Pancreatic Insufficiency in HIV: Is It Possible?

Abstract

Pancreatic involvement in AIDS is very important and common, but there are few studies in the literature concerning the pancreas in AIDS. Therefore, our research involves an important issue in the pancreatic field. The objective of the study was to evaluate the profile of HIV-infected patients with probable exocrine pancreatic insufficiency and its relation to the degree of human immunodeficiency virus (HIV) infection. This is a cross-sectional study carried out at Faculdade de Medicina do ABC in partnership with the basic health care unit Vila Guiomar in Santo André. We selected 118 individuals divided into four groups (a control group and three other groups composed of AIDS patients, separated according to CD4 levels); participants had an interview, completed a questionnaire, and had laboratory and imaging tests. The only clinical variables with significant differences among the studied groups were the use of highly active antiretroviral therapy (HAART), the incidence of opportunistic infections, the administration of chemoprophylaxis, and weight loss. There were no differences in the amylase, lipase, and steatocrit dosages among the groups. Levels of fecal elastase 1 were lower in the HIV patient groups (2, 3, and 4) when compared with the control group, although all of them showed average levels that were much higher than the cutoff point (200 μg/g). Only nonalcoholic individuals showed a relationship between diarrhea and alterations in elastase levels. A relationship between the use of HAART and exocrine pancreatic insufficiency in different phases of HIV infection could not be verified.

Introduction

HIV infection is a major public health problem all over the world. WHO's annual report on the progression of the epidemic in the world states that 34 million people were infected by the HIV virus by the end of 2010. Among those, there were around 2.7 million new cases. Since the beginning of the decade the mortality rate caused by complications related to HIV infection decreased from 2 million cases in 2002 to 1.8 million cases in 2010. However, as of the end of 2010, 6,650,000 individuals had access to antiretroviral therapy.¹

In Latin America, by the end of 2010, 15 million people were estimated to be seropositive for the HIV virus and there were around 60,000 HIV-related deaths. In Brazil, according to the epidemiologic bulletin published by the Ministry of Health (2011), in the past 30 years there were 608,230 reported cases of AIDS. In 30 years the number of reported HIV-related deaths was 241,649. In the state of São Paulo there were 207,077 reported cases in the same period.²

Pancreatic dysfunctions are frequent in cases of HIV infection, but are not often studied due to the fact they are evidenced only histologically. Because the pancreas is a retroperitoneal organ, biopsy performances are very restrict. Situations such as the use of antiviral drugs, the onset of opportunistic infections, and the action of HIV also corroborate with pancreatic lesions.

In a study carried out before the introduction of highly active antiretroviral therapy (HAART), postmortem histological analysis of the pancreas in individuals with AIDS revealed an alteration in 90% of the cases according to three distinct patterns: a severe calorie–protein malnutrition pattern (kwashiorkor), an inflammatory pattern, and a mixed type. Pancreatic dysfunction caused by opportunistic microorganisms (Mycobacterium sp., Toxoplasma gondii, fungi, cytomegalovirus, Pneumocystis jiroveci) and by HIV were also frequent.³

These patterns were confirmed in a different study, in which the ultrastructural evaluation of the pancreas in deceased AIDS patients revealed acinar atrophy, depletion of zymogen granules, an increase in lipofuscin pigment, Golgi complex rarefaction, increased and dilated rough endoplasmic reticulum, and an increase in size and number of mitochondria. Although these findings are compatible with exocrine pancreatic insufficiency and calorie–protein undernutrition, there is no clinical manifestation of pancreatic insufficiency.⁴

Is it possible that this histological exocrine pancreatic insufficiency could clinically manifest itself? When studying children infected with HIV and with steatorrhea, Carrocio et al. showed that those children had clinical improvement after the replacement of pancreatic enzymes orally; thus, the answer to the formulated question is affirmative.⁵

