A large outbreak due to Vancomycin Resistant Enterococci: Description of clonal features and management strategies
A large outbreak with different clones of vancomycin-resistant enterococci (VRE) affected the Bern University Hospital for several months. The aim of this study was to describe the extent of the outbreak and the infection control measures implemented to control transmission.
Triggered by two cases of VRE bloodstream infections on our hemato-oncologic ward, an outbreak investigation was started. Microbiological diagnosis of VRE was obtained by culture. Epidemiological links were assessed by meticulous chart review and supplemented with whole genome sequencing (WGS) analyses. Multiple infection control measures were implemented to avoid further transmissions.
Between 29 December 2017 and 21 April 2018, 2877 screening samples were obtained from 1200 patients. 83 patients (6.9%) were found to be colonized with VR Enterococcus faecium. Of those, 76 (91.6%) had a strain carrying vanB, with 70 (84%) isolates virtually identical by cgMLST. The remaining eight patients (9.6%) were colonized with vanA carrying strains belonging to five different STs. Five patients (7%) developed an invasive infection (four bloodstream infections, one abdominal abscess). In order to control the outbreak, a VRE task force was formed and extensive infection control measures were implemented: temporary admission stop, patient isolation and cohorting, staff cohorting, active contact tracing and targeted screening, ward screening of “at risk” wards (irrespective of exposure), reinforcement of hand hygiene compliance, intensified environmental cleaning and proactive communication. By April 2018 the outbreak was contained with these specific measures, which required a substantial effort from the infection prevention team and the involved healthcare workers.
This VRE outbreak was characterized by a rapid spread of the causative clone. Epidemiological analyses supplemented by WGS typing results were essential for the reconstruction of transmission pathways and to guide contact tracing. A multi-faceted infection control led to the containment of the outbreak.
In vitro metabolomic studies on the Echinococcus multilocularis metacestode
The metacestode (larval stage) of the tapeworm Echinococcus multilocularis is the causative agent of alveolar echinococcosis (AE), which is a severe and in many cases incurable disease in humans and other mammals. Livelong benzimidazole chemotherapy is often the only option for AE patients. Benzimidazoles can cause substantial side effects and therefore novel therapeutic treatment strategies are urgently needed.
We follow the strategy to discover new treatment options against AE by investigating the host-dependent nutritional requirements of the parasite. Like this, specific factors could be targeted to starve the parasite within the host.
We applied metabolomic profiling by 1H Nuclear Magnetic Resonance (NMR) spectroscopy to investigate the nutritional requirements of the E. multilocularis metacestode in an in vitro model.
Of the unambiguously detected metabolites, six were significantly consumed from the medium and thirteen released into the medium. We confirmed the above-described NMR results by amino acid quantification assays, and specific measurements of the energy metabolism and metabolic products. Several released metabolites are involved in the anaerobic malate dismutation pathway, which could offer a potential drug target, as it is not found in mammals. Among the most consumed metabolites was the amino acid threonine. Metacestode growth in vitro was accelerated in L-threonine enriched medium. Currently, we investigate the threonine pathways of the E. multilocularis metacestode to specifically identify key steps in threonine consumption. Preliminary experiments showed increased parasite respiration upon addition of threonine to E. multilocularis metacestodes, which indicates that threonine could be used as a substrate for the energy metabolism of Echinococcus. This could offer new innovative ways for starving the parasite in the future.
Pentraxin-3 Polymorphisms and Invasive Mold Infections in Patients Undergoing Myeloablative Chemotherapy for Acute Leukemia
Aims: Single nucleotide polymorphisms (SNPs) in Pentraxin-3 (PTX3) have been associated with the risk of invasive mold infections (IMIs) in hematopoietic cells (HCT) and solid organ transplant recipients. However, similar associations have not been reported in acute leukemia (AL) patients. The aim of this study was to analyze the role of two PTX3 SNPs on susceptibility for IMIs in AL patients undergoing myeloablative chemotherapy.
Methods: All adult patients hospitalized in the isolation Unit of the Lausanne University Hospital for acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), or refractory anemia with excess blast-2 (RAEB-2) between 2007 and 2017 who signed an informed consent for genetic biobanking were included. Clinical data were prospectively collected during complete chemotherapy courses, associating induction, consolidation and potential reinduction cycles for up to 6 months. The association between IMI and PTX3 SNPs was assessed by cumulative incidence (CI) curves and uni/multivariable Cox regression models with censoring on last hospitalization date or upon HCT administration, considering death as a competing risk.
