COVID-19-Associated Pulmonary Aspergillosis, Fungemia, and Pneumocystosis in the Intensive Care Unit: a Retrospective Multicenter Observational Cohort during the First French Pandemic Wave

ABSTRACT The aim of this study was to evaluate diagnostic means, host factors, delay of occurrence, and outcome of patients with COVID-19 pneumonia and fungal coinfections in the intensive care unit (ICU). From 1 February to 31 May 2020, we anonymously recorded COVID-19-associated pulmonary aspergillosis (CAPA), fungemia (CA-fungemia), and pneumocystosis (CA-PCP) from 36 centers, including results on fungal biomarkers in respiratory specimens and serum. We collected data from 154 episodes of CAPA, 81 of CA-fungemia, 17 of CA-PCP, and 5 of other mold infections from 244 patients (male/female [M/F] ratio = 3.5; mean age, 64.7 ± 10.8 years). CA-PCP occurred first after ICU admission (median, 1 day; interquartile range [IQR], 0 to 3 days), followed by CAPA (9 days; IQR, 5 to 13 days), and then CA-fungemia (16 days; IQR, 12 to 23 days) (P < 10−4). For CAPA, the presence of several mycological criteria was associated with death (P < 10−4). Serum galactomannan was rarely positive (<20%). The mortality rates were 76.7% (23/30) in patients with host factors for invasive fungal disease, 45.2% (14/31) in those with a preexisting pulmonary condition, and 36.6% (34/93) in the remaining patients (P = 0.001). Antimold treatment did not alter prognosis (P = 0.370). Candida albicans was responsible for 59.3% of CA-fungemias, with a global mortality of 45.7%. For CA-PCP, 58.8% of the episodes occurred in patients with known host factors of PCP, and the mortality rate was 29.5%. CAPA may be in part hospital acquired and could benefit from antifungal prescription at the first positive biomarker result. CA-fungemia appeared linked to ICU stay without COVID-19 specificity, while CA-PCP may not really be a concern in the ICU. Improved diagnostic strategy for fungal markers in ICU patients with COVID-19 should support these hypotheses. IMPORTANCE To diagnose fungal coinfections in patients with COVID-19 in the intensive care unit, it is necessary to implement the correct treatment and to prevent them if possible. For COVID-19-associated pulmonary aspergillosis (CAPA), respiratory specimens remain the best approach since serum biomarkers are rarely positive. Timing of occurrence suggests that CAPA could be hospital acquired. The associated mortality varies from 36.6% to 76.7% when no host factors or host factors of invasive fungal diseases are present, respectively. Fungemias occurred after 2 weeks in ICUs and are associated with a mortality rate of 45.7%. Candida albicans is the first yeast species recovered, with no specificity linked to COVID-19. Pneumocystosis was mainly found in patients with known immunodepression. The diagnosis occurred at the entry in ICUs and not afterwards, suggesting that if Pneumocystis jirovecii plays a role, it is upstream of the hospitalization in the ICU.

