a similar position paper indicating their willingness to cover the cost of PET/CT for certain conditions, such as cardiac sarcoid and a long list of malignancies, but not for the investigation of FUO, aortitis/large vessel vasculitis, Takayasu’s arteritis, or other noninfectious inflammatory disorders. Of course, when you order a PET/CT for your diagnostic workup, how do you know whether you are going stumble across one of the 25 malignancies listed or something else that would not be covered? Ironically, Medicare does cover the cost of nuclear leukocyte imaging, which is more costly than PET/CT, although far less useful.

High Rates of Thromboembolism in COVID-19

SOURCE: Wichmann D, Sperhake J-P, Lütgehetmann M, et al. Autopsy findings and venous thromboembolism in patients with COVID-19: A prospective cohort study. Ann Intern Med 2020;173:268-277.

Prospectively obtained autopsy data from a single academic center in Hamburg, Germany, revealed a significant frequency of occult thromboembolism in the first 12 consecutive COVID-19 deaths. Two of the patients died in the outpatient setting following unsuccessful cardiac resuscitation, five patients died in intensive care, and five patients had advance directives for best active care on a medical unit. The median age was 73 years, and three-fourths were men. Coronary artery disease (50%) and obesity and chronic obstructive pulmonary disease (25%) were common. The cause of death in all 12 cases was found in the lungs or pulmonary vasculature. Although venous thromboembolism (VTE) was not suspected prior to death in any of the patients, autopsy revealed that seven of the 12 patients (58%) had VTE. Four patients died of massive pulmonary embolism (PE), all derived from the deep veins of the lower extremity, and the other three had lower extremity deep vein thrombosis (DVT) without PE. D-dimers varied from 20 nmol/L to > 1,905 nmol/L, with a median of 495 nmol/L. Eight of the patients had histopathologic evidence of diffuse alveolar damage, consistent with early acute respiratory distress syndrome (ARDS), with microvascular thromboemboli, capillary congestion, and interstitial edema. In addition, six of nine men had fresh thrombosis in the prostatic venous plexus.

PE should be suspected in any COVID-19 patient with an abrupt deterioration in respiratory or hemodynamic status. Further, we have been routinely screening COVID-19 patients for upper and lower extremity thromboembolism, especially those with elevated D-dimers. Although the radiology department resists performing these noninvasive studies in the rooms of these patients in airborne and contact isolation, the yield has been surprisingly high.