Results: Among 185 patients included, 172 completed a single chemotherapy course (N = 172) and 20 two chemotherapy courses (N = 40), for a total of 212 chemotherapy courses (150 in AML, 45 in ALL and 17 in RAEB-2 patients). A total of 26 IMIs (10 proven and 16 probable) occurred. Homozygosity for rs2305619 and/or rs3816527 was associated with an increased risk of IMI (CI = 21% versus 10%, P = 0.04). After stratification according to the absolute neutrophils count (ANC) at leukemia presentation, the association was significant in patients without pre-existing neutropenia (ANC ≥ 500/mm3, CI = 27% versus 6%, P = 0.006) but not in the other (ANC < 500/mm3, P = 0.6). PTX3 SNPs were independent risk factors for IMIs (HR = 5.06, 95% confidence interval 1.68-15.2, P = 0.004) in the non-neutropenic group. These associations remained significant in multivariable analyses (hazard ratio = 4.0, 95% confidence interval 1.55-10.3, P = 0.004) adjusted for age, sex and type of hematological malignancy.
Conclusions: The new evidence for a robust association between PTX3 SNPs and IMIs among AL patients undergoing myeloablative chemotherapy makes PTX3 one of the most promising markers for novel management strategies, especially in this population which may be particularly well suited for genetically-targeted antifungal prophylaxis.
Validation of the Management of Aortic Graft Infection Collaboration criteria for the diagnosis of vascular graft infection: results from the prospective Vascular graft cohort study
Aims: There is still a lack of evidence-based, validated diagnostic criteria for vascular graft infections (VGI), although VGIs cause major morbidity, mortality and economic costs. Recently, a formal case definition has been issued by the multidisciplinary Management of Aortic Graft Infection Collaboration (MAGIC)1. We aimed to validate the proposed diagnostic standard for VGI in the prospective Vascular Graft Infection Cohort Study (VASGRA).
Methods: We investigated participants with VGI, suspicion of VGI and contemporary control patients after vascular graft implantation. Using the MAGIC criteria for evaluation, VGI were defined as diagnosed (two major criteria; one major criterion and one minor criterion from a different category), suspected (one isolated major criterion; at least two minor criteria from different categories) or rejected (no criterion, one minor criterion,). We assessed the diagnostic accuracy of the MAGIC criteria by calculating sensitivity, specificity and the positive (PPV) and negative predictive value (NPV).
Results: We analyzed 202 predominantly male (85%) VASGRA participants with a median age of 68 years (Interquartile range [IQR] 58.5-75). Thereof, VGI was diagnosed by our multidisciplinary group in 103 patients (101 definite, 2 possible) and rejected by consensus in 34 patients. As a control group, we included 65 VASGRA participants undergoing routine vascular surgery. Assuming that patients with possible VGI according to our judgement and suspected VGI according to the MAGIC criteria were actually diseased, a sensitivity of 100% (95% Confidence Interval (C.I.) 97-100%), a specificity of 61% (95% C.I. 50-70%), a PPV 73% (95% C.I. 64-78%), and a NPV 100% (95% C.I. 94-100%) was reached. Counting possible VGI (our rating) and suspected VGI according to the MAGIC criteria as not diseased, the diagnostic accuracy improved with a sensitivity of 95% (95% Confidence Interval [C.I.] 89-98%), a specificity of 92% (95% C.I. 84-96%), a PPV 91% (95% C.I. 84-96%), and a NPV 95% (95% C.I. 88-98%).
Conclusion: The current MAGIC criteria offer a high sensitivity and good specificity for the diagnosis of VGI, and therefore they may be safely used for VGI case definition. Considering some shortcomings of the MAGIC scheme (lack of specificity for possible VGI; lack of consideration of histopathology, molecular diagnostics and serology; rating of PET/CT as minor radiologic criterion) further modifications will be needed in the near future.
Long-term immune response to yellow fever vaccination in HIV-infected individuals depends on HIV-RNA suppression status: Implications for vaccination schedule
Background. In human immunodeficiency virus (HIV)–infected individuals, the immune response over time to yellow fever
vaccination (YFV) and the necessity for booster vaccination are not well understood.
Methods. We studied 247 participants of the Swiss HIV Cohort Study (SHCS) with a first YFV after HIV diagnosis and determined their immune responses at 1 year, 5 years, and 10 years postvaccination by yellow fever plaque reduction neutralization titers (PRNTs) in stored blood samples. A PRNT of 1: ≥ 10 was regarded as reactive and protective. Predictors of vaccination response were analyzed with Poisson regression.
