Journal of Virological Methods 317 (2023) 114743 Contents lists available at ScienceDirect Journal of Virological Methods journal homepage: www.elsevier.com/locate/jviromet Protocols Detection of cytomegalovirus in bronchoalveolar lavage fluid from immunocompromised patients with pneumonitis by viral culture and DNA quantification C. Berengua a, b, e, *, 1, E. Miró a, b, e, C. Gutiérrez b, M. Sánchez b, A. Mulero b, P. Ramos b, M. del Cuerpo b, A. Torrego c, e, I. García-Cadenas d, e, V. Pajares c, e, F. Navarro a, b, e, R. Martino d, e, N. Rabella a, b, e a Genetics and Microbiology Department. Universitat Autònoma de Barcelona, Spain b Microbiology Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain c Pneumology Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain d Hematology Department, Hospital de la Santa Creu I Sant Pau, Barcelona, Spain e Sant Pau Institute of Biomedical Research (IIb Sant Pau), Barcelona, Spain A R T I C L E I N F O A B S T R A C T Keywords: Purpose: To compare the detection of human cytomegalovirus (HCMV) in bronchoalveolar lavage (BAL) fluid by Cytomegalovirus viral culture and quantitative polymerase chain reaction (qPCR), and to establish a viral load threshold that can HCMV identify cases of HCMV replication indicative of pneumonitis. There is currently no universal viral load cut-off to Viral culture differentiate between patients with and without pneumonitis, and the interpretation of qPCR results is Quantitative PCR Viral load challenging. Pneumonitis Methods: 176 consecutive BAL samples from immunosuppressed hosts with signs and/or symptoms of respiratory Immunosuppressed infection were prospectively studied by viral culture and qPCR. Results: Concordant results were obtained in 81.25% of the BAL samples. The rest were discordant, as only 34% of the qPCR-positive BAL samples were positive by culture. The median HCMV load was significantly higher in culture-positive than in culture-negative BAL samples (5038 vs 178 IU/mL). Using a cut-off value of 1258 IU/mL of HCMV in BAL, pneumonia was diagnosed with a sensitivity of 76%, a specificity of 100%, a VPP of 100% and VPN of 98%, and HCMV was isolated in 100% of the BAL cultures. Conclusion: We found that a qPCR-negative was a quick and reliable way of ruling out HCMV pneumonitis, but a positive result did not always indicate clinically significant replication in the lung. However, an HCMV load in BAL fluid of ≥ 1258 IU/mL was always associated with disease, whereas < 200 IU/mL rarely so. 1. Background as a mononucleosis-like syndrome. In immunosuppressed hosts, it can cause severe systemic or end-organ disease, such as pneumonitis, colitis, Human cytomegalovirus (HCMV) is a DNA virus belonging to the or retinitis, both after primary infection and upon reactivation of the family Herpesviridae, subfamily Betaherpesvirinae. The seroprevalence in latent virus. Despite the reduction in the incidence of HCMV disease, due adults is between 50% and 98%, being higher in developing countries to pre-emptive antiviral therapy or prophylaxis and better diagnostic (Boeckh and Geballe, 2011). After primary infection, the virus may tools, HCMV remains the most important and feared opportunistic remain latent in different cells and tissues and reactivate throughout life. pathogen associated with transplant patients, especially those with In immunocompetent hosts, primary infection very rarely causes severe impaired T-cell-mediated immunity (Boeckh and Geballe, 2011). disease, usually being asymptomatic, although it can present clinically HCMV pneumonitis causes significant morbidity and mortality in * Correspondence to: Microbiology Department, Hospital de la Santa Creu i Sant Pau, Universitat Autònoma de Barcelona, Sant Quintí Street, 89, 08041 Barcelona, Spain. E-mail address: cberengua@santpau.cat (C. Berengua). 1 ORCID: 0000–0001-9963–3809 https://doi.org/10.1016/j.jviromet.2023.114743 Available online 26 April 2023 0166-0934/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). C. Berengua et al. Journal of Virological Methods 317 (2023) 114743 immunosuppressed patients (transplant recipients, HIV-positive, or fluid by viral culture and qPCR for the diagnosis of pneumonitis in receiving immunosuppressive drug therapy), even with the correct immunosuppressed hosts, as well as to establish a cut-off value for the antiviral treatment. In allogeneic haematopoietic stem cell transplant HCMV load in BAL samples that discriminates between cases of active patients, HCMV pneumonitis is one of the most dangerous infectious virus replication and viral shedding, using culture as the reference complications, with a mortality rate of up to 70% (Boeckh and Geballe, technique. The secondary objective was to assess plasma HCMV load in 2011; Ibrahim et al., 1997; Erard et al., 2015; Boivin et al., 1996). The patients with suspected pneumonitis and its correlation with the VL in symptoms and radiological findings of HCMV pneumonitis are BAL fluid. non-specific, which hinders and delays diagnosis (Franquet et al., 2003). At the Fourth International Conference on HCMV in Paris in 1993, a 3. Study design report was drawn up to define and unify the concepts of HCMV infection and disease, with a recent update in 2017 (Ljungman et al., 2017). 