This is a useful report documenting the scope of fungal infections complicating SARS-CoV2 in France. However the data are presented in a haphazard way that I think would confuse the reader. Specific comments follow to try to improve the flow of the paper. 1. The authors focused on 244 subjects that required intensive care and then go on to present risk factors such as age, sex and mortality. However to be useful and to place these data in context it would be important to compare this group with COVID-19 without evidence of fungal infection. Is fungal infection an independent risk factor of mortality independent of age and sex? 2. The statistical test used to analyze the data in Figure 1 should be added to the Legend. 3. What was the denominator for the 283 fungal infections? What was the incidence of fungal infection? 4. Page 6: What are groups 1, 2, and 3? They are not defined at all in the results section. How were they chosen? Were they defined a priori, prior to the study or defined in a post-hoc manner?
Reviewer #2 (Comments for the Author): This is a French multicentre retrospective study during 4 months (February 1st to May 31st 2020) reporting 3 encountered fungal co-infections in COVID-19 patients admitted to ICU. The authors present a large collection of enrolled patients and the study is of interest for mycologists and ICU doctors alike. I have a few thoughts for authors consideration. The authors suggest that CAPA could be hospital acquired because the median time to diagnosis was apparently 9 days. This can only be considered if screening of respiratory samples was done for all patients from day 0 onwards. Generally patients are already admitted with Aspergillus colonization, the latter is most common in patients with compromised airways for example due to COPD. But also smoking habits predisposes to Aspergillus colonization. The authors rightfully tone down the diagnosis of CAPA especially in the early months when no guidelines were available but the significance of the presence of Aspergillus in airway specimens (detected by culture, galactomannan antigen or specific PCR) remains to be fully understood. (I suggest to discuss this, also French study more in detail which was recently published in Mycoses 2021;64(9):980-988.). The authors discuss the occurrence of influenza associated aspergillosis to occur at a median of 3 days (line 262) and suggest herewith that this is community acquired. However influenza destroys the bronchial epithelium much more and quicker than SARS-CoV2 therefore the later diagnosis of CAPA may reflect this. Secondly, I suppose that in the period feb-may very few patients may have been treated with systemic steroids (N Engl J Med. 2020; 10.1056/NEJMoa2021436). This may have impacted the diagnosis of CAPA (less due to less steroid use) in this first wave compared to the later waves. It is not clear how many patients received full doses of steroids. Line 317-318 states finding of a higher mortality in patients receiving steroids while line 320-321 states lack of details of steroid treatment. This may be biased since dexa use in your study was not standard use for every patient as recommended by the Recovery trial. Sickest patients might have been selected in your study to receive "last resort" dexa treatment. This influence may be discussed in more detail i.e. what was proportion of steroids used in CAPA cases. There is only one (small) study comparing CAPA during first and second wave in Europe. Thirdly, the authors state that "Aspergillus azole resistance was published only once" (line 204 and 316). Does this mean that all > 110 cultured Aspergillus was tested with AFST or azole plate screening to support this statement? If yes the results need to be mentioned (for example by stating the MIC-ranges and MIC90) Line 315-316." A fumigatus azole resistance was mentioned only once?" or do you mean was "published only once (9)"? The latter is of course not correct since azole resistance and CAPA was also found once in Ireland (Med Mycol Case Rep. 2020; 10.1016/j.mmcr.2020.06.005) and twice (2 cases out of 13 CAPA patients) in the Netherlands (Mycoses. 2021; 64(4): 457-464). Line 361: does it also support routine AFST of patients with CAPA? Line 364. CAPA could be hospital-acquired because it was diagnosed 1 week after ICU admission is speculation in the absence of systematic screening from day 0 onwards. See also above. Minor suggestions: In introduction: The study concerns CAPA however, only influenza associated aspergillosis is referenced (1,2). I don't see this relation especially since the latter disease is very different from CAPA. Therefore I suggest to discuss the first published (European) cases of CAPA in May 2020 (Mycoses.2020;63 (6)

Preparing Revision Guidelines
To submit your modified manuscript, log onto the eJP submission site at https://spectrum.msubmit.net/cgi-bin/main.plex. Go to Author Tasks and click the appropriate manuscript title to begin the revision process. The information that you entered when you first submitted the paper will be displayed. Please update the information as necessary. Here are a few examples of required updates that authors must address: • Point-by-point responses to the issues raised by the reviewers in a file named "Response to Reviewers," NOT IN YOUR COVER LETTER. • Upload a compare copy of the manuscript (without figures) as a "Marked-Up Manuscript" file. • Each figure must be uploaded as a separate file, and any multipanel figures must be assembled into one file. For complete guidelines on revision requirements, please see the journal Submission and Review Process requirements at https://journals.asm.org/journal/Spectrum/submission-review-process. Submissions of a paper t hat does not conform t o Microbiology Spect rum guidelines will delay accept ance of your manuscript . " Please return the manuscript within 60 days; if you cannot complete the modification within this time period, please contact me. If you do not wish to modify the manuscript and prefer to submit it to another journal, please notify me of your decision immediately so that the manuscript may be formally withdrawn from consideration by Microbiology Spectrum.
If you would like to submit an image for consideration as the Featured Image for an issue, please contact Spectrum staff.
If your manuscript is accepted for publication, you will be contacted separately about payment when the proofs are issued; please follow the instructions in that e-mail. Arrangements for payment must be made before your article is published. For a complete list of Publicat ion Fees, including supplemental material costs, please visit our website.
Corresponding authors may join or renew ASM membership to obtain discounts on publication fees. Need to upgrade your membership level? Please contact Customer Service at Service@asmusa.org.
Thank you for submitting your paper to Microbiology Spectrum.