Results. At vaccination, 82% of the vaccinees were taking combination antiretroviral therapy (cART), 83% had suppressed HIV RNA levels ( < 400 copies/mL), and their median CD4 T-cell count was 536 cells/μL. PRNT was reactive in 46% (95% confidence interval [CI], 38%–53%) before, 95% (95% CI, 91%–98%) within 1 year, 86% (95% CI, 79%–92%) at 5 years, and 75% (95% CI, 62%–85%) at 10 years postvaccination. In those with suppressed plasma HIV RNA at YFV, the proportion with reactive PRNTs remained high: 99% (95% CI, 95%–99.8%) within 1 year, 99% (95% CI, 92%–100%) at 5 years, and 100% (95% CI, 86%–100%) at 10 years.
Conclusions. HIV-infected patients’ long-term immune response up to 10 years to YFV is primarily dependent on the control
of HIV replication at the time of vaccination. For those on successful cART, immune response up to 10 years is comparable to that of non-HIV-infected adults. We recommend a single YFV booster after 10 years for patients vaccinated on successful cART, whereas those vaccinated with uncontrolled HIV RNA may need an early booster.
Pseudo-outbreak of a carbapenem-resistant Klebsiella pneumoniae on an intensive care unit potentially linked to a contaminated bronchoscope
Aim. We report the detection and management of a pseudo-outbreak of carbapenem-resistant Klebsiella pneumoniae (CR-KP) in our surgical intensive care unit (SICU).
Methods. A patient colonized with a CR-KP (index patient) was hospitalized and contact-isolated in our SICU in December 2017. Upon detection of the same CR-KP in a respiratory sample of another SICU patient, an outbreak investigation was started. This included screening for and isolation of secondary cases, observation of work processes, reinforcement of hygiene measures, generation of transmission hypotheses, and environmental screening. Case patients were screened at other body sites for CR-KP colonization and whole-genome sequencing was used to characterize and to compare isolates.
Results. We detected CR-KP in 3 secondary SICU patients with the identical resistance profile as the index case, all in broncho-alveolar lavage (BAL) specimens collected with the same bronchoscope. Multiple rectal, urine (in case of urinary catheter), and follow-up sputum/BAL samples remained negative for all 3 patients; no infections developed. Using whole-genome sequencing, a K. pneumoniae multi-locus sequence type (MLST) 37 harbouring the TEM1B β-lactamase, 2 extended-spectrum β-lactamases (OXA-1, SHV-11), and an AmpC β-lactamase (DHA-7) was detected; cgMLST analysis showed identical isolates without allelic differences. Several steps in the cleaning process of the bronchoscope with potential for improvement were identified, such as suboptimal pre-treatment of instruments; high humidity in the storage room; or spatially close stowing of clean and used bronchoscopes. However, swabs taken from the orifices of the incriminated bronchoscope and cultures of rinsing liquid of both the working and suction channels remained negative, even after sonification was performed. After removal of the instrument, no new cases were detected beyond February 2017.
Discussion. A pseudo-outbreak of a CR-KP in our SICU involving the index and three additional cases was potentially related to a contaminated bronchoscope. Outbreak investigations revealed several breaches of hygiene in the cleaning process and storage of the bronchoscopes, which were addressed and resolved.
Point prevalence study of antibiotic appropriateness in a Swiss University Hospital to tailor antibiotic stewardship interventions.
Background: The emergence and spread of bacterial resistance are an increasing public health concern. Epidemiological studies support a close correlation between consumption and bacterial resistance. Proactive antibiotic stewardship interventions are limited in Switzerland, although inappropriate antimicrobial therapies have been reported. In our hospital, we have limited data on appropriateness of antibiotic prescriptions. An evaluation was needed to tailor interventions.
Methods: We conducted a point prevalence study in 31 acute medical and surgical units. Intermediate and intensive care units were excluded. All hospitalised patients receiving antibiotics (treatment or prophylaxis) on the day of evaluation were included. Antibiotic appropriateness, including indication, duration, route of administration, spectrum and dosing was evaluated by an infectious disease specialist based on patients’ charts, clinical data, microbiological documentation, local antibiotic guidelines and expert opinion.