To evaluate pancreatic function, direct or indirect methods can be used. The direct methods represent the gold standard, with pancreatic secretion dosage after the infusion of secretagogues. The indirect methods can be applied with pancreatic enzyme dosages after the intake of nutrients, pancreatic enzyme determinations, and the amount of fat in feces and urine (bentiromide and pancreolauryl tests).⁶

Elastase 1 is an enzyme secreted by the pancreas. It is highly stable and resistant to proteolytic degradation during its passage through the gastrointestinal tract. In addition, the immune enzymatic reaction by ELISA is specific for human elastase 1. Because of this, the elastase 1 dosage in feces obtained by this method was shown to be a good marker in the presence of exocrine pancreatic insufficiency. De Oliveira et al. recently standardized this method in the area, showing good sensibility and specificity.⁷

To use this precise method, we standardized fecal elastase 1 with a group with neither pancreatic disease nor HIV infection. The treatment of HIV patients with HAART promoted not only an important decrease in the incidence of opportunistic infections, but also a significant increase in life expectancy. This increase generated the onset of new challenges: metabolic disturbances, lipodystrophy and dyslipidemia, and increasing rates of cardiovascular disease. Would there be any sort of pancreatic manifestation by the immune reconstitution syndrome? Would the increase in life expectancy caused by HAART have any influence on the pancreas?

A possible influence of HAART in the development of an acute or chronic pancreatic lesion, whether preventing alterations in the control of HIV or triggering inflammation as a result of the toxicity of the drugs used, has been widely discussed. A clinical study conducted by Manfredi et al. with patients using HAART showed there was no significant difference between the relationship to the drugs used by individuals with AIDS and pancreatitis and those with isolated laboratory alterations. Laboratory manifestations, such as the increase of amylase, isoamylase, or lipase (with or without pancreatitis per se), were more frequently detected in patients who had been sick longer and who had presented with lower CD4 levels.⁸

Thus, the profile of pancreatic dysfunctions in HIV/AIDS patients as well as a possible interference by HAART have not been well elucidated in the current medical literature. Further studies are necessary so that these subjects can be clarified.

The objective of this study was to evaluate the profile of HIV-infected patients with probable exocrine pancreatic insufficiency and its relationship to the degree of HIV infection.

Materials and Methods

A cross-sectional study was carried out at Faculdade de Medicina do ABC (FMABC), in partnership with a basic health care unit,Vila Guiomar, in Santo André, Brazil, a reference unit for contagious infectious diseases.

The research was approved in accordance with the guidelines of the FMABC Research Ethics Committee according to protocol number 077/2002. All the patients enrolled at the study read and signed an informed consent sheet and gave their writen permission. The candidate screening process occurred between August 2003 and November 2004. Patients were submitted to an interview and completed a questionnaire that covered demographic and clinical data [alcoholism, body mass index (BMI), smoking, illicit drug use, fever, weight loss, diarrhea, and use of antiretrovirals].

Alcoholism was defined by the consumption of at least 50 g of alcohol in a daily routine. Smoking was defined by the consumption of nicotine cigarettes in a daily routine. A sample of 118 adult patients over 18 years of age was included; the patients were HIV positive or negative and had different CD4 levels. Patients with diabetes mellitus were excluded.

All patients were subjected to the following laboratory tests: urea, creatinine, sodium, potassium, amylase, lipase, albumin, aspartate aminotransferase (AST), alanine aminotransferase (ALT), and glucose, as well as hemogram, urine I, protoparasitological stool test (Hoffman and Faust methods), coproculture, steatocrit, fecal elastase 1, d-xylose, and abdominal ultrasonography.

Pancreatic ultrasound abnormality was indicated by an abnormal increase in parenchymal texture. An abnormal liver ultrasound parameter was defined as raised texture/heterogen texture/steatosis. To facilitate the reading of the tables we defined the ultrasound parameter as normal or abnormal.

These tests were necessary to evaluate all possible pancreatic disfunctions. By excluding these disfunctions, it would be possible to isolate any exocrine pancreatic antiretroviral adverse reaction.