3.1. Clinical samples Accordingly, it defines proven HCMV pneumonitis as the detection of the virus in lung tissue (by culture, histopathology, immunohisto­ A prospective study was carried out between 1 January 2021 and 31 chemistry or DNA hybridisation techniques), accompanied by signs July 2022 at the Microbiology Department of a tertiary level university and/or symptoms of pneumonia. However, this requires a lung biopsy, hospital in Barcelona. which is invasive and often contraindicated, limiting the possibility of Consecutive BAL samples (n = 176) from immunosuppressed pa­ making this diagnosis. Fortunately, bronchoalveolar lavage (BAL) has tients with signs and/or symptoms of respiratory infection, received in become a widely accepted alternative to lung biopsy for the evaluation our laboratory for virological study, were included. of lung disease (Stover et al., 1984), whereby probable HCMV pneu­ The reference methods for HCMV detection in BAL samples in our monitis can be diagnosed by virus detection in BAL fluid by culture or laboratory are traditional viral culture (in tubes) and centrifugation DNA quantification, accompanied by signs and/or symptoms of culture (shell vial). In addition, an HCMV qPCR assay was performed on pneumonia. these samples. Viral culture is considered the gold standard for the diagnosis of Retrospectively, we established whether these patients had been HCMV infection and positive cultures are highly correlated with disease tested for the HCMV load in plasma (DNAemia) in the week before or (Razonable et al., 2002; Kotton, 2013). In the late 1980 s, after BAL fluid collection. culture-centrifugation (shell vial) was introduced and became the mainstay of HCMV diagnosis, providing faster results than traditional 3.2. BAL fluid collection by bronchoscopy culture, and reducing the turnaround time from several weeks to one day (Rabella and Drew, 1990). Despite being the reference techniques Bronchoscopy was performed using standard procedures in accor­ for the diagnosis of viral infections, in most laboratories culture-based dance with the Spanish Society of Pulmonology and Thoracic Surgery methods have been replaced, due to their laboriousness and longer (SEPAR) protocol no. 28, "Quality plan for the collection and processing turnaround time, by molecular techniques such as qPCR. However, of endoscopic specimens" (Manual 28 de procedimientos SEPAR). The although qPCR analysis is more sensitive and rapid than viral culture, it technique consists of inserting the bronchoscope into the subsegment of is limited in that it detects viral genetic material without being able to the lung pre-selected by radiological criteria until the tip is wedged in a distinguish between clinically relevant HCMV replication (disease) and bronchiole with visualization of the distal airway in the center of the viral shedding or latent HCMV infection (Leuzinger et al., 2021 Jun). image. Sterile normal saline (room temperature) is injected via a Due to this lack of specificity, guidelines recommend quantification of handheld syringe and then gradually withdrawn back into the syringe. A the viral load (VL) of HCMV in BAL fluid as an aid to interpret the total of up 150 cc of sterile saline was injected in 3 aliquots of 50 cc positive qPCRs (Ljungman et al., 2017). Nevertheless, a universal VL syringes. The first fraction recovered was discarded due to risk of cut-off value that discriminates between patients with and without contamination. After collection, the sample was placed in the corre­ HCMV pneumonitis has not been established to date. This is partly due sponding specific sterile microbiological recipients and sent to the to the pathogenesis of the virus itself (potentially establishing latency or Microbiology Department within minutes of collection. From collection asymptomatic replication without causing disease) and also to the to processing (always less than 72 h) the samples were stored at 2–6ºC. numerous variables that can affect DNA quantification in BAL samples (type of