Opportunistic Infections among People Living with HIV (PLHIV) with Diabetes Mellitus (DM) Attending a Tertiary Care Hospital in Coastal City of South India

Background HIV/AIDS and Diabetes Mellitus are the diseases’ known to supress cell mediated immunity and predispose patients for opportunistic infections. Hence, we conducted a study to compare the common opportunistic infections (OIs) between People Living with HIV with DM (PLHIV-DM) and PLHIV without DM (PLHIV). Methodology PLHIV with DM and without DM (1:1) were prospectively included in the study from January 2011 to January 2012 at a tertiary care hospital in Mangalore city. Patients were classified as Diabetic if their fasting plasma glucose was ≥ 7.0mmol/l (126mg/dl) or 2–h plasma glucose was ≥11.1mmol/l (200mg/dl). Standard procedures and techniques were followed for diagnosis of OIs as per WHO guidelines. The data was entered and analyzed using Statistical Package for Social Sciences (SPSS) version 11.5. Findings The study included 37 PLHIV with DM and 37 PLHIV without DM and both groups were treated with Anti-Retroviral Therapy (ART). The median age was 47 years (IQR: 41-55years) for PLHIV-DM as compared to 40 years (IQR: 35–45.5 years) for PLHIV (p<0.0001). PLHIV-DM had median CD4 counts of 245 (IQR: 148–348) cells/μl compared to 150(IQR: 70–278) cells/μl for PLHIV (p = 0.02). Common OIs included oral candidiasis (49% of PLHIV-DM and 35% of PLHIV); Cryptococcal meningitis (19% of PLHIV-DM and 16% of PLHIV); Pneumocystis jiroveci pneumonia (5% of PLHIV-DM and 18% of PLHIV); extra pulmonary tuberculosis (22% of PLHIV-DM and 34.5% of PLHIV); and Cerebral toxoplasmosis (11% of PLHIV–DM and 13.5% of PLHIV). Microbiological testing of samples from PLHIV- DM, C krusei was the most common Candida species isolated from 9 out of 18 samples. Out of six pulmonary TB samples cultured, four grew Non-tuberculosis mycobacteria (NTM) and two Mycobacterium tuberculosis complexes. Conclusions Study did not identify any significant difference in profile of opportunistic infections (OIs) between PLHIV with and without Diabetes.


Findings
The study included 37 PLHIV with DM and 37 PLHIV without DM and both groups were treated with Anti-Retroviral Therapy (ART).

Introduction
India is facing the double burden of communicable and non-communicable diseases, with the third largest number of People Living with HIV in the world. According to the "Joint United Nations programme on AIDS" report 2013, an estimated 210,000 people are infected by HIV in India [1]. Out of 1.21 billion population in the country, an estimated 40.9 million people are affected with Diabetes Mellitus. [2] [3]. There is adequate evidence from cohort studies regarding a higher incidence of glucose intolerance and DM among HIV infected individuals on combined Anti-Retroviral Therapy (ART) when compared to those not on ART [4].
Impaired glucose metabolism is implicated to be multifactorial, the role of protease inhibitors (PIs) and nucleoside reverse transcriptase inhibitors (NRTIs) leading to DM has been demonstrated by few prospective studies [5,6]. Additionally, DM is known to supress the cell mediated immunity and increases the frequency of infections [7,8].
According to a study conducted by Brown et al in America, diabetes mellitus was 4 times more prevalent in HIV positive individual on ART [9]. Though we have numerous studies addressing common opportunistic infection among PLHIV/AIDS, there is limited data available about opportunistic infections affecting PLHIV/AIDS with diabetes. Early recognition and treatment of OI in these individuals will improve their quality of life. Hence, this study was conducted with the objectives to study the pattern of opportunistic infections in those PLHIV with DM and differences in OI's between those PLHIV, with and without DM.