Patients were divided into four groups; group 1 was the control and the others were stratified according to the degree of HIV infection, classified by CD4 cell levels:

• Group 1: control—no pancreatic disease and HIV negative (23 patients);

• Group 2: CD4>500 cells/mm³ (32 patients);

• Group 3: CD4 between 250 and 500 cells/mm³ (33 patients);

• Group 4: CD4<250 cells/mm³ (30 patients).

We included a control group with neither pancreatic disease nor HIV infection because the fecal elastase 1 method was not standardized in Brazil. Therefore, a control group with no pancreatic disorder was necessary as fecal elastase 1 was the main parameter used to evaluate exocrine pancreatic disfunction in this study.

The groups were compared based on parametric and nonparametric variables. The first two tables compared the four groups together with demographic and clinical data in Table 1 and laboratory and ultrasound parameters in Table 2. We then divided the 118 patients comparing the fecal elastase 1 result with clinical and laboratory data. The fourth and fifth relationships studied alcohol intake and fecal elastase 1 of the 118 patients with clinical and laboratory data.

Table 1.

Demographic and Clinical Data

	Group 1 control n=23	Group 2 CD4 >500 n=32	Group 3 (CD4 250–500) n=33	Group 4 (CD4 <250) n=30	p
Age (years)^a	28.1±11.3	39.4±12.2	38.3±7.8	40.2±8.6	<0.001
BMI (kg/m²)	23.5±2.8	24.2±3.9	23.4±4.1	23.3±3.8	0.792
Male sex	12	20	18	23	0.220
Ethnicity
White	20	21	17	13	0.012
Smoking	7	14	19	19	0.038
Alcohol intake
No	21	19	24	18
Up to 50 g/day	2	8	2	7	0.011
From 50 to 100 g/day	0	4	5	2
More than 100 g/day	0	1	2	3
Use of illicit drugs	1	8	7	10	0.180
HIV group risk	0	32	32	28	<0.001
HAART	0	21	26	11	<0.001
Fever	1	3	5	3	0.624
Weight loss	1	11	9	5	0.046
Anorexia	0	3	6	4	0.114
Depression	2	3	5	4	0.848
Opportunistic infection	0	6	18	14	<0.001
Chemoprophylaxis	0	2	6	12	<0.001
Diarrhea	1	12	10	5	0.022

Mean and standard deviation.

BMI, body mass index; HAART, highly active antiretroviral therapy.

Table 2.

Laboratory Data and Ultrasound Abnormalities

Variable	Reference range	Group 1 control n=23	Group 2 (CD4 >500) n=32	Group 3 (CD4 250–500) n=33	Group 4 (CD4 <250) n=30	p
Viral load ×10³ (units/ml)^a	0	0	4.94±18.33	40.64±85.02	119.9±127.8	<0.001^*
Urea (mg/dl)	17–50	29.4±7.59	29.8±18.6	26±11.9	27.5±7.14	0.692
Creatinin (mg/dl)	0.7–1.3	1.0±0.1	1.1±1.7	0.8±0.1	0.9±0.26	0.693
Amylase (UI/ml)	80–180	51.7±17.2	61.2±35.3	47.9±21.1	64.2±29.9	0.070
Lipase (UI/ml)	10–140	58.7±70	112.2±115.4	107.2±42.6	97.7±122.4	0.186
Albumin (g/dl)	3.5–5.6	4.7±0.4	4.2±0.4	2.8±0.3	4.2±1	<0.001^*
AST (U/liter)	10–37	26.9±5.1	44.2±28.7	38.8±25.4	46.1±36.8	0.630
ALT (U/liter)	8–39	21.4±6.7	39.2±31.4	32.2±23	35.6±24.2	0.540
Glucose (mg/dl)	60–99	91.4±10	110.5±32.9	106.4±41.5	97.6±22.6	0.078
Hemoglobin (g/dl)	13.5–18	14.5±1.5	14.6±1.4	14.2±1.7	13.5±1.8	0.025^*
Hematocrit (%)	40–54	45.1±4	43.4±4.6	41.2±4.7	40.1±5.6	0.001^*
MCV (u³)	82–94	82.3±8	105.4±10.6	103.3±13.9	102.2±11.2	<0.001^*
White cell count×10³ (mm³)	4.5–11	6.3±1.4	20.4±1.7	17±2.1	14.1±1.6	0.525
Platelets (10³/mm³)	150–350	232.3	212.7	199.1	194.6±62.5	0.161
Stool positive worms presence	Negative	0	4	8	0	0.004^*
Fecal elastase 1 (μg/g)	≥200	500	337.9	351.1	367.8±182.9	<0.001^*
Steatocrit (%)	≤2	4.3	5.9	5.4	5.1±3.1	0.265
Abnormal d-xylose (mg/dl)		0	5	1	2	NS
Liver ultrasound						0.129
Normal		23	28	28	29
Abnormal		0	4	5	1
Pancreas ultrasound						0.347
Normal		23	31	30	29
Abnormal		0	1	3	1