patient, the BAL procedure and subsequent processing, or the 3.3. BAL viral culture DNA quantification assay) (Ljungman et al., 2017; Razonable et al., 2002; Piñana et al., 2019; Chemaly et al., 2004; Lodding et al., 2018). To avoid culture contamination, each sample was first treated with a Several studies have sought to establish whether the HCMV load in mixture of antibiotics (vancomycin 500 µL/mL and gentamycin 2.5 mg/ blood (DNAemia) is useful for the diagnosis of pneumonitis, and mL) and an antifungal agent (amphotericin 0.05 mg/mL) in a 10% whether the blood VL has a cut-off value indicative of clinically signif­ volume/volume ratio, for 30 min. icant replication in the lung. Again, no unanimous conclusions have been reached. While some reports indicate that it has a high predictive - Tube culture. 0.3 mL of the pre-treated sample was inoculated into a value for lung disease (Meyers et al., 1990; Saksirisampant et al., 2022; cylindrical tube containing a monolayer of human MRC5 cells (RD- Lee et al., 2017), others found that a considerable percentage of pneu­ Biotech, France) with maintenance medium (Minimal Essential Me­ monitis patients have undetectable HCMV in the blood as a consequence dium x10 with fetal bovine serum 2.5%), and incubated in a hori­ of local replication in the lung, concluding that a negative-DNAemia zontal position at 36 ± 2ºC for up to 21 days. Cell monolayers were does not exclude pneumonitis (Leuzinger et al., 2021 Jun; Beam et al., examined daily with an inverted microscope (x40). A culture was 2018; Westall et al., 2004). considered positive when a characteristic cytopathic effect was This lack of consensus in interpreting VL values in both BAL fluid and observed. blood renders the diagnosis of pneumonitis a challenge. Therefore, more - Centrifugation culture (shell vial). This method is used to detect well-designed prospective studies, ideally based on reference microbi­ immediate early HMCV antigen and is based on the centrifugation of ological techniques, are needed (Piñana et al., 2019). samples on a cell monolayer (MRC5 cells) arranged on a 12 mm diameter lens. Thus, 0.3 mL of each pre-treated sample was inocu­ 2. Objective lated into two vials with the cell monolayer. To increase viral ab­ sorption, the vials were centrifuged at 700 x g for 30 min. The sample The primary objectives were to compare HCMV detection in BAL was then decanted and 1 mL of growth medium (Minimal Essential 2 C. Berengua et al. Journal of Virological Methods 317 (2023) 114743 Medium x10 with fetal bovine serum 10%) was added and incubated at 36 ± 2ºC (one vial for 16 h and the other for 40 h). Then, specific monoclonal antibodies and a fluorescein conjugate (Vircell, Spain) were added, and each culture was observed under a fluorescence microscope (x400) at 16 h or 40 h. The shell vial was considered positive if one or more fluorescent nuclei of MRC5 cells were observed. - Samples were considered culture-positive if either the tube culture or shell vial were positive. 3.4. Quantification of HCMV DNA (VL) ELITe MGB® Kit CMV reagent in an ELITe InGenius® instrument (ELITechGroup, France) was used for this assay. This fully automated system performs the extraction and amplification directly from the sample, and the results are interpreted by the instrument software. The primers and HCMV probe are specific for the exon 4 region of the HCMV MIEA gene (major immediate early antigen, HCMV UL123). The primers and probe for the internal control are specific for the promoter and 5′ UTR region of the human beta-globin gene to ensure that both the BAL cell number and extraction-amplification processes occur correctly. Fig. 1. Boxplot analysis of HCMV qPCR Ct values showing differences in dis­ For both BAL fluid and plasma, 0.2 mL of sample was used. HCMV tribution and median values between BAL samples with positive and negative DNA quantification was expressed as IU/mL in BAL and plasma. Ac­ culture results. cording to manufacturer’s indications 1 IU/mL = 5 copies/mL (“1st WHO International Standard for HCMV for Nucleic Acid Amplification The median HCMV load and qPCR Ct values of culture-positive and Techniques”). -negative BAL samples were compared (Table 2, Picture 1). In culture- The system considered these samples to be positive if the HCMV load positive samples, the median VL was 4860 IU/mL (24,298 copies/mL) was above the detection limit of the reagent (178 IU/mL for BAL and higher and the Ct value lower by 6.17 units compared to those culture- plasma). The test was considered invalid if the internal control was not negative, both differences being statistically significant (p < 