Results: A total of 564 patients were reviewed, of whom 186 (33%) received one or more topic or systemic antibiotics: 64 (34%) as a prophylaxis and 122 (66%) as a treatment +/- prophylaxis. Among these latter, 90 (75%) were on iv +/- another route of administration, 30 (25%) on oral route only and 2 (1%) on another route only. 69 (58%) presented at least one episode of microbiologically documented infection and 56 (45%) were followed at least once by the infectious diseases team. 70 patients (58%) presented at least one opportunity of treatment adaptation: 23 (19%) were eligible for treatment interruption, 14 (11%) had no indication for at least one prescribed antibiotic, 12 (10%) could benefit from an iv to oral switch, 10 (8%) could have an adaptation of spectrum, 10 (8%) needed a dosing adaptation and 2 (2%) received antibiotics on an inappropriate oral route. Of patients receiving prophylaxis, 37/64 (58%) had no longer indications for antibiotics.
Conclusion: In our hospital, 58% of patient receiving antibiotics on a given day presented at least one opportunity of treatment adaptation, although most of them had been evaluated by the infectious diseases team during the days before. These results highlight the need of joint efforts together with a dedicated antimicrobial stewardship team to improve antibiotics prescriptions and increase prescribers’ awareness of necessary daily reassessment of antimicrobial therapies.
A simplified dolutegravir monotherapy is non-inferior compared to cART in patients with early ART: a randomized controlled trial
Background: Patients (pt) who started combination antiretroviral therapy (cART) during acute/recent HIV-1 infection show a smaller HIV-1 reservoir size compared to pt who started cART during chronic infection. Thus, we hypothesized that a smaller HIV-1 reservoir size translates in sustained virological suppression after simplification of cART to dolutegravir (DTG) monotherapy (MT).
Methods: In this randomized, open label, non-inferiority trial, we recruited pt > 18 yr with documented primary HIV-1 infection (PHI) who started cART < 180 days after estimated date of infection (EDI), who were fully suppressed for > 48wks (< 50 cp/mL plasma). Exclusion criteria were previous virological failure (VF) or treatment interruption and major resistance associated mutations (RAM) to integrase inhibitors. We randomly assigned patients in a 2:1 ratio to MT with DTG 50 mg once daily or to continuation of standard cART. The primary endpoint was virological response, defined as HIV-1 RNA < 50 cp/mL plasma at wk 48, in the per-protocol (PP) population, with a non-inferiority margin of 10% (ClinicalTrials.gov, NCT02551523).
Results: Between 11/2015-3/2017, we randomly assigned 101 patients (68 to simplification to DTG MT, 33 to continuation of cART). Median age was 42 yr and 83% were MSM. At week 48 in the PP population, 67/67 (100%) had virological response in the DTG MT group versus 31/31 (100%) in the cART group (difference 0%, 95%-CI [-1, 0.047)], showing non-inferiority at the prespecified level. In the intention-to-treat population, 1 pt in the DTG MT group experienced viral failure at week 36 (viral load 382 cp/mL) and 2 patients in the cART group left the study before wk 48 because they moved abroad. The pt in the MT group who experienced viral failure was found to be chronically infected at the start of first cART and therefore violated entry criteria. Resistance test at time of viral failure revealed no RAMs and he was re-suppressed on cART. Overall, 14 severe adverse events occurred (DTG MT 10 [15%]; cART 4 [12%]), none related to study drugs.
Conclusion: In our randomized simplification trial, MT with once daily DTG was effective, safe, and non-inferior to cART in pt with a documented PHI who initiated cART < 180 days after EDI and were virologically suppressed for at least 48 weeks. Our results suggest that future simplification studies should use a stratification according to time of infection at start of first cART.
A five-day course of oral antibiotics followed by feacal transplantation to eradicate carriage of multidrug-resistant Enterobacteriaceae: A Randomized Clinical Trial
Background: Fecal microbiota transplantation (FMT) has been suggested to eradicate intestinal carriage with Extended spectrum beta-lactamase (ESBL-E) and carbapenemase-producing Enterobacteriaceae (CPE).
