Sequential diagnostics of pulmonary fungal infections

Invasive pulmonary aspergillosis- or mucormycosis are the most common pulmonary
fungal infections (PFI) diagnosed in patients receiving intensive chemotherapy
for myeloid malignancies or as a conditioning regimen for allogeneic stem cell
transplantation. Prolonged neutropenia, mucosal damage and reduced cellular
immunity are risk factors that contribute to the high incidence of mold
infections in these patients. Several studies have proven the efficacy of
primary mold active azole prophylaxis in reducing the incidence and mortality
of PFI in these patients and prophylaxis is recommended by various guidelines
during cytotoxic treatment up until immunological recovery which can take weeks
to months. However, there are negative effects of prophylaxis that were not
studied in the aforementioned trials. For example, there are important
bidirectional drug interactions with new targeted antileukemic and antiviral
drugs that were introduced after these trials were published and that were not
studied in the registration trials. There is also increasing azole resistance
in Europe and breakthrough infections with azole susceptible or resistant molds
are also frequently reported. The alternative to universal prophylaxis is a
pre-emptive or diagnostic driven approach to PFI management, where mold active
treatment is started only when an invasive mold infection is suspected because
of clinical sings of infection (e.g. fever during neutropenia or pulmonary
symptoms) or positive serum aspergillus antigenemia while results of definitive
diagnostic testing are pending. We have applied this approach in our
institution; using serial aspergillosis antigenemia monitoring and low dose
CT-imaging for persistent fever has led to a reduction of the use of azole
drugs to just 15% of our population which would have otherwise all received
universal azole prophylaxis, thus preventing overtreatment in the majority of
the patients. Surprisingly, all of our PFI cases were diagnosed primarily by
CT-imaging and subsequent CT-directed bronchoalveolar lavage of suspected
pulmonary lesions, while serum antigenemia remained negative. These findings
indicate that we cannot rely on positive antigenemia as an early marker for
PFI, at the same time they point to the advantage of CT-imaging in the early
detection of PFI . This is in line with current studies supporting the role of
CT-imaging as a screening tool for PFI even before the start of chemotherapy
given that around 20% of patients already have a PFI before starting treatment.
Moving forward with our diagnostic approach to PFI we will integrate an early
CT-scan of the chest as a screening tool before the start of chemotherapy
starting in 2023. While this baseline CT will offer early diagnosis of PFI to
some, we foresee some important questions that have remained unanswered in the
current literature. First, even though early CT-imaging is already routine
practice in several institutions, there are no known reports of the evolution
kinetics of PFI lesions early during the course of chemotherapy-induced
neutropenia, before fever or positive serum antigenemia appear. In fact we do
not know when fungal lesions first become apparent on CT-imaging nor do we know
at what speed they develop or when they incite fever. Similarly, response to
treatment currently relies on follow up imaging but there are no
recommendations on lesion response criteria. In this diagnostic strategy
protocol we want to explore computational assisted sequential CT-pulmonary
imaging as a tool for early detection and quantification of PFI lesions and
also to guide early initiation and continuation of antifungal treatment during
intensive chemotherapy courses for myeloid malignancies or allogeneic stem cell
transplantation.

To determine the clinical utility of sequential computational assisted
CT-guided treatment of PFI in the hemato-oncology population.

non-randomised prospective feasibility study with additional diagnostic
intervention

One additional non contrast enhanced low dose CT thorax.
Routine of care follow up or clinically triggered CT thorax will be performed
following a weekly interval if feasible

CT scanning is an imaging modality that is extensively used in standard care.
Low dose CT scans are performed at 100kV with automated exposure control (Sure
Exposure). Image reconstruction is performed using AiCE deep learning
reconstruction with a slice thickness of 0.5 in lung and mediastinal window.
Average dose-length-product (DLP) is estimated at 40.In this study one
additionalCT scan will be performed which is needed to better understand the
kinetics of development of invasive pulmonary fungal infections and the
combined strategy will identify patients that have fungal infections earlier
and also identify patients that are in need of antifungal treatment even in the
absence of fever. On the other hand it also contributes to safety, in patients
that have no lesions it is safe to (continue to) withhold antifungal treatment
even when there is persistent fever. In the future this study will provide
evidence to decrease the number of patients that are unnecessarily exposed to
antifungal drugs, decrease the duration in which they are exposed to these
drugs and also standardize radiological reporting for the clinician and future
research.