0.005). detected. When the HCMV load in BAL samples was < 200 IU/mL (<1000 copies/ mL), the probability of virus isolation in culture was 4.3%, as only one 3.5. Statistical analysis positive culture was obtained from the 24 BAL samples with this VL value. However, with > 900 IU/mL (>4500 copies/mL), the probability Results are presented as the number of cases and percentages for increased to 100%, as all 13 BAL samples with this HCMV load tested categorical data and as medians and the other two quartiles for ordinal positive by culture (Fig. 2). Using a qPCR cut-off value of 1258 IU/mL data. Comparison of cycle threshold (Ct) values and VL between positive (6290 copies/mL) HCMV in BAL, pneumonia was diagnosed with a and negative cultures was performed with the nonparametric Mann- sensitivity of 76%, a specificity of 100%, a VPP of 100% and VPN of 98% Whitney test. The statistical significance level was 5% (a=0.05), and (AUC of 0.987) (Fig. 3). two-tailed tests were used throughout. All analysis was performed using In all qPCR assays, the internal control (human beta-globin gene) IBM-SPSS software (version 26; SPSS. Armonk, NY). was amplified to ensure that both DNA extraction and amplification were performed correctly. 3.6. Ethical approval Retrospective analysis of DNAemia values, obtained in 58 patients in the week before or after BAL fluid collection (Table 3), revealed that The study protocol was evaluated and approved by the HSCSP Ethics 100% of patients whose BAL samples yielded positive results by both Committee (IIBSP-VIR-2014–41). qPCR and culture analysis also tested positive by qPCR in plasma. This percentage dropped to 93% in patients with BAL samples positive by 4. Results qPCR but negative by culture, and to 33% in those with negative results by both techniques. The median value of HCMV load in plasma of the A total of 176 BAL samples from 158 immunosuppressed patients first group of patients (4805 IU/mL) was more than 4000 IU/mL higher with signs and/or symptoms of respiratory infection were tested for compared to the other two groups (222.5 IU/mL and 178 IU/mL HCMV by culture and qPCR during this period. The results of viral respectively). culture and qPCR are shown in Table 1. Concordant results were ob­ tained in 81.25% of the BAL samples, the remaining 18.75% giving discordant results. Of the 50 BAL samples positive for HCMV by qPCR, only 17 (34%) were positive by culture. Fifteen (88.2%) of them were Table 2 positive by tube culture and shell-vial, while the other two (11.8%) were Median viral load (VL) and qPCR cycle threshold (Ct) value in BAL of HCMV only positive by shell-vial and negative by tube culture.Fig. 1. qPCR-positive samples according to the culture result. Positive culture Negative culture p value Table 1 (17) (33) HCMV culture and qPCR results obtained from BAL samples (PPV: positive Median VL 5038 178 p < 0.001 predictive value; NPV: negative predictive value). (Q1; Q3) BAL (IU/mL) (1008; 39,594) (178; 248) Median VL 25,188 890 p < 0.001 n = 176 Positive culture Negative culture (Q1; Q3) BAL (copies/ (5042; 197,974.5) (890; 1239) Positive qPCR 17 (9.7%) 33 (18.7%) PPV 34% mL) Negative qPCR 0 126 (71.6%) NPV 100% Median qPCR Ct value 32.10 38.27 p < 0.001 Total S 100% E 79.2% (Q1; Q3) BAL (27.22; 34.11) (36.11; 38.97) 3 C. Berengua et al. Journal of Virological Methods 317 (2023) 114743 Fig. 2. HCMV culture results in BAL samples as a function of viral load value (IU/mL), and percentage of culture positivity as a function of viral load value. Fig. 3. ROC curve for BAL HCMV qPCR for diagnosis of HCMV pneumonia. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) values as a function of HCMV load in BAL. be a highly feared disease in immunosuppressed patients (Erard et al., Table 3 2015). Currently, the diagnosis of proven HCMV pneumonitis is much Results of the HCMV DNAemia study and median HCMV viral load (IU/mL) as a less common than that of probable pneumonitis. Performing lung bi­ function of BAL culture and qPCR result. opsies is contraindicated in many cases, and less invasive methods such DNAemia study BAL Positive BAL Negative BAL Negative as BAL are preferred. Likewise, reference techniques for the diagnosis of culture and culture and culture and positive qPCR (17) positive qPCR (33) negative qPCR HCMV infection such as culture have been displaced in most laboratories (126) by simpler, faster, and more sensitive methods such as qPCR. Despite these improvements, diagnosis of pneumonitis and the interpretation of Not done (n) 23.5% (4) 54.5% (18) 76.2% (96) Negative (n) 0 3% (1) 15.9% (20) results remains challenging (Ljungman et al., 2017). Positive (n) 