Risk of mortality in people with chronic liver diseases hospitalized for Coronavirus disease of 2019 (COVID-19) in a tertiary hospital in Lombardy, Italy.

The impact of severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) infection on patients with pre-existing chronic liver diseases (CLD) remains elusive. The aim of this study was to investigate the in-hospital mortality in patients hospitalized for Coronavirus disease of 2019 (COVID-19) with CLD (CLD group) compared to those without CLD (non-CLD group). We performed a retrospective cohort study including patients with confirmed SARS-CoV-2 infection, hospitalized at San Raffaele Hospital (Milan), stratified according to the presence or absence of CLD. A propensity score was estimated and used to match the two groups by age, gender, body mass index, type 2 diabetes mellitus, and hypertension. Predictors of mortality were assessed using univariate and multivariate logistic regression model. Among 1210 patients with COVID-19, 41 (3.4%) were included in the CLD group and 1169 (96.6%) in the non-CLD group. Using a propensity score, we matched 41 patients in the CLD group with 123 in the non-CLD group. At admission, patients in the CLD group had worse liver function, lower platelets count, and lower c-reactive protein levels. By multivariate analysis, the CLD group showed a higher risk of death: OR 4.04 (95% CI 1.29-12.70; p= 0.017). Our study showed that COVID-19 with chronic liver diseases has a higher risk of mortality during hospitalization.

Detection of low-level HCV variants in DAA treated patients: comparison amongst three different NGS data analysis protocols.

BACKGROUND: Notwithstanding the efforts of direct-acting antivirals (DAAs) for the treatment of chronically infected hepatitis C virus (HCV) patients, concerns exist regarding the emergence of resistance-associated substitutions (RAS) related to therapy failure. Sanger sequencing is still the reference technique used for the detection of RAS and it detects viral variants present up to 15%, meaning that minority variants are undetectable, using this technique. To date, many studies are focused on the analysis of the impact of HCV low variants using next-generation sequencing (NGS) techniques, but the importance of these minority variants is still debated, and importantly, a common data analysis method is still not defined. METHODS: Serum samples from four patients failing DAAs therapy were collected at baseline and failure, and amplification of NS3, NS5A and NS5B genes was performed on each sample. The genes amplified were sequenced using Sanger and NGS Illumina sequencing and the data generated were analyzed with different approaches. Three different NGS data analysis methods, two homemade in silico pipeline and one commercially available certified user-friendly software, were used to detect low-level variants. RESULTS: The NGS approach allowed to infer also very-low level virus variants. Moreover, data processing allowed to generate high accuracy data which results in reduction in the error rates for each single sequence polymorphism. The results improved the detection of low-level viral variants in the HCV quasispecies of the analyzed patients, and in one patient a low-level RAS related to treatment failure was identified. Importantly, the results obtained from only two out of the three data analysis strategies were in complete agreement in terms of both detection and frequency of RAS. CONCLUSIONS: These results highlight the need to find a robust NGS data analysis method to standardize NGS results for a better comprehension of the clinical role of low-level HCV variants. Based on the extreme importance of data analysis approaches for wet-data interpretation, a detailed description of the used pipelines and further standardization of the in silico analysis could allow increasing diagnostic laboratory networking to unleash true potentials of NGS.

Comparison of three therapeutic regimens for genotype-3 hepatitis C virus infection in a large real-life multicentre cohort.

BACKGROUND & AIMS: In the direct-acting antiviral era, treatment of genotype-3 HCV (HCV-GT3) is still challenging. Real-life comparisons between recommended regimens, sofosbuvir (SOF)+daclatasvir (DAC), SOF/velpatasvir (VEL), glecaprevir/pibrentasvir (GLE/PIB), are scarce. We aimed at filling this data gap. METHODS: Sustained virological response 12 weeks after treatment completion (SVR12) was assessed for all HCV-GT3 patients consecutively treated within the Lombardia web-based Navigatore HCV-Network; differences in SVR12 across regimens were evaluated by logistic regression. RESULTS: Of the 2082 subjects with HCV-GT3, 1544 were evaluable for comparisons between regimens: SOF + DAC (1023, 66.2%), SOF/VEL (369, 23.9%), GLE/PIB (152, 9.8%). Patients treated with former regimens were more frequently male, cirrhotic, HIV-positive, pretreated, used ribavirin in their regimen, and had lower baseline HCV-RNA. SVR12 was similar across groups: 94.8% in SOF + DAC, 97.6% in SOF/VEL, 96.7% in GLE/PIB (P = .065). At univariate analysis, SVR12 was associated with female gender (97.9% vs 94.8%, P = .007) and lower median pretreatment Log10 HCV-RNA (5.87 vs 6.20, P = .001). At multivariate logistic regression analysis, treatment with SOF/VEL was associated with a higher likelihood of SVR12 than SOF + DAC, but only in the absence of ribavirin (98% vs 90.3%). Female gender and lower pretreatment HCV-RNA were independently associated with SVR12. CONCLUSIONS: In a large real-life setting of HCV-GT3-infected patients with a high proportion of cirrhosis, the success rate was remarkable. The slight advantage of SOF/VEL on SOF + DAC was significant only without ribavirin. The current prescription shift towards novel regimens (ie SOF/VEL and GLE/PIB) in easier-to-treat patients allows ribavirin-free and shorter schedules without mining SVR12 in this <> genotype.