p values statistically significant.

Variables are expressed as mean values and standard deviation.

MCV, mean corpuscular volume; NS, not significant.

Either Fisher's exact test or the Chi-square test was applied for the analysis of nonparametric variables whereas the Student's t-test was chosen for the analysis of parametric variables. The two-factor variance analysis was used to evaluate the significance of the many variables in relationship to the presence or absence of alcoholism and the results of fecal elastase 1 measurements. Values of p<0.05 were considered significant.

Results

The four characterized groups, as described before, were analyzed according to epidemiologic variables (Table 1). In the control group (1) patients were predominantly white and the age average was significantly lower than in the other groups.

In the HIV groups (2, 3, and 4) there was a prevalence of smokers, alcoholics, and carriers with a higher number of risk factors for the acquisition of HIV infection compared to the control group. Educational attainment in group 1 was predominantly higher. The other variables did not present significant differences among the four groups.

In Table 1 the use of HAART, the incidence of opportunistic infections, the use of chemoprophylaxis, and weight loss were the only studied clinical variables that demonstrated statistical differences among the groups.

Table 2 shows relevant laboratory tests. Differences in the average levels of hemoglobin and hematocrits with significantly lower values among group 4 patients were found when compared to the other groups. The dosages of amylase, lipase, and steatocrit did not show any significant difference among the four groups. The dosage of fecal elastase 1 presented a sensibly lower average in the HIV groups (2, 3, and 4) than in the control group, although all of them had much higher average levels than the test cutoff point (200 μg/g). The albumin dosage was significantly lower in group 3. The protoparasitological stool (p=0.004) and the HIV viral load test analyses (p<0.001) showed a significant statistical difference among the groups. There was no difference in ultrasonographic tests among the evaluated patients.

The target in Table 3 was pancreatic insufficiency. Patients were separated according to elastase results (normal or abnormal) by comparing the clinical and laboratory variables among them. There was some difference in the fecal elastase 1 result between the control group and the other groups. This difference, however, was not significant when stratified by CD4 level (p=0.121). Alcoholism was a factor that also proved to be a variable with important differences between the two groups (p=0.002). The other analyzed parameters were not significant.

Table 3.

Relationship Between Fecal Elastase-1 with Clinical and Laboratory Data

	Normal	Abnormal	p
Albumin (g/dl)	4.3	4.0	0.378
BMI (kg/m²)	23.7	23.9	0.782
MCV (u³)	100.1	102.8	0.615
Viral load (units/ml)	65,402	44,013	0.170
CD4
Control	23	0
>500	26	6	0.121
250–500	27	6
<250	23	4
Age (years)	37	37.2	0.951
Steatocrit (%)
Positive	16	4
Negative	43	4	0.345
Indetermined	39	8
Smoking (individuals)	46	11	0.052
Glucose (mg/dl)	101.3	108.5	0.133
Alcohol intake
No alcohol intake	73	7
Up to 50 g/day	16	2	0.002
From 50 to 100 g/day	5	6
More than 100 g/day	4	1
HAART
Yes	48	8	0.970
No	49	8
Emaciation
Yes	22	3	1
No	77	13
Diarrhea
Yes	21	6	0.202
No	78	10
d-Xylose (average)	45.3	31	0.135
Pancreatic ultrasound
Normal	96	14
Abnormal	3	2	0.142

MCV, mean corpuscular volume; HAART, highly active antiretroviral therapy.