Materials and Methods
A prospective study was conducted at the Kasturba Medical College Hospital, Mangalore. This hospital has registered about 1500 AIDS patients and has a fully functioning ART centre supported by National AIDS control organization, Government of India. The study was conducted from January 2011-January 2012. PLHIVs' with and without DM (1:1) were included in the study. PLHIV without DM where those individuals who had tested HIV positive without documented DM. Patients were classified as Diabetic if their fasting plasma glucose was 7.0mmol/l (126mg/dl) or 2-h plasma glucose was 11.1mmol/l (200mg/dl) [10]. Further, clinical staging of PLHIV with and without DM was done based on World Health Organization guidelines (WHO) [11]. PLHIV-DM admitted in the hospital were approached and a written informed consent was obtained. A list of PLHIV without DM admitted in the hospital during study period was obtained and every 5th patient was selected for the comparison group. Their laboratory reports were reviewed for data related to opportunistic infection. A semi structured pro-forma was used to collect socio demographic data, clinical features, CD4 counts and treatment. All opportunistic infections were diagnosed based on WHO guidelines [12].
Though most of the opportunistic infections diagnosed for PLHIV-DM had microbiological or histopathological evidence, there were cases of extra Pulmonary Tuberculosis and Cytomegalovirus (CMV) retinitis which were diagnosed clinically based on WHO criteria.

Laboratory diagnosis
Due to resource constraints, laboratory testing was done on samples collected from PLHIV-DM based on the presenting symptoms and clinical features. Samples collected included sputum, oral swab, stool, urine, cerebrospinal fluid (CSF), lymph node aspirate& blood. All samples were collected following standard precautions in suitable sterile universal container. Blood samples were collected aseptically and inoculated into aerobic culture bottles only. The samples were transported to the laboratory immediately and processed without delay in a class II biological safety cabinet. Appropriate methods and procedures for diagnosing opportunistic infections were carried out at Kasturba Medical College, Microbiology laboratory which is a NABL (National Accreditation Board for testing and calibration of Laboratories) accredited laboratory. CSF samples were screened for pathogens using gram stain, ZN stain and India ink staining and cultured for pathogens. Sputum culture was done for patients suspected to have pulmonary tuberculosis on Lowenstein Jenson media. Mycobacterial speciation was done for this cultures using Immuno-chromatographic test called SD TB Ag MPT 64 (S.D. BIO STANDARD DIAGNOSTICS, South Korea). Speciation of candida was done using Chrome agar plates (HiMedia Laboratories).
Serological tests were performed for Cytomegalovirus (CMV) (IgM and IgG), Hepatitis B surface antigen (HBsAg), Herpes Simplex Virus (HSV1 & 2-IgM and IgG). Those patients with clinical suspicion of toxoplasmosis underwent radiological diagnosis followed by serological tests on serum samples (IgM and IgG). CD4 counts were collected from medical records as the test was performed at the ART centre as part of routine care.

Data Analysis
The data was entered and analysed using Statistical Package for Social Sciences (SPSS) version 11.5 (SPSS Inc., 233 South Wacker Drive, 11th floor, Chicago, IL 60606-6412). The results were expressed as proportions and presented in the tables. Chi-square or Fisher exact test were used to examine the relation between categorical variable and Mann-Whitney U tests for relation between Medians.

Ethical issues
Ethical clearance was obtained from Institutional ethics committee at Kasturba Medical College, Mangalore. Permission to conduct the study was obtained from the medical superintendent of the hospital.

Results
The study included 37 PLHIV-DM and 37 PLHIV and all the cases and controls were on ART admitted in the hospital. Median age for PLHIV-DM group was 47 years (IQR: 41-55years) as The most common clinically diagnosed opportunistic infections was oral candidiasis among 49% of PLHIV-DM and 35% of PLHIV. Cryptococcal meningitis was diagnosed among 19% of PLHIV-DM and 16% of PLHIV and Pneumocystis jiroveci pneumonia was diagnosed among 5% of PLHIV-DM compared to 18% of PLHIV. Among bacterial infections extra pulmonary tuberculosis was diagnosed among 22% of PLHIV-DM and 34.5% of PLHIV. Cerebral toxoplasmosis was diagnosed among 11% of PLHIV-DM compared to 13.5% of PLHIV (Table 2).
Of 18 samples collected from patients with candida infections, Candida krusei species was isolated in 9/37, C albicans in 5/37, C tropicalis in 3/37 and C glabrata in 1/37. Out of the 6 sputum samples cultured for pulmonary TB, four samples were identified as Non-tuberculosis mycobacteria (NTM) and two Mycobacterium tuberculosis complexes (Table 2).
All OI s in PLHIV-DM and PLHIV had Microbiological or Histopathological confirmation except out of the 8 samples from suspected cases of extra pulmonary tuberculosis among PLHIV-DM, 3 samples did not show any growth even after prolonged incubation hence