Relapse of Symptomatic Cerebrospinal Fluid HIV Escape.

PURPOSE OF REVIEW: Symptomatic cerebrospinal fluid (CSF) HIV escape defines the presence of neurological disease in combination antiretroviral therapy (cART)-treated persons due to HIV replication in CSF despite systemic suppression or to higher viral replication in CSF than in plasma. The aim was to search for cases of recurrent symptomatic CSF escape and to define their characteristics. RECENT FINDINGS: By review of the literature, we identified symptomatic CSF escape relapses in three patients who had shown clinical remission of a first escape episode following cART optimization. By examination of our cohort of 21 patients with symptomatic CSF escape, we identified five additional patients. In the latter, viral escape relapsed over a median follow-up of 108 months because of low adherence or upon treatment simplification of a previously optimized regimen. cART reoptimization based on resistance profile and potential drug neuropenetration and efficacy led to relapse resolution with no further episodes after a median follow-up of 50 months from relapse. The observation that CSF escape may relapse highlights the importance of long-term neuro-suppressive regimens after a first episode and supports the role of the brain as a reservoir for HIV.

HIV-1 recombinant forms in immigrants regularly residing in Milan, northern Italy.

To assess the HIV -1subtypes distribution in HIV-1 positive migrants living in Milan we studied 77 HIV-1 patients followed at the San Raffaele Hospital of Milan. Twenty subjects were born in Europe, 43 in the Americas, 10 in Africa and 4 in Asia. Unsafe heterosexual activity prevailed in migrants born in Africa and male homosexuality in those born in European, American and Asian countries (p = 0.05). The phylogeny showed that 38/77 (49.3%) subjects carried HIV-B subtype while the remaining strains were classified as not pure HIV-1 B subtypes 13/77 (16.9%) or recombinant forms 26/77 (33.8%). Female gender more frequently showed HIV-1 non-B strains and rarely HIV-1 B subtypes (12/39, 30.8% vs. 3/38, 7.9%, p = 0.02). Transmitted drug resistance was identified in 10/77 (13%) patients predominately with B subtype. Our data underscore a large heterogeneity in HIV-1 subtypes and a large proportion of recombinant forms.

Molecular epidemiology of HIV-1 infection in immigrant population in northern Italy.

Human immunodeficiency virus-1 (HIV-1) is characterised by a vast genetic diversity classified into distinct phylogenetic strains and recombinant forms. We describe the HIV-1 molecular epidemiology and evolution of 129 consecutive HIV-1 positive migrants living in Milan (northern Italy). Polymerase gene sequences of 116 HIV-1 subtype-B positive patients were aligned with HIV-1 reference sequences (https://www.ncbi.nlm.nih.gov/) by using MAFFT alignment and edited by using Bioedit software. A maximum likelihood (ML) phylogenetic tree was performed by MEGA7 and was visualised by using FigTree v1.4.3. Of 129 migrants, 35 were born in Europe (28 in Eastern Europe), 70 in the Americas (67 in South America), 15 in Africa and nine in Asia; 76.4% were men who have sex with men (MSM). The serotype HIV-1-B prevailed (89.9%), followed by -C, -F1, -D and -A. Compared with 116 HIV-B patients, the 13 with HIV-non-B showed lower Nadir of CD4+ cell/mmc (P = 0.043), more frequently had sub Saharan origin (38.5 vs. 1.72%, P = 0.0001) and less frequently were MSM (40 vs. 74.5%, P = 0.02). The ML phylogenetic tree of the 116 HIV-1 subtype-B positive patients showed 13 statistically supported nodes (bootstrap > 70%). Most of the sequences included in these nodes have been isolated from male patients from the Americas and the most common risk factor was MSM. The low number of HIV-1 non-B subtype patients did not allow to perform this analysis. These results suggest a shift of HIV-1 prevention projects' focus and a continuous monitoring of HIV-1 molecular epidemiology among entry populations. Prevention efforts based on HIV molecular epidemiology may improve public health surveillance setting.