Table 4 evaluates the existence of exocrine pancreatic insufficiency among the groups separated by fecal elastase 1 values (normal or abnormal). It was observed that not only alcoholic, but also nonalcoholic patients showed a statistically significant difference (p=0.031) in serum albumin values. The other analyzed parameters indicated that alcoholism is not an isolated prognostic factor in the cases studied.

Table 4.

Relationship Between Fecal Elastase-1 of Alcoholics and Nonalcoholics with Clinical and Laboratory Data

Variable	+ Alcohol ^a	− Alcohol ^a	p^b	+ Alcohol ^c	− Alcohol ^c	p^d	P^e
Albumin (g/dl)	4.36±1.10	4.03±0.53	0.050	3.78±0.29	4.3±0.41	0.147	0.031
BMI (kg/m²)	23.03±3.39	23.88±3.86	0.577	22.70±1.79	25.58±5.10	0.605	0.326
MCV (u³)	107.96±10.90	97.49±12.60	0.884	107.88±14.66	96.42±9.84	0.868	0.001
Viral load×10³ (U/ml)	52.14±78.99	69.70±132.10	0.577	68.11±98.23	10.26±20.37	0.605	0.335
Age (years)	39.48±9.48	36.09±11.45	0.897	36.22±3.80	38.57±13.98	0.864	0.340
Glucose (mg/dl)	100.12±14.97	101.73±36.42	0.361	107.33±13.35	110.14±15.36	0.795	0.994
d-Xylose (mg/dl)	38.86±20.15	46.51±27.13	0.322	26.98±9.44	40.36±7.45	0.249	0.752

Normal fecal elastase.

p value for alcohol intake and normal fecal elastase.

Abnormal fecal elastase.

p value for alcohol intake and abnormal fecal elastase.

p value for normal and abnormal fecal elastase and alcohol intake.

MCV, mean corpuscular volume; BMI, body mass index.

Discussion

Although the HIV groups were at different stages of disease progression and they were patients receiving ambulatory follow-up care in good general condition, they showed a great homogeneity as to demographic and clinical and laboratory profiles. The studied population is a good sample of infected patients in our environment, who, according to the last issue of the epidemiological bulletin, are predominantly young white males.

Although alcoholism showed a significant difference among the groups, when only HIV-positive individuals were evaluated and divided according to a higher or lower intake of 50 g/day of ethanol, the difference ceased to exist. This fact demonstrates that in spite of some heterogeneity among the groups, levels of alcoholism that could lead to the development of pancreatic insufficiency were not different among HIV patients. A probable explanation could be the suspension of alcohol intake and the prescription of large quantities of drugs with the onset of the disease.

The occurrence of diarrhea and the loss of weight, typical in patients with exocrine pancreatic insufficiency, seems to be interrelated in all tested groups. Nevertheless, these symptoms were more prevalent in groups 2 (CD4>500 mm³) and 3 (CD4 between 250 and 500 mm³) than in group 4 (CD4<250 mm³); therefore, there was no direct relationship between the status of viral infection and these alterations.