Reviewer comments: Reviewer #1 (Comments for the Author):
This is a useful report documenting the scope of fungal infections complicating SARS-CoV2 in France. However the data are presented in a haphazard way that I think would confuse the reader. Specific comments follow to try to improve the flow of the paper. 1. The authors focused on 244 subjects that required intensive care and then go on to present risk factors such as age, sex and mortality. However to be useful and to place these data in context it would be important to compare this group with COVID-19 without evidence of fungal infection. Is fungal infection an independent risk factor of mortality independent of age and sex? A: As explicitly explained in the manuscript (lines 350 to 362), the present study is a multicenter declarative study, and not a multicenter prospective longitudinal study with scheduled testing / imaging at entry and during hospitalization in all patients to decipher the risk factors for mortality in COVID-19 patients. This later study design was not possible in the emergency context of the first wave of COVID-19. Most ICU patients during the first wave were not tested for IFDs. To compare as much as possible with the overall mortality, we obtain the overall mortality in some of the corresponding centers during the same period. This mortality was much lower than the mortality observed in our patients (Line 254: 22.6% vs 50.6% vs 22.6%, p <10-8), except for Pneumocystis positive patients. Figure 1 should be added to the Legend. A: Done 3. What was the denominator for the 283 fungal infections? What was the incidence of fungal infection? A: See comment above. In the absence of scheduled collection of microbiological tests in ICU patients, it is not possible to calculate an incidence. It is clearly presented in the discussion, where we did not speak about incidence (lines 350-362).

The statistical test used to analyze the data in
4. Page 6: What are groups 1, 2, and 3? They are not defined at all in the results section. How were they chosen? Were they defined a priori, prior to the study or defined in a post-hoc manner? A: The three groups have been clearly defined in the Methods section, which unfortunately comes after the Results. We have put the information in the results (lines 185-187) to avoid any questions. Risk factors were included in the questionnaire sent to each participant prior to the analysis which intended to compare the three groups.

Reviewer #2 (Comments for the Author):
This is a French multicentre retrospective study during 4 months (February 1st to May 31st 2020) reporting 3 encountered fungal co-infections in COVID-19 patients admitted to ICU. The authors present a large collection of enrolled patients and the study is of interest for mycologists and ICU doctors alike. I have a few thoughts for authors consideration.
1/ The authors suggest that CAPA could be hospital acquired because the median time to diagnosis was apparently 9 days. This can only be considered if screening of respiratory samples was done for all patients from day 0 onwards. Generally patients are already admitted with Aspergillus colonization, the latter is most common in patients with compromised airways for example due to COPD. But also smoking habits predisposes to Aspergillus colonization. A: We agree that screening at entry might help diagnose the timing of CAPA acquisition, but it would be difficult to perform BAL or even tracheal aspiration before these procedures are needed, except in one controlled study with ethical approval. We cannot affirm from the literature and in our experience that all the patients had airway colonization. Indeed we think that is not expected in patients without a respiratory history (60% of the present CAPAs, Table 1). Of course, patients could be colonized between onset of COVID-19 and intensive care hospitalization, but this is highly speculative and there is no easy tool to determine this putative colonization. On the other hand, discussing the hospital acquisition as suspected by time of occurrence after admission (9 days) allows one to suggest preventive measures. For example, there is no mention in the literature that the air in intensive care units could be controlled for the fungal contamination of the environment or tat basic measures to avoid Aspergillus contamination should be implemented and controlled, which could be discussed with our data (see lines 265-266 ref # 16).
The authors rightfully tone down the diagnosis of CAPA especially in the early months when no guidelines were available but the significance of the presence of Aspergillus in airway specimens (detected by culture, galactomannan antigen or specific PCR) remains to be fully understood. (I suggest to discuss this, also French study more in detail which was recently published in Mycoses 2021;64(9):980-988.). A: We agree that the diagnosis of CAPA is not straightforward and that the definitions may help report epidemiological data. However, the comment we introduced in our discussion is that the more markers you have, the worse the prognosis. Therefore, waiting to be sure of the diagnosis could delay the treatment decision. Hence our comment (lines 292-293) to rely on culture, the main initial finding that is frequent as already done in ICU (ref # 23-24). Of course, understanding the meaning of a positive culture is mandatory, even though this question has been pending for many years.
We were not aware of this article when we submitted our manuscript. The authors point out that age, COPD and a positive GM are the main prognostic factors of CAPA. In our study, COPD was not the group of patients with the highest mortality (which is hematology as expected in our opinion). For GM serum considered alone, positivity was not associated with mortality (p = 0.085). On the other hand, we show that it is the association of mycological markers which is associated with mortality (see Figures 2 and 3). Overall, it is review of 35 studies reporting 134 clinical cases from different countries, with different diagnostic to and treatments thus clearly completely different in design from ours which is also not limited to CAPA. We thus don't think discussing these specific results rather than the references we already selected is more relevant since we will mainly only underline the differences in design.