Natural polymorphisms in the resistance associated sites of HCV-G1 NS5B domain and correlation with geographic origin of HCV isolates.

BACKGROUND: We evaluated the frequency of naturally occurring resistance associated substitutions (RASs) and their characteristic of polymorphic or non-polymorphic amino acid change to direct acting antivirals (DAAs) in NS5b HCV subtypes 1a and 1b according to different geographic origin of isolates. METHODS: Using a public database we retrieved 738 worldwide NS5b sequences (for which was available the geographic origin) from HCV genotype (G)1 infected patients naive to DAAs. NS5b sequences clustering with G1a were more conserved in regard of RASs than G1b isolates, (14% vs 57% RASs, P < 0.0001). RESULTS: In G1a, RASs were differently distributed between isolates from Europe (24%) and USA, (12%) P = 0.0186. In particular, 421 V associated with resistance to non-nucleoside inhibitor beclabuvir was polymorphic in Europe and USA, being detected in 24% and 11% of sequences, respectively, P = 0.0140. In G1b, RASs were found in 45% of sequences from Europe, in 54% of isolates from USA and in 70% of sequences from Asia (P = 0.0051). The 316 N polymorphism was detected in 54% of Asian isolates and at lower frequency, in 28% of isolates from USA and in 20% of European sequences (P < 0.0001). CONCLUSIONS: In conclusion, a higher prevalence of RASs in G1b respect to G1a was found and a geographical distribution of RASs and polymorphic aa changes was observed in G1a as well in G1b. The clinical and therapeutic impact of the geographic distribution of RASs to polymerase inhibitors remains to be established, particularly in patients with virologic failure to DAAs and/or advanced liver disease.

Ombitasvir/Paritaprevir/Ritonavir and Dasabuvir Combination Treatment in Patients with HIV/HCV Co-Infection: Results of an Italian Compassionate Use Program.

Patients co-infected with hepatitis C virus (HCV) and human immunodeficiency virus (HIV) are at high risk of liver disease progression. We report a favorable safety profile and SVR12 rates of 96.7% among HIV/HCV co-infected patients participating in an Italian compassionate-use program of ombitasvir/paritaprevir/ritonavir + dasabuvir (OBV/PTV/r + DSV) ± ribavirin (RBV).

Pharmacokinetics of dolutegravir and rilpivirine in combination with simeprevir and sofosbuvir in HIV/hepatitis C virus-coinfected patients with liver cirrhosis.

Objectives: To evaluate the plasma trough concentrations ( C trough ) of dolutegravir and rilpivirine used in combination with simeprevir and sofosbuvir in HIV/hepatitis C virus (HCV)-coinfected patients with liver cirrhosis. Virological efficacy and safety of both ART and anti-HCV therapy were assessed. Patients and methods: A prospective observational study in HIV/HCV-coinfected patients with liver cirrhosis on ART with dolutegravir plus rilpivirine and treated with simeprevir plus sofosbuvir (±ribavirin) was conducted. Dolutegravir, rilpivirine, GS-331007 (sofosbuvir metabolite) and simeprevir C trough were evaluated with a validated HPLC method at anti-HCV treatment baseline and weeks 2 and 4. Geometric means were calculated to summarize C trough values. Results: Twelve patients were evaluated: 75% were males and the median (IQR) age was 53 (53-55) years. All patients were Child-Pugh stage A, except one who was stage B. The geometric mean (95% CI) of C trough of rilpivirine and dolutegravir did not change between baseline and week 4 ( P = 0.654 and P = 0.268, respectively), with corresponding overall values of 135 (102-177) and 1357 (970-1897) ng/mL. The overall geometric mean (95% CI) of GS-331007 and simeprevir C trough was 370 (268-512) and 2537 (1569-4101) ng/mL, respectively, without significant variation between weeks 2 and 4 ( P = 0.643 and P = 0.179, respectively). All patients completed anti-HCV treatment, achieving sustained virological response. All but two patients maintained undetectable HIV-RNA up to post-treatment week 24. Conclusions: Dolutegravir and rilpivirine C trough appeared not to be affected by concomitant treatment with simeprevir plus sofosbuvir in these HIV/HCV-coinfected patients with liver cirrhosis, supporting the use of this antiretroviral regimen in this setting.