Materials/methods: This publically funded, multicenter (Geneva [G], Paris [P], Utrecht [U], Tel Aviv [T]) open-label, randomized trial examined whether a 5-day course of oral antibiotics followed by FMT is superior to no intervention for the eradication of intestinal carriage of ESBL-E and/or CPE. Adult, immunocompetent patients with ESBL-E or CPE carriage were eligible. Patients were randomized 1:1 to either 5 days of colistin sulphate 2 million IU per os 4x/day & Neomycin sulphate 500 mg (salt) per os 4x/day for 5 days followed by FMT (either by administration of 15 FMT capsules on two consecutive days [G, P] or by a single administration of 80ml of faecal material via nasogastric tube [U,T]). FMT was obtained from healthy, unrelated donors [G, P, U] or from a stool bank [T]. Stool cultures for ESBL-E/CPE carriage were obtained 8-15 days [V2], 16-28 days [V3], 35-48 days [V4, primary outcome] and 5-7 months [V5] after randomization. The targeted sample size was 16 patients per centre. The primary analysis was “intention-to-treat”. ClincalTrials.gov NCT02472600.
Results: Between 02/2016-06/2017 39 patients were randomized [G=14; P=16; U=7; T=2], 22 to the intervention (21/22 underwent the intervention) and 17 to the control. Recruitment stopped in 06/2017 due to lack of further funding. Median age was 65 years (range 22-89), 20 were female, 36 of 39 patients were colonized with ESBL-E and 11 with CPE (8 had both). Of the 22 patients in the intervention arm, 9 (41%) were negative for ESBL-E/CPE at V4. In the control arm, 4 patients were negative (24%), 12 positive and 1 was lost to follow-up (imputed as negative). The intervention had no significant effect (OR for decolonization; 2.0 [95%CI 0.5-7.6]). Study drugs were overall well tolerated.
Conclusions: While this clinical trial failed to achieve its targeted sample size, the results suggest only a small effect of oral non-absorbable antibiotics followed by FMT on the eradication of intestinal carriage of ESBL-E and/or CPE.
Patient selection for blood culture sampling: A prospective cohort study to compare a prediction score including biomarkers with non-standardized clinical judgement
Aim: Although still established as gold standard for detection of bacteremia, blood cultures (BC) have a low diagnostic yield resulting in high diagnostic costs. Guidelines for effective BC sampling (BCS) are lacking. A retrospective study reported a high predictive value for BC positivity when different clinical parameters included in the “Shapiro score” were combined with procalcitonin (PCT).
Methods: This single-center, prospective cohort study was designed to validate the SPA criteria (i.e., “Shapiro score” ≥ 3 points combined with PCT > 0.25µg/l) for positive BC prediction in patients with a systemic inflammatory response syndrome (SIRS) and suspected infection at the emergency ward. Predefined overruling criteria for BCS were severe sepsis, immunosuppression, suspected endocarditis or meningitis. The algorithm was tested against individual doctors’ decisions for BCS. We calculated logistic regression analysis with odds ratios (OR) and area under the receiver operating characteristic curve (AUROC) to study associations of predictors and positive BC.
Results: The overall population in this study included 1438 patients with routine BCS, the diagnostic yield for positive BC was 15% (n=215). In patients with positive SPA criteria, the diagnostic yield of BC increased to 31% (173/555), corresponding to an OR of 9.07 (95%CI 6.34-12.97). In patients with either positive SPA or overruling criteria (n=749) the number of positive BC increased to 194 with a slight reduction in diagnostic yield (26%), corresponding to an OR of 11.12 (95%CI 6.99-17.69). In contrast, the diagnostic yield for positive BC in patients without SPA or overruling criteria was only 3% (21/689) with an OR of 0.09 (95%CI 0.06-0.14). Sensitivity, specificity and AUROC were as follows: SIRS (93%, 35%, 0.640), SPA or overruling (81%, 69%, 0.746), PCT > 0.25µg/l (91%, 42%, 0.667), Shapiro ≥ 3 (91%, 40%, 0.657). In the 21 positive BC missed by the algorithm, two had a symptom onset < 24 h and therefore still low PCT, six pathogens were detected alternatively (3 urine, 1 synovial fluid, 1 vertebral tissue culture, 1 urine pneumococcal antigen test).
Conclusions: Our data validated the high prognostic value of SPA combined with overruling criteria for accurate prediction of positive BC with an increase in diagnostic yield from 15% to 26% compared to usual care. The algorithm allows to reduce BC by 48% while still detecting 202/215 organisms (94%), representing a novel diagnostic stewardship tool.
Antimicrobial resistance of Salmonella isolated from wild animals and reared grasscutters used as animal source food in Côte d’Ivoire (West Africa)
Aims: The study aims to contribute to the surveillance of food safety and of antimicrobial resistance (AMR) in wildlife and animal source food by assessing the AMR prevalence of Salmonella spp isolated from wild animals and reared grasscutters and by estimating the frequency of risky consumption.
