Anne Thiebaut-Bertrand

Variability of Voriconazole Plasma Concentrations after Allogeneic Hematopoietic Stem Cell Transplantation: Impact of Cytochrome P450 Polymorphisms and Comedications on Initial and Subsequent Trough Levels

ABSTRACT Voriconazole (VRC) plasma trough concentrations (Cmin) are highly variable, and this could affect treatment efficacy and safety in patients undergoing allogeneic hematopoietic stem cell transplantation (AHSCT). We aimed to describe the intra- and interindividual variation of VRC Cmin throughout the course of VRC therapy and to identify the determinants of this variation. Clinical data, medications, and VRC Cmin (n = 308) of 33 AHSCT patients were retrospectively collected. Cytochrome P450 (CYP450) genotypes of CYP2C19, CYP3A4, and CYP3A5 patients were retrospectively determined before allografting, and a combined genetic score was calculated for each patient. The higher the genetic score, the faster the metabolism of the patient. The VRC Cmin inter- and intraindividual coefficients of variation were 84% and 68%, respectively. The VRC dose (D) was correlated to VRC Cmin (r = 0.412, P < 0.0001) only for oral administration. The administration route and the genetic score significantly affected the initial VRC Cmin. Considering oral therapy, patients with a genetic score of <2 had higher initial VRC Cmin/D than patients with a genetic score of >2 (P = 0.009). Subsequent VRC Cmin remained influenced by the genetic score (P = 0.004) but were also affected by pump proton inhibitor comedication (P < 0.0001). The high variability of VRC Cmin in AHSCT patients is partially explained by the route of administration, treatment with pump proton inhibitors, and the combined genetic score. This study suggests the interest in combined genetic score determination to individualize a priori the VRC dose and underlines the need for longitudinal therapeutic drug monitoring to adapt subsequent doses to maintain the VRC Cmin within the therapeutic range.

Characteristics and clinical relevance of the quantitative touch-down major surface glycoprotein polymerase chain reaction in the diagnosis of Pneumocystis pneumonia

The evaluation of quantitative polymerase chain reaction (PCR) characteristics can increase the accuracy of the laboratory diagnosis of Pneumocystis pneumonia (PCP). Between July 2008 and September 2009, 66 non-sequential prospective bronchoalveolar lavage (BAL) samples, obtained from five HIV-infected and 49 non HIV-infected patients were investigated, using a quantitative-touch-down-PCR to determine the number of copies of major surface glycoprotein (MSG) genes of Pneumocystis jirovecii (q-TD-MSG-PCR). PCP was confirmed by microscopic observation of Pneumocystis, radio-clinical and therapeutic data in 18/54 patients. For PCP, the cut-off was 54.3 MSG copies per ml of BAL fluid. The PCR was positive in these same 18 cases and it was the only positive assay in two cases and the earliest diagnosis test in one case of PCP relapse. The likelihood positive ratio, sensitivity and specificity of the q-TD-MSG-PCR were 44, 100% and 97.7%, respectively. The Predictive Negative Value was 100% and the Predictive Positive Value of 95.5%, the intra- and inter-assay variability values were 2.7% (at more than 30 MSG copies) and 11.7% (at 10,000 MSG copies), respectively. Quantitative PCR can help diagnose PCP even in cases of low Pneumocystis load and might decrease morbidity in association with very early specific treatments.

Seven-year surveillance of nosocomial invasive aspergillosis in a French University Hospital

OBJECTIVES This study aims at describing the evolution of the epidemiology of invasive aspergillosis (IA) in a French University Hospital focussing on nosocomial cases, in order to assess the efficiency of the environmental preventive measures which were implemented. METHODS From 2003 to 2009, IA cases were reviewed monthly and classified according to the EORTC/MSG criteria and the origin of contamination. RESULTS Five proven and 65 probable IA cases were diagnosed. Most of the cases (74.3%) occurred in patients with haematological malignancies. Incidences of IA and nosocomial IA (NIA) were 0.106 and 0.032 cases per 1000 admissions, respectively. All the 21 NIA cases occurred in the absence of air treatment (laminar air flow facilities or Plasmair decontamination units) and/or during construction works. The 3-month and 1-year overall survival rates were 50.6% [38.2-61.7] and 31.1% [20-42.9] respectively, and did not differ according to the origin of contamination. CONCLUSION Nosocomial IA still accounted for a third of all IA cases diagnosed from 2003 to 2009 and mainly occurred in the absence of environmental protective measures, which were confirmed to be effective when applied. Our results show that extension and/or reinforcement of these measures is needed, especially in the haematology unit and during construction works.