Discussion
This is one of the first studies to the best of our knowledge reporting common opportunistic infections among the subgroup of patients with PLHIV with DM compared with PLHIV without DM from South India. The study detected fungal infections among majority of our patients followed by protozoal and bacterial infections in both groups. Majority of patients in both groups were males, with a significant difference in age between the groups; PLHIV-DM being around 47 years, comparable with the age group affected by type II DM [13]. Despite both groups having been HIV positive and received ART for the same duration of time, those PLHIV diagnosed to have DM had better CD4 count. Among PLHIV with DM being a diabetic could have led to more frequent follow up visits, which could have indirectly improved adherence to ART leading to higher CD4 counts or PLHIV with DM may have complications of OIs which are less likely to be successfully treated on outpatient hence, these patients may have high CD4 counts on admission. Influence of the above factors on CD4 levels need to be further studied.
Oral candidiasis was the most common opportunistic infection in both the groups, being higher in those with DM. This difference could be because candida is the most common  [19] and Ayyagari in Northern India reported 5.6%cases [20]. In the current study, it was identified in about 19% of PLHIV-DM cases, falling within this wide range of findings, but almost twice as much, when compared to the findings of the study conducted by Saldana et al [19] in the same setup. Tuberculosis (TB) is the most common opportunistic infection leading to 25% mortality among AIDS patients [21] and Diabetes is a risk factor for tuberculosis [22]. Pulmonary tuberculosis was seen equally among both groups, while that extra-pulmonary tuberculosis was diagnosed in a slightly higher number among PLHIV patients when compared to PLHIV-DM patients. The difference could be due to significant difference in CD4 counts between the groups, with the latter having a higher CD4 count. Among pulmonary TB isolates, NTMs' were identified in four out of six patients when compared to Mycobacterium tuberculosis complex (two out of six patients). These findings are similar to findings of a study done by Horsburgh C in Atlanta, in PLHIV (AIDS) [23]. A study among HIV positives (non AIDS) individuals has shown that NTMs are associated with fewer cases of pulmonary infections when compared to M tuberculosis complex [24].
A slightly higher number of PLHIVs were affected by cerebral toxoplasmosis when compared to PLHIV-DM (13.5% Vs 11%). The findings in this study are higher than some Indian studies (Sharma et al in north India among PLHIV-3.7%) [25]. Pneumocystis jerovecii pneumonia (PCP pneumonia) is the most common OI in western world [26]. Studies related to PCP pneumonia from the Indian subcontinent are very limited. One such study done by Udwadia et al in Mumbai showed 13% of HIV admissions were due to PCP pneumonia [27]. Eighteen percent of PLHIVs in our study had PCP as compared to 5% in PLHIV-DM. The difference in PCP and cerebral toxoplasmosis rates between PLHIVs with and without DM could be attributed to significant difference in CD4 counts in our study.
The study was limited to a very small group of patients and was from a single tertiary care hospital. Factors such as adherence to ART, type of diabetic medication received by those with DM, and Glycaemic control were not studied. Inclusion of the above factors would make interpretation of findings easier, and these factors need to be addressed in subsequent studies.

Conclusion
The group of patients with DM were older. All the PLHIV with DM had at least one OI with one third of them having more than one OIs. Although PLHIV-DM had higher proportion of oral candidiasis and Cryptococcal meningitis than their counterparts, this study did not identify any significant difference in profile of opportunistic infections (OIs) between PLHIV with and without Diabetes.
Institute of Medical sciences, Bangalore, Karnataka, India and Dr Sudarshan Papanna, Assistant Professor, Department of Social Sciences Central University, Gujarat, India for their valuable inputs during the study. We also wish to thank all the participants.