Methods: The study was conducted in Abidjan, located in South-Côte d’Ivoire (CI). We collected swabs and feces among reared grascutters in farms and swabs of dead wild animals sold in markets. Overall 425 samples have been analyzed for Salmonella spp isolation according the protocol ISO 6579. Biochemical and agglutination tests have been used for confirmation and determination of serogroups of strains isolated and then the susceptibility of Salmonella spp to thirteen antimicrobial agents was determined using the Kirby-Baüer method. In addition, consumption interviews were conducted with hunters, butchers, vendors of wildlife and households in Toumodi (south-central of CI), one of the greatest hub in wild animal meat (bushmeat) traffic.
Results: Salmonella were isolated from 14.6% (CI 95 11.4-18.3) of samples with a significant higher frequency in wild animals (22.4%) compared to the reared grasscutters (7%) (OR = 3.8, CI 95 2 – 7.6). The isolated strains belonged to the serogroups B (39.4%), E or G (29.6%), D (16.9%) and C (14.1%). Resistance to at least on antibiotic agent was found in 63% (CI 95 50.2 - 74.7) of strains. Multi-resistances went up to seven antibiotics. No significant difference of AMR was observed between wild animal and reared animals. The highest rates of resistance are observed for Ciprofloxacin (53.8%), Tobramycin (43%), Gentamicin (23%) and Ceftriaxone (13.8%), whereas we found 100% of sensitivity of strains for the following agents: Amoxicilin + Clavulanic acid, Trimethoprim + Sulfamethoxazole and Chloramphenicol. Data analyses are ongoing.
Preliminary conclusion: People handling wild animals are seemingly 4 times more exposed to Salmonella than those in contact with farmed grasscutters. Infrequent use of antibiotics in grasscutter farms likely explains the limited resistances among reared animals. However, the relative risk for human health exists with regard to high resistances observed for some antibiotics, particularly of Ciprofloxacin which is often used for treating human enteric fever.
Key words: wild animal, grasscutter, antimicrobial resistance, food
Immunogenicity and Safety of Seasonal Influenza Vaccine with Topical Imiquimod in Immunocompromised Patients: a Randomized Controlled Pilot Trial
Aim: Application of a cream containing the Toll-like receptor 7 agonist imiquimod before intradermal injection of the influenza vaccine has been shown to significantly increase the vaccine immunogenicity in the immunocompetent host, but the effect in immunocompromised patients is unknown. We aimed to assess the impact of an imiquimod-based vaccination strategy in immunocompromised patients and the effect of imiquimod before intramuscular vaccination.
Methods: Kidney transplant recipients (KT) and HIV infected (HIV+) patients were randomized to receive the intramuscular vaccine alone (IM) or, respectively, the intramuscular (IMI-IM) or the intradermal (IMI-ID) vaccine after topical imiquimod. Anti-influenza antibody titers were measured by hemagglutination inhibition assay before and 3 and 24 weeks after vaccination. Vaccine response was defined as seroconversion (4-fold increase in antibody titer) to at least one viral strain 3 weeks after vaccination, and seroprotection as a titer ≥ 1:40. Predictors of vaccine response were analyzed by logistic regression. Participants were followed for 6 months.
Results: Seventy patients (35 KT and 35 HIV+) received the IM (24), the IMI-IM (22) or the IMI-ID (24) vaccine. Baseline characteristics were comparables between groups. Fourteen (61 %) patients in the IM group, 13 (59 %) in the IMI-IM and 15 (65 %) in the IMI-ID group responded to the vaccine (P = 0.909). Vaccine response was significantly better in HIV+ when compared to KT, regardless of the imiquimod application or route of injection (OR 3.7, 95 % CI (1.3 – 10.3), P = 0.015). After vaccination the majority of the patients were seroprotected to all 3 viral strains without differences between groups (19 / 24 (78 %), 15 / 22 (68 %) and 16 / 23 (70 %) in the IM, IMI-IM and IMI-ID groups, P = 0.657). We did not observe any vaccine-related severe adverse event or episode of acute rejection during the study period. One KT and 1 HIV+ patient (1 in the IMI-ID and 1 in the IMI-IM group) developed laboratory confirmed influenza.
Conclusions: Although safe and well tolerated, topical imiquimod before intradermal or intramuscular injection did not improve the immunogenicity of influenza vaccine in KT and in HIV+. Thus, application of topical imiquimod seems not to be an appropriate strategy to improve the immunogenicity of the influenza vaccine in immunocompromised patients.