Even in the post-HAART era diarrhea is still a major clinical problem for HIV carriers, not only because of a worsening in the quality of life,⁹ but also for its intimate connection with disabsorptive cases that cause weight loss. This, in fact, becomes an independent predictor of mortality¹⁰ once adequate nutrition helps strengthen the immune system of HIV-infected patients by reducing the severity of opportunistic infections in these patients.¹¹

Regardless of their CD4 level, all HIV-infected groups were HAART users. There were fewer cases of diarrhea among patients with lower CD4 levels, and these individuals are precisely the ones who benefit most from the therapy. HAART was used according to current medical literature recommendations.¹²

An interesting point concerning the evaluation of these patients is the fact that almost all of them were in good general condition and receiving ambulatory follow-up care. Laboratory tests, considered significant in the studied population, showed undernutrition in these patients. This fact could be confirmed in the differences among the groups in relation to hemoglobin and hematocrit levels and mean corpuscular volume (MCV). Group 3 (CD4 250–500 cells/mm³) was the group that presented lower levels of hematocrit and hemoglobin. However, this group did not show higher MCV levels, a fact that indicates that the anemia in these patients was not caused by only one nutrient deficiency.

Albumin, also used as a possible marker of undernutrition, was significantly reduced in group 3, which was, in turn, the group with a higher prevalence of verminosis shown in positive protoparasitological tests.

As seen in the studies of Chehter et al.,^3,4 alterations similar to the ones found in patients suffering from severe undernutrition, such as kwashiorkor, had already been found in HIV carriers in the pre-HAART era. From a clinical point of view the current therapy seems not to have resulted in great differences concerning the probable causes of undernutrition in these patients. This reinforces the idea that this complication is multifactorial.

Although the average levels of fecal elastase 1 were much lower in the HIV groups in relationship to the control group, they were not enough to diagnose exocrine pancreatic insufficiency, as shown in Table 2. Nevertheless, when individuals with abnormal fecal elastase 1 are analyzed, suggesting exocrine pancreatic insufficiency, the presence of HIV was remarkable, despite the fact that there was no significant relationship among the groups with different CD4 levels, as shown in Table 3. These findings are similar to the ones already presented by Carrocio et al. ^13,14

Alcoholism also showed, as expected, a relationship with the fecal elastase 1 tests result. Carrocio et al. could not demonstrate an inverse relationship between steatocrit and fecal elastase 1 levels¹³ in HIV-positive adult patients. The same relationship was significant when the studied group was composed of HIV-positive children¹⁴; still, steatocrit values seem to decrease during therapy with orally administrated pancreatic enzymes, and the clinical improvements of some patients were considered significant.⁵ The same study also could not demonstrate a relationship between the presence of altered fecal elastase 1 and the severity of HIV infection, or the presence or absence of undernutrition.¹³

When the groups were separated by the presence or absence of exocrine pancreatic insufficiency, alcoholic and nonalcoholic patients presented differences in albumin values. The other parameters indicated that alcoholism by itself did not prove to be an isolated prognostic factor in the studied cases.

The nonparametric analysis, which took alcoholism into consideration, revealed a significant difference according to the concentrations of fecal elastase 1 only for the presence or absence of diarrhea. This finding shows that diarrhea does not seem to be necessarily linked to exocrine pancreatic insufficiency.

Although in some studies^15,16 the use of HAART was related to the development of pancreatitis, in this study it was not confirmed based on the results presented.^17
–19 Differences after the systematic use of HAART were not identified either, a result similar to the one found by Riedel et al.²⁰ Therefore, there is no basis on which to claim that these drugs are really connected to the development of pancreatic insufficiency.

In conclusion, evidence that would prove the presence of exocrine pancreatic insufficiency in the different stages of HIV infection was not found. The increase in life expectancy, the immunological reconstruction, and the reduction in the incidence of opportunistic infections may explain the onset of pancreopathies similar to the ones found in HIV-negative patients.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Global HIV/AIDS Response. Epidemic update and health sector progress towards Universal AccessProgress Report 2011. [Internet]. WHO [updated November 30, 2011 and cited January 25, 2012]www.who.int/hiv/pub/progress_report2011. 2011 December 5.

Brazilian AIDS Bulletin 2011: [Internet]. Brazilian Health Ministry[updated November 28, 2011 and cited January 25, 2012]www.aids.gov.br/publicacao/2011/boletim_epidemiologico_2011. 2011 December 5.