Added Value of Next-Generation Sequencing for Multilocus Sequence Typing Analysis of a Pneumocystis jirovecii Pneumonia Outbreak1

Pneumocystis jirovecii is a major threat for immunocompromised patients, and clusters of pneumocystis pneumonia (PCP) have been increasingly described in transplant units during the past decade. Exploring an outbreak transmission network requires complementary spatiotemporal and strain-typing approaches. We analyzed a PCP outbreak and demonstrated the added value of next-generation sequencing (NGS) for the multilocus sequence typing (MLST) study of P. jirovecii strains. Thirty-two PCP patients were included. Among the 12 solid organ transplant patients, 5 shared a major and unique genotype that was also found as a minor strain in a sixth patient. A transmission map analysis strengthened the suspicion of nosocomial acquisition of this strain for the 6 patients. NGS-MLST enables accurate determination of subpopulation, which allowed excluding other patients from the transmission network. NGS-MLST genotyping approach was essential to deciphering this outbreak. This innovative approach brings new insights for future epidemiologic studies on this uncultivable opportunistic fungus.

Contribution of a Simple Bioassay in Effective Therapeutic Drug Monitoring of Posaconazole and Voriconazole.

Background: With the constantly growing incidence of invasive fungal infections, any failure of antifungal treatment is worrying. Azole antifungals present high variability of their plasma trough concentrations (Cmin), justifying their therapeutic drug monitoring (TDM). The authors aimed to develop a simple bioassay to determine the in vitro growth inhibition diameter (ID) and to correlate this ID with Cmin in patients treated with voriconazole or posaconazole. Methods: The bioassay determined the ID for Candida parapsilosis using a disk diffusion method. Calibration curves were built for posaconazole and voriconazole in water and in 45% plasma. ID was determined in plasma from patients currently undergoing TDM for posaconazole (n = 73) or voriconazole (n = 90). Results: In water or plasma spiked with antifungals and patient samples, cubic regression between ID and Cmin gave coefficient of determination values of 0.997, 0.999, and 0.819, respectively, for posaconazole and 0.996, 0.990 and 0.925, respectively, for voriconazole (P < 0.001 for each curve). Calibration curves with or without plasma did not differ. For voriconazole, Cmin of 1 and 4.7 mg/L corresponded to 54% and 90% of maximal ID, respectively. For posaconazole, Cmin of 0.5, 0.7, and 1 mg/L corresponded to 26%, 40%, and 53% of maximal ID, respectively. Conclusions: Bioassay could be useful to better characterize the antifungal therapeutic range and brings additional information to the interpretation of TDM in patients for whom Cmin alone is insufficient to adjust the antifungal dosage.

Usefulness of pan-fungal NASBA test for surveillance of environmental fungal contamination in a protected hematology unit.

A pan-fungal nucleic acid sequence based applification (NASBA) test was adapted and used for the first time to detect and quantify the level of filamentous fungi in environmental samples. Surface samples (n = 356) collected in a controlled air flow hematology ward were tested by mycological culture and the pan-fungal NASBA test. The overall percentage of agreement between culture and NASBA was 88%, the Kappa coefficient was equal to 0.61 (95%CI = [0.51; 0.72]). This pan-fungal NASBA test could be a promising tool to rapidly monitor the absence of molds in controlled environments.

Minimum inhibitory concentrations of amphotericin B against Candida krusei isolates from a French teaching hospital laboratory: a retrospective study over 8 years.