Systemic Inflammatory Response Is a Prognostic Marker in HIV-Infected Patients with Hepatocellular Carcinoma.

BACKGROUND: Hepatocellular carcinoma (HCC) is increasingly prevalent in people living with HIV. Systemic inflammation is a prognostic factor requiring validation in HIV-associated HCC. AIMS: Using a multi-centre database of consecutive HCC cases, we investigated the prognostic role of a panel of inflammatory markers, including neutrophil to lymphocyte ratio (NLR), using univariate and multivariate survival analyses. RESULTS: Fifty-nine patients with HIV-associated HCC secondary to hepatitis C (69%) or B virus infection (32%) were identified. The median survival was 22 months. A raised NLR independently predicted patients' survival and was correlated with advanced Barcelona Clinic Liver Cancer stage (p = 0.003) and poor performance status (p < 0.001) but not with HIV RNA or CD4 counts. CONCLUSION: Systemic inflammation, as measured by NLR, is a prognostic determinant associated with adverse pathological features of malignancy, but not coexisting HIV infection, suggesting a tumour-promoting role of the innate immune response that warrants further investigation in mechanistic studies.

NS3 protease resistance-associated substitutions in liver tissue and plasma samples from patients infected by hepatitis C virus genotype 1A or 1B.

The presence of naturally occurring resistance-associated substitutions (RASs) in the HCV-protease domain has been poorly investigated in the liver, the main site of HCV replication. We evaluated the natural resistance of the virus to NS3 protease inhibitors in liver tissue and plasma samples taken from HCV-infected patients. RASs were investigated by means of viral population sequencing in liver tissue samples from 18 HCV-infected patients harbouring genotype 1a or genotype 1b; plasma samples from 12 of these patients were also available for virological investigation. A discordant genotype was found in two of the 12 patients (16.6%) who provided samples from both compartments. Sequence analysis of the NS3 protease domain showed the presence of RASs in four of the 18 liver tissue samples (22.2%), two of which showed cross-resistance to protease inhibitors in clinical use or phase 2-3 trials. The analysis of the 12 paired tissues and plasma samples excluded the presence of RASs in the plasma compartment. The dominance of discordant genotypes in the paired liver and plasma samples of some HCV-infected patients suggests mixed infection possibly leading to the selective advantage of different genotype in the two compartments. The presence of RASs at intra-hepatic level is not uncommon and may lead to the early emergence of cross-resistant strains.

Dynamic change in the NS3 protease domain in HIV/HCV-coinfected patients naïve to anti-HCV protease inhibitors.

The genetic analysis of THE natural protease inhibitors (PI) resistance of HCV genotype (GT)1 involves subtypes 1a and 1b. NS3 protease domain was sequentially analysed in 10 HIV/HCV GT1-coinfected individuals naïve to HCV treatment. Analysis at different time points showed that 2/3 GT1b patients were infected by a GT1a clade1 during follow-up. In one patient a switch from clade1 to clade2 and in one other patient a switch from clade2 to clade1 was revealed. Four out of ten patients had resistance-associated substitutions (RASs) at baseline. The dynamics of the dominant infecting subtype suggests the presence of mixed infection in some patients.

Diagnostic accuracy of APRI, FIB-4 and Forns for the detection of liver cirrhosis in HIV/HCV-coinfected patients.

We determined the diagnostic accuracy and optimal cut off of three indirect fibrosis biomarkers (APRI, FIB-4, Forns) compared with liver stiffness (LS) for the detection of liver cirrhosis in HIV/HCV-coinfected patients. An observational retrospective study on HIV/HCV-coinfected patients with concomitant LS measurement and APRI, FIB-4 and Forns was performed. The presence of liver cirrhosis was defined as a LS ≥13 KPa. The diagnostic accuracy and optimal cut-off values, compared with LS categorization (<13 vs ≥13 KPa), were determined by receiver operating characteristics (ROC) curves. The study sample included 646 patients. The area-under-the ROC curve (95% confidence interval) for the detection of liver cirrhosis were 0.84 (0.81-0.88), 0.87 (0.84-0.91) and 0.87 (0.84-0.90) for APRI, FIB-4 and Forns, respectively. According to the optimal cut off values for liver cirrhosis (≥0.97 for APRI, ≥2.02 for FIB-4 and ≥7.8 for Forns), 80%, 80% and 82% of subjects were correctly classified by the three indirect fibrosis biomarkers, respectively. Misclassifications were mostly due to false positive cases. The study suggests that indirect fibrosis biomarkers can help clinicians to exclude liver cirrhosis in the management of HIV/HCV co-infected patients, reducing the frequency of more expensive or invasive assessments.