Chehter

, Longo

, Laudanna

, Duarte

. Involvement of the pancreas in AIDS: A prospective study of 109 post-mortems. AIDS, 2000; 14:1879–1886.

Chehter

, Duarte

, Takakura

, Longo

, Laudanna

. Ultrastructural study of pancreas in AIDS. Pancreas, 2003; 26,2:153–159.

Carroccio

, Guarino

, Zuin

, Verghi

, Canani

et al. Efficacy of oral pancreatic enzyme therapy for the treatment of fat malabsorption in HIV-infected patients. Aliment Pharmacol Ther, 2001; 15:1619–1625.

Chehter

. Pancreas diseases. Clinical Medicine: Diagnosis and Treatment. Sarvier: São Paulo, 2008; 933–952.

de Oliveira

, Fonseca

FLA

, Salotto

, Chicoli

, Doria

et al. Validation of fecal elastase-1 determination using immunoenzymatic assay in HIV-infected patients. J Clin Lab Anal, 2008; 22:286–290.

Manfredi

, Calza

, Chiodo

. A case-control study of HIV-associated pancreatic abnormalities during HAART era. Focus on emerging risk factors and specific management. Eur J Med Res, 2004; 9:537–544.

Siddiqui

, Bini

, Chandarana

, Leong

, Ramsetty

, Schiliro

et al. Prevalence and impact of diarrhea on health-related quality of life in HIV-infected patients in the era of highly active antiretroviral therapy. J Clin Gastroenterol, 2007; 41:484–490.

10.

Mangili

, Murman

, Zampini

, Wanke

. Nutrition and HIV infection: Review of weight loss and wasting in the era of highly active antiretroviral therapy from the nutrition for healthy living cohort. Clin Infect Dis, 2006; 42:836–842.

11.

Oguntibeju

, van den Heever

, Van Schalkwyk

. The interrelationship between nutrition and the immune system in HIV infection: A review. Pak J Biol Sci, 2007; 10:4327–4338.

12.

Thompson

, Aberg

, Cahn

, Montaner

, Rizzardini

et al. Antiretroviral treatment of adult HIV infection. 2010 recommendations of the International AIDS Society–USA panel. JAMA, 2010; 304:321–333.

13.

Carroccio

, Di Prima

, Di Grigoli

, Soresi

, Farinella

et al. Exocrine pancreatic function and fat malabsorption in human immunodeficiency virus-infected patients. Scand J Gastroenterol, 1999; 34:729–734.

14.

Carroccio

, Fontana

, Spagnuolo

, Zuin

, Montalto

et al. Pancreatic dysfunction and its association with fat malabsorption in HIV infected children. Gut, 1998; 43:558–563.

15.

Price

, Schmid

, Ong

, Adjukeiwicz

, Peaston

et al. Pancreatic exocrine insufficiency in HIV-positive patients. HIV Med, 2005; 6:33–36.

16.

Dutta

, Ting

, Lai

. Study of prevalence, severity and etiological factors associated with acute pancreatitis in patients infected with human immunodeficiency virus. Am J Gastroenterol, 1997; 92:2044–2048.

17.

Manfredi

, Calza

. HIV infection and the pancreas: Risk factors and potential management guidelines. Int J STD AIDS, 2008; 19:99–105.

18.

Reisler

, Murphy

, Redfield

, Parker

. Incidence of pancreatitis in HIV-1-infected individuals enrolled in 20 adult AIDS clinical trials group studies: Lessons learned. J Acquir Immune Defic Syndr, 2005; 39:159–166.

19.

Bush

, Kosmiski

. Acute pancreatitis in HIV-infected patients: Are etiologies changing since the introduction of protease inhibitor therapy? Pancreas, 2003; 27:e1–5.

20.

Riedel

, Gebo

, Moore

, Lucas

. A ten-year analysis of the incidence and risk factors for acute pancreatitis requiring hospitalization in an urban HIV clinical cohort. AIDS Patient Care STDS, 2008; 22:113–121.