76.5% (13) 42.5% (14) 7.9% (10) Viral culture is recognized as a reliable technique capable of • Median VL • 4805 (461; • 222.5 (178; • 178 (178; detecting clinically relevant HCMV involvement in the lower respiratory (Q1; Q3) (IU/ 35,998) 982) 314.25) tract, as HCMV isolation is highly correlated with the disease (Razonable mL) et al., 2002; Kotton, 2013). Since the introduction of molecular tech­ niques in the 1990 s, several studies have compared the results of HCMV 5. Discussion DNA quantification in BAL fluid with those obtained by culture-based techniques, and have attempted to establish an HCMV load in BAL In the last years there have been major advances in the diagnosis and fluid indicative of clinically significant virus replication (pneumonitis). management of HCMV infection, which has led to a reduction in its Most have concluded that qPCR in BAL fluid is more sensitive and has a incidence and mortality. Despite this, HCMV pneumonitis continues to higher negative predictive value compared to culture but has less 4 C. Berengua et al. Journal of Virological Methods 317 (2023) 114743 specificity and positive predictive value (Razonable et al., 2002; Leu­ Two recent studies of lung transplant recipients, which did not use zinger et al., 2021 Jun; Boeckh et al., 2017; Liesnard et al., 1994; Riise culture, propose 4545 IU/mL (Lodding et al., 2018) and 46,000 et al., 2000; Chemaly et al., 2005; Tan et al., 2016). copies/mL (Westall et al., 2004) as VL threshold values. In both cases, To date, defining a universal VL threshold in BAL fluid that allows HCMV DNA quantification was carried out by manual qPCR, which re­ differentiation between patients with or without pneumonitis has quires prior DNA extraction and, as mentioned, provides less repro­ proved elusive (Ljungman et al., 2017; Piñana et al., 2019). Several ducible results. In contrast, several studies have concluded that it is not factors are responsible, such as the pathogenesis of HCMV itself (its possible to establish a cut-off point that allows patients with pneumo­ ability to establish latency or asymptomatic replication without causing nitis to be differentiated from those without (Piñana et al., 2019; Riise disease), or the lack of standardized procedures for qPCR analysis in BAL et al., 2000; Tan et al., 2016). samples (BAL fluid extraction and processing or the qPCR assay used). In the present study, DNAaemia was not measured in all patients, The inherent variability of the bronchoscopy technique and the dilu­ which limits the conclusions that can be drawn. All patients with HCMV- tional effect also make it difficult to obtain conclusive results. To collect positive BAL fluid (by culture and qPCR) also had HCMV-positive BAL fluid, the instilled serum is distributed heterogeneously in the lung, DNAaemia, and their plasma VL was significantly higher than in pa­ and both the bathed lung area and the amount of sample recovered tients with culture-negative BAL samples. These results agree with other varies greatly between patients. Additionally, the abandonment of cul­ reports (Ibrahim et al., 1997; Chemaly et al., 2004; Meyers et al., 1990; ture techniques by most laboratories means that many studies do not Saksirisampant et al., 2022; Lee et al., 2017; Beam et al., 2018) of a compare qPCR results with those of gold standard techniques and base significant positive correlation between HCMV DNAaemia and pneu­ their conclusions on statistical analysis and clinical interpretation. monitis. However, it should be noted that the absence of HCMV DNA in Despite attempts by the WHO to standardize HCMV qPCR results, plasma does not exclude pneumonitis. Recently, two studies found that establishing HCMV threshold values to facilitate clinical decisions has approximately one third of patients with HCMV pneumonitis had an also proved difficult for plasma or blood samples. In this case, stan­ undetectable plasma VL, demonstrating that local replication of the dardization is hindered by the variability of HCMV qPCR protocols, virus can occur in the lower respiratory tract without systemic which differ in assay performance (detection and quantification limits), involvement (Leuzinger et al., 2021 Jun; Lodding et al., 2018; Westall DNA extraction methods, target genes, and amplicon size (Razonable et al., 2004). et al., 2020). One of the strengths of this study is its use of both traditional viral The results of our study show that shell-vial was more sensitive than and shell vial culture as reference techniques, both of which are fully tube culture, as 11.8% of positive cultures were only detected by shell- established in our laboratory routine, to