Recent studies have shown decreased susceptibility of Candida krusei to amphotericin B (AmB), in addition to its inherent resistance to fluconazole. The susceptibility of C. krusei to AmB was studied in the Parasitology–Mycology laboratory of Grenoble Teaching Hospital, France. Between 2003 and 2011, we analysed 200 C. krusei isolates from 130 patients. The isolates were mainly collected in intensive care, cardio‐thoracic and cancer/haematology units. Minimum inhibitory concentrations (MICs) were determined by the E‐test method. The modal MIC was 0.5 μg ml−1; the MIC50 and MIC90 (MICs encompassing 50% and 90% of all isolates tested, respectively) were 0.5 μg ml−1 and 1 μg ml−1. The Cuzick’s and Kendall’s tests showed a significant increase in MIC values between 2003 and 2011 (P = 0.001 and P ≤ 0.001, respectively), regardless of age or gender. No statistical difference was reached with these tests when the first 100 or 50 data were excluded. Despite the increase observed in the first period of the study, our results confirm the low AmB MICs reported in previous studies. However, some authors have recently reported much higher MICs. This discrepancy cannot be explained by method biases and could reflect C. krusei epidemiological differences among populations.

Inflammation is the main risk factor of voriconazole overdose in hematological patients

Aim: Voriconazole (VRC) overdoses are frequent and expose patients at high risk of adverse effects. This case-control study performed in hematological patients who benefited from VRC therapeutic drug monitoring from January 2012 to December 2015 aimed to identify risk factors of VRC overdose. Methods: Pharmacogenetic, biological, and demographic parameters at the time of VRC trough concentration (Cmin) were retrospectively collected from medical records. Cases (VRC overdose: defined by a VRC Cmin ≥ 4 mg/l; n = 31) were compared to controls (no VRC overdose: defined by VRC Cmin < 4 mg/L; n = 31) using non-parametric or Chi-square tests followed by multivariable analysis. Results: VRC overdoses were significantly associated with high CRP and bilirubin levels, intra-venous administration, and age in univariable analysis. In contrast, the proportion of CYP genotypes (CYP2C19, CYP3A4, or CYP3A5, considered alone or combined in a genetic score1) were not significantly different between patients who experienced a VRC overdose and those who did not. In multivariable analysis, the class of CRP level (defined by median CRP levels of 96 mg/l) was the sole independent risk factor of VRC overdose (p < 0.01). Patients with CRP levels > 96 mg/l had a 27-fold (IC 95%: [6-106]) higher risk of VRC overdose than patients with CRP levels ≤ 96 mg/l. Conclusion: This study demonstrates that inflammatory status, assessed by CRP levels, is the main risk factor of VRC overdose in French hematological patients, whereas pharmacogenetic determinants do not appear to be involved.

IgG1 anti-cell wall and IgG2 anti-phosphopeptidomannan antibodies in the diagnosis of invasive candidiasis and heavy Candida colonization

We conducted a retrospective study to evaluate the usefulness of immunoglobulin G (IgG) subclasses against Candida cell wall fragments (CW) and phosphopeptidomannan (PPM) for the diagnosis of invasive candidiasis (IC). We analyzed 54 patients with IC (n = 19), Candida heavy colonization (HC; n = 16), and controls (no IC or HC, n = 19).In nonneutropenic patients (n = 47), the sensitivity and specificity values of IgG1 anti-CW and IgG2 anti-PPM in IC were 88%, 59%, and 88%, 94%, respectively. The areas under the receiver operating characteristic curves were 0.69 (0.51-0.88) and 0.901 (0.78-1.02), respectively. IgG1 mean values (arbitrary units) and 95% confidence interval were 46 (20-71), 42 (-0.38 to 84) and 20 (8.3-32) in IC, HC, and in controls, respectively, and discriminated IC but not HC from controls (P = .032, and P = .77, respectively). IgG2 mean values were 26 (9.2-42), 19 (4.4-33), and 3.2 (0.28-6.6) in IC, HC, and in controls, respectively, and discriminated both IC and HC from controls (P < .0001 and P = .035, respectively) but did not separate IC from HC (P = .2). IgG2 showed positivity as early as one day after the IC diagnosis. Antibodies were detected in only two out of a total of seven neutropenic patients.For both IC and HC patients, the diagnostic performance of IgG2 anti-PPM was better than the one of IgG1 anti-CW. In nonneutropenic patients, IgG2 anti-PPM accurately identified not only IC patients but also HC patients at high risk for IC. This marker may help clinicians in the initiation of early preemptive therapy.