Non-invasive fibrosis biomarkers - APRI and Forns - are associated with liver stiffness in HIV-monoinfected patients receiving antiretroviral drugs.

BACKGROUND & AIMS: HIV-monoinfected patients are susceptible to liver injury by different factors and may develop liver fibrosis, which requires adequate clinical management in terms of therapy and disease monitoring. We aimed to evaluate the presence of liver fibrosis identified by transient elastography (TE), its relationships with indirect biochemical markers [the aspartate aminotransferase/platelet ratio index (APRI), the Forns index and FIB-4] and its predictive factors in HIV-monoinfected patients receiving antiretroviral therapy (ART). METHODS: Seventy-two HIV-monoinfected patients underwent TE and were evaluated using APRI, Forns and FIB-4. The clinical, immunological, virological and other biochemical characteristics were evaluated at the time of TE, together with their history of ART. RESULTS: Seven patients (10%) had liver stiffness (LS) values predicting cirrhosis, and 12 (17%) had values predicting significant or advanced fibrosis. Higher indirect biochemical scores of liver fibrosis were significantly associated with higher LS values [APRI rs  = 0.4296 (P < 0.001); Forns rs  = 0.4754 (P < 0.001); FIB-4 rs  = 0.285 (P = 0.015)]. At multivariable analysis, APRI (ß = 2.7405; P = 0.036), Forns (ß = 1.4174; P = 0.029) and triglyceride levels (ß = 1.3028; P = 0.007) were independently associated with LS. CONCLUSIONS: Indirect fibrosis biomarkers may increase the probability to detect liver injury enhancing a specific diagnostic workup and so contribute to improving the clinical management of HIV-monoinfected patients with clinically suspected liver disease.

Case Report: Lenvatinib for the treatment of recurrent hepatocellular carcinoma in people living with HIV: a report of two cases.

The use and choice of the best systemic treatment is gaining increasing interest in people living with HIV (PLWH) because hepatocellular carcinoma (HCC) represents an increasing cause of morbidity and mortality in this setting and most HCCs are diagnosed in the advanced stage. Ten years ago, the multi-kinase inhibitor lenvatinib was approved in the first-line setting. However, to date, no data on the efficacy and tolerability of lenvatinib in PLWH from clinical trials and real-life studies are available. Case 1 was a gentleman with hepatitis B virus-related cirrhosis who underwent orthotopic liver transplantation for HCC and developed peritoneal metastasis several years later. Lenvatinib treatment was selected at HCC recurrence. This participant maintained undetectable HIV viremia and a relatively preserved immune status during 6 months of systemic treatment with lenvatinib. After 6 months, he discontinued lenvatinib for progression of the disease (growing of peritoneal metastasis) and uncontrolled hypertension. Case 2 was a gentleman with hepatitis C-genotype 1a-related cirrhosis who experienced unresectable recurrences after radiofrequency thermal ablation of the tumor. At the first recurrence, HCC was treated with six cycles of trans-catheter arterial chemoembolization; at the second recurrence, the participant underwent trans-catheter arterial radioembolization; and at the third recurrence, he received lenvatinib. A week after the start of lenvatinib, the participant had liver decompensation and discontinued therapy. The presently reported cases showed low tolerability of systemic therapy with lenvatinib in PLWH. Cumulative data are necessary to define the position of lenvatinib in this setting.

Survival in People Living with HIV with or without Recurrence of Hepatocellular Carcinoma after Invasive Therapy.