compare the results of qPCR vial. Furthermore, qPCR analysis of BAL fluid was shown to have a analysis in BAL samples. Furthermore, the samples were selected sensitivity and negative predictive value of 100% for the diagnosis of consecutively and prospectively, and submitted to culture and qPCR pneumonitis, using culture as the reference method, which indicates analysis as soon as they arrived at the laboratory. Finally, for optimum that qPCR is an efficient technique to rule out this disease. However, due reproducibility and accuracy, each step, from bronchoscopy procedure to its low specificity (79.2%) and positive predictive value (34%), a to sample processing for culture and qPCR, was protocolized and positive result by qPCR in BAL fluid should be treated with caution, and standardised as much as possible. Moreover, a fully automated assay for other values should be considered to avoid overdiagnosis and unnec­ HCMV DNA quantification was used, including DNA extraction and essary treatment. The HCMV load and the qPCR Ct value in BAL samples amplification and interpretation of the results. Additionally, the inclu­ may be helpful to interpret the results. In our study, in line with pub­ sion of human beta-globin as the internal control target in the qPCR lished literature, significantly higher median VL and significantly lower assay ensured that BAL samples tested contained a sufficient number of qPCR Ct values were obtained in culture-positive compared to culture- cells. The intern control was detected in all qPCR, indicating that sam­ negative BAL samples. In addition, the probability of virus isolation in ples were of sufficient quality (and reaction occurred correctly), while if culture increased with higher loads. it was not detected (the number of cells was deemed insufficient or an Based on our results, an HCMV load of 1258 IU/mL (6290 copies/ incorrect reaction) the test was declared invalid to avoid false negatives. mL) or higher in BAL fluid may be considered as indicative of clinically Finally, our study reports the results of VL in BAL in IU/mL and copies/ relevant replication in the lower respiratory tract, with high sensitivity mL, according to “1st WHO International Standard for HCMV for Nucleic and NPV, and full specificity and PPV. Virus was isolated in culture in Acid Amplification Techniques”, allowing comparison with other 100% of these cases. In the literature, the VL values in BAL fluid reported studies. to be diagnostic of pneumonitis vary widely from 500 IU/mL to The main limitation of the study was that various types of immu­ > 500.000 copies/mL (Leuzinger et al., 2021 Jun; Piñana et al., 2019; nosuppressed patients were included, which could limit the generaliz­ Lodding et al., 2018; Saksirisampant et al., 2022; Lee et al., 2017; Beam ability of the VL threshold values obtained. Additionally, although we et al., 2018; Westall et al., 2004; Boeckh et al., 2017; Riise et al., 2000; have controlled that all BAL fluids had an adequate number of cells to Chemaly et al., 2005; Tan et al., 2016). consider the results valid (using B-globin as an internal control), this Three studies published in recent years also compare the results of does not eliminate the impact that the number of cells contained in each viral culture and qPCR for HCMV detection in BAL fluid. Based on an BAL fluid may have on the VL detected. For this, the number of cells in analysis of 1109 BAL samples from immunosuppressed patients, Leu­ each sample should be quantified to obtain normalised VL, as has been zinger et al. (2020) determined that > 10,000 copies/mL was an indi­ done for other viral infections (Santos Bravo et al., 2021). Lastly, the cator of active replication. A limitation compared with the present study number of culture-positive BAL samples was low, and not all patients is that only shell vial culture was used, and not traditional culture; also, had undergone a DNAemia study, which limits the conclusions that can some of the samples (88%) were frozen and processed for qPCR analysis be drawn on this subject. retrospectively, and the qPCR assay used was not fully automated (it required prior DNA extraction), which can reduce the reproducibility of 6. Conclusion the results. After analyzing 565 BAL samples, Young et al. (2017) re­ ported that 28,774 copies/mL was diagnostic of HCMV pneumonitis, A negative qPCR for HCMV in BAL fluid quickly rules out pneumo­ whereas Boeckh et al. (2017), in a study of 271 BAL samples, proposed a nitis, but a positive qPCR in BAL is not always an indicator of the disease. much lower cut-off point of 500 IU/mL. The latter two studies only An HCMV load in BAL fluid of ≥ 1258 IU/mL (6290 copies/mL) would included patients with hematologic malignancies, so their conclusions be indicative of clinically relevant viral replication in the lung, whereas are limited to this type of patient. They also used less automated qPCR with < 200 IU/mL (<1000 copies/mL) the probability that the virus is assays than here, and did this assay on frozen samples (− 80%). replicating in the lungs is less than 5%. Even though this is a single- 5 C. Berengua et al. Journal of Virological Methods 317 (2023) 114743 center experience, and the data cannot be directly extrapolated to other Lee, H.Y., Rhee, C.K., Choi, J.Y., Lee, H.Y., Lee, J.W., Lee, D.G., 2017. Diagnosis of cytomegalovirus pneumonia by quantitative polymerase chain reaction using scenarios, the consistency of our results could be a guide for other bronchial washing fluid from patients with hematologic malignancies. Jun 13 centers that do not apply viral culture. Oncotarget 8 (24), 39736–39745. https://doi.org/10.18632/oncotarget.14504. Leuzinger, K., Stolz, D., Gosert, R., Naegele, K., Prince, S.S., Tamm, M., Hirsch, H.H., Funding 2021. Comparing cytomegalovirus diagnostics by cell culture and quantitative nucleic acid testing in broncho-alveolar lavage fluids. J. Med. Virol. 93 (6), 3804–3812. https://doi.org/10.1002/jmv.26649. Epub 2020 Nov 10. This work was supported by own funding. Liesnard, C., De Wit, L., Motte, S., Brancart, F., Content, J., 1994. Rapid diagnosis of cytomegalovirus lung infection by DNA amplification in bronchoalveolar lavages (Aug). Mol. Cell Probes 8 (4), 273–283. https://doi.org/10.1006/mcpr.1994.1039. Declaration of Competing Interest Ljungman, P., Boeckh, M., Hirsch, H.H., Josephson, F., Lundgren, J., Nichols, G., Pikis, A., Razonable, R.R., Miller, V., Griffiths, P.D., 2017. Disease definitions None of the authors report any conflict of interest. working group of the cytomegalovirus drug development forum. definitions of cytomegalovirus infection and disease in transplant patients for use in clinical trials. Jan 1 Clin. Infect. Dis. 64 (1), 87–91. https://doi.org/10.1093/cid/ciw668. Acknowledgments Lodding, I.P., Schultz, H.H., Jensen, J.U., Kirkby, N., Perch, M., Andersen, C., Lundgren, J.D., Iversen, M., 2018. Cytomegalovirus viral load in bronchoalveolar lavage to diagnose lung transplant associated cmv pneumonia (Feb). Transplantation We thank Ignasi Gich from Epidemiology Department (Hospital de la 102 (2), 326–332. https://doi.org/10.1097/TP.0000000000001927. Santa Creu i Sant Pau, Barcelona) for his contribution to the statistical Manual 28 de procedimientos SEPAR. Plan de calidad para la toma y procesado de las analysis and Luzy Brzoska for her contribution to the English edition. muestras endoscópicas. 〈https://issuu.com/separ/docs/manual_28〉. Meyers, J.D., Ljungman, P., Fisher, L.D., 1990. Cytomegalovirus excretion as a predictor of cytomegalovirus disease after marrow transplantation: importance of References cytomegalovirus viremia (Aug). J. Infect. Dis. 162 (2), 373–380. https://doi.org/ 10.1093/infdis/162.2.373. Beam, E., Germer, J.J., Lahr, B., Yao, J.D.C., Limper, A.H., Binnicker, M.J., Razonable, R. Piñana, J.L., Giménez, E., Gómez, M.D., Pérez, A., González, E.M., Vinuesa, V., R., Cytomegalovirus, C.M.V., 2018. DNA quantification in bronchoalveolar lavage Hernández-Boluda, J.C., Montoro, J., Salavert, M., Tormo, M., Amat, P., Moles, P., fluid of immunocompromised patients with CMV pneumonia (Jan). Clin. Transpl. 32 Carretero, C., Balaguer-Roselló, A., Sanz, J., Sanz, G., Solano, C., Navarro, D., 2019. (1). https://doi.org/10.1111/ctr.13149. Pulmonary cytomegalovirus (CMV) DNA shedding in allogeneic hematopoietic stem Boeckh, M., Geballe, A.P., 2011. Cytomegalovirus: pathogen paradigm and puzzle (May). cell transplant recipients: implications for the diagnosis of CMV pneumonia (May). J. Clin. Investig. 121 (5), 1673–1680. https://doi.org/10.1172/JCI45449. PMID: J. Infect. 78 (5), 393–401. https://doi.org/10.1016/j.jinf.2019.02.009. 21659716; PMCID: PMC3083799. Rabella, N., Drew, W.L., 1990. Comparison of conventional and shell vial cultures for Boeckh, M., Stevens-Ayers, T., Travi, G., Huang, M.L., Cheng, G.S., Xie, H., detecting cytomegalovirus infection. J. Clin. Microbiol. 28 (4), 806–807. https://doi. Leisenring, W., Erard, V., Seo, S., Kimball, L., Corey, L., Pergam, S.A., Jerome, K.R., org/10.1128/jcm.28.4.806-807.1990. 