BACKGROUND AND AIMS: To address the overall survival (OS) and recurrence (RE) in people living with HIV (PLWH) treated with invasive therapy (IT) for hepatocellular carcinoma (HCC). METHODS: This is a retrospective cohort study on 41 PLWH with HCC receiving IT, defined as liver resection (LR), orthotopic liver transplantation (OLT), radiofrequency thermo-ablation (RFTA) trans arterial chemo, or radioembolization (CRE). OS and RE were investigated by Kaplan-Meier curves. The Cox proportional hazard regression model was used for multivariate analyses. RESULTS: Recurrence occurred in 46.3% PLWH; in 36.7% of participants at 2 years and in 52% at 5 years from HCC diagnosis; it was less frequent in males, p = 0.036. Overall, 2- and 5-year survival after HCC diagnosis was 72% and 48%, respectively. Two-and five-year survival was 100% and 90.9%, respectively, in PLWH receiving OLT, compared to other IT (60.9% and 30.6%, respectively) log-rank p = 0.0006. Two- and five-year survival in participants with no-RE was 70.5% and 54.6%, respectively, and 73.7% and 42.1% among RE, respectively, log-rank p = 0.7772. By multivariate analysis, AFP at values < 28.8 ng/mL, at HCC diagnosis, was the only factor predicting survival. CONCLUSIONS: Fifty percent of PLWH survived five years after HCC diagnosis; 90.9% among OLT patients. Recurrence after IT was observed in 46% of HCC/PLWH. AFP cut-off levels of 28.8 ng/mL were the only independent variable associated with survival.

Treatment success of rescue regimens after dual therapy failure in people living with HIV in a real-life setting.

OBJECTIVES: Few data on management of two-drug regimen (2DR) failure in people living with HIV (PLWH) are available. METHODS: Retrospective study of treatment-experienced PLWH on a 2DR who experienced virological failure (VF) [two consecutive viral loads (VLs) ≥50 copies/mL, single VL ≥1000 copies/mL, or antiretroviral therapy (ART) switch after single VL ≥50 copies/mL with previous blips] or discontinuation for toxicity (baseline). Integrase strand transfer inhibitor (INSTI)-based [one INSTI plus one nucleoside reverse transcriptase inhibitor (NRTI) (n = 78) or one non-NRTI (n = 20)] or boosted protease inhibitor (PI/b)-based [one PI/b plus one NRTI (n = 116) or one INSTI (n = 12)] 2DRs were included. Probabilities of treatment success (TS), VF and discontinuation for any other cause of rescue regimens were estimated by Kaplan-Meier curves. A stepwise Cox model was performed to assess predictors of TS. RESULTS: Overall, 226 PLWH were evaluated: at baseline, 144 individuals discontinued 2DR for toxicity and 82 had VF [median viraemia 81 (63-212) copies/mL]; 171 switched therapy (49.7% to triple regimen, 40.9% to different 2DR), while 55 (exclusively with VF) maintained failing regimens. Probabilities of 12- and 24-month TS were 75.6% and 64.7%, respectively. Higher TS probabilities were observed in individuals who switched ART at 2DR failure (P = 0.003) and PLWH who discontinued 2DR for toxicity (P = 0.008). Therapy switch was the only predictor of TS (P = 0.002). CONCLUSIONS: Overall probability of rescue regimens' TS introduced after 2DR failure is good. Prompt ART switch after 2DR failure is advisable.

Risk of type 2 diabetes among HIV-infected and healthy subjects in Italy.

Type 2 diabetes mellitus is a growing problem in HIV population and a comparison with the general population may help screening and prevention. In this cross-sectional study the authors determined the prevalence of type 2 diabetes mellitus in 4,249 HIV-infected subjects attending the San Raffaele Infectious Diseases Department compared with 9,148 healthy controls recruited in 15 Italian regions, and identified risk factors associated with of type 2 diabetes mellitus. Type 2 diabetes mellitus was defined as reported diabetes, a fasting plasma glucose concentration ≥7.0 mmol/l, or current use of anti-diabetic medication. Prevalence of type 2 diabetes mellitus was higher in HIV-infected than healthy subjects (4.1 vs. 2.5 %; P < 0.0001). At multivariable analysis, HIV-infected subjects (odds ratio 1.70, 95 % CI, 1.12-2.51; P = 0.009), older age (P < 0.0001), higher BMI (P < 0.0001) and hypertension (P = 0.039) were associated with a higher risk of diabetes. Among HIV-infected patients, the risk of type 2 diabetes mellitus increased with older age (P < 0.0001), higher BMI (P = 0.003), higher triglycerides (P = 0.015) lower total cholesterol (P = 0.008), longer duration of HIV infection (P = 0.036) lower nadir CD4 (P = 0.027). Prevalence of type 2 diabetes mellitus in HIV-infected subjects was almost two-fold increased than healthy subjects and it was associated with the typical risk factors of the general population and also to longer duration of HIV infection and lower nadir CD4.