2017. Cytomegalovirus (CMV) DNA quantitation in bronchoalveolar lavage fluid Razonable, R.R., Paya, C.V., Smith, T.F., 2002. Role of the laboratory in diagnosis and from hematopoietic stem cell transplant recipients with CMV pneumonia. May 15 management of cytomegalovirus infection in hematopoietic stem cell and solid- J. Infect. Dis. 215 (10), 1514–1522. https://doi.org/10.1093/infdis/jix048. PMID: organ transplant recipients (Mar). J. Clin. Microbiol 40 (3), 746–752. https://doi. 28181657; PMCID: PMC5461426. org/10.1128/JCM.40.3.746-752.2002. PMID: 11880387; PMCID: PMC120290. Boivin, G., Olson, C.A., Quirk, M.R., Kringstad, B., Hertz, M.I., Jordan, M.C., 1996. Razonable, R.R., Inoue, N., Pinninti, S.G., Boppana, S.B., Lazzarotto, T., Gabrielli, L., Quantitation of cytomegalovirus DNA and characterization of viral gene expression Simonazzi, G., Pellett, P.E., Schmid, D.S., 2020. Clinical diagnostic testing for human in bronchoalveolar cells of infected patients with and without pneumonitis. J. Infect. cytomegalovirus infections. J. Infect. Dis. 221 (Suppl 1), S74–S85. https://doi.org/ Dis. 173 (6), 1304–1312. https://doi.org/10.1093/infdis/173.6.1304. 10.1093/infdis/jiz601. Chemaly, R.F., Yen-Lieberman, B., Castilla, E.A., Reilly, A., Arrigain, S., Farver, C., Riise, G.C., Andersson, R., Bergström, T., Lundmark, A., Nilsson, F.N., Olofsson, S., 2000. Avery, R.K., Gordon, S.M., Procop, G.W., 2004. Correlation between viral loads of Quantification of cytomegalovirus DNA in BAL fluid: a longitudinal study in lung cytomegalovirus in blood and bronchoalveolar lavage specimens from lung transplant recipients (Dec). Chest 118 (6), 1653–1660. https://doi.org/10.1378/ transplant recipients determined by histology and immunohistochemistry (May). chest.118.6.1653. J. Clin. Microbiol 42 (5), 2168–2172. https://doi.org/10.1128/JCM.42.5.2168- Saksirisampant, G., Kawamatawong, T., Promsombat, K., Sukkasem, W., Liamsombut, S., 2172.2004. PMID: 15131185; PMCID: PMC404658. Pasomsub, E., Bruminhent, J., 2022. A prospective study of plasma and Chemaly, R.F., Yen-Lieberman, B., Chapman, J., Reilly, A., Bekele, B.N., Gordon, S.M., bronchoalveolar lavage fluid cmv dna load quantification for the diagnosis and Procop, G.W., Shrestha, N., Isada, C.M., Decamp, M., Avery, R.K., 2005. Clinical outcome of cmv pneumonitis in immunocompromised hosts (Oct). J. Clin. Virol. utility of cytomegalovirus viral load in bronchoalveolar lavage in lung transplant 155, 105243. https://doi.org/10.1016/j.jcv.2022.105243. recipients (Mar). Am. J. Transpl. 5 (3), 544–548. https://doi.org/10.1111/j.1600- Santos Bravo, M., Nicolás, D., Berengua, C., Fernandez, M., Hurtado, J.C., Tortajada, M., 6143.2005.00747.x. Barroso, S., Vilella, A., Mosquera, M.M., Vila, J., Marcos, M.A., 2021. Severe acute Erard, V., et al., 2015. Reduced mortality of cytomegalovirus pneumonia after respiratory syndrome coronavirus 2 normalized viral loads and subgenomic RNA hematopoietic cell transplantation due to antiviral therapy and changes in detection as tools for improving clinical decision making and work reincorporation. transplantation practices. Clin. Infect. Dis. 61 (1), 31–39. Jul 1. Oct 28 J. Infect. Dis. 224 (8), 1325–1332. https://doi.org/10.1093/infdis/jiab394. Franquet, T., Lee, K.S., Müller, N.L., 2003. Thin-section CT findings in 32 Stover, D.E., Zaman, M.B., Hajdu, S.I., Lange, M., Gold, J., Armstrong, D., 1984. immunocompromised patients with cytomegalovirus pneumonia who do not have Bronchoalveolar lavage in the diagnosis of diffuse pulmonary infiltrates in the AIDS. Ajr. Am. J. Roentgenol. 181 (4), 1059–1063. https://doi.org/10.2214/ immunosuppressed host. Ann. Intern. Med. 101 (1), 1–7. https://doi.org/10.7326/ ajr.181.4.1811059. 0003-4819-101-1-1. Ibrahim, A., Gautier, E., Roittmann, S., Bourhis, J.H., Fajac, A., Charnoz, I., Terrier, P., Tan, S.K., Burgener, E.B., Waggoner, J.J., Gajurel, K., Gonzalez, S., Chen, S.F., Pinsky, B. Salord, J.M., Tancrède, C., Hayat, M., Bernaudin, J.F., Pico, J.L., 1997. Should A., 2016. Molecular and culture-based bronchoalveolar lavage fluid testing for the cytomegalovirus be tested for in both blood and bronchoalveolar lavage fluid of diagnosis of cytomegalovirus pneumonitis. Feb 10 Open Forum Infect. Dis. 3 (1), patients at a high risk of CMV pneumonia after bone marrow transplantation? (Jul). ofv212. https://doi.org/10.1093/ofid/ofv212. Br. J. Haematol. 98 (1), 222–227. https://doi.org/10.1046/j.1365- Westall, G.P., Michaelides, A., Williams, T.J., Snell, G.I., Kotsimbos, T.C., 2004. Human 2141.1997.1752987.x. cytomegalovirus load in plasma and bronchoalveolar lavage fluid: a longitudinal Kotton C.N. CMV: Prevention, Diagnosis and Therapy. Am J Transplant. 2013 Feb;13 study of lung transplant recipients. Sep 15 J. Infect. Dis. 190 (6), 1076–1083. Suppl 3:24–40; quiz 40. doi: 10.1111/ajt.12006. PMID: 23347212. https://doi.org/10.1086/422327. 6