ABSTRACT
Introduction:
In individuals with acquired immunodeficiency syndrome (HIV/AIDS), abdominal pathologies rank second in frequency only to pulmonary illnesses. An essential imaging method for assessing abdominal diseases is ultrasonography (USG). In this study, abdominal pathologies in HIV/AIDS patients were evaluated using USG, and their relationship to CD4 count was further examined.
Materials and Techniques:
400 HIV+ subjects with aberrant abdominal USG participated in the current investigation. The subjects were assessed and graded as per the CD4 counts. Later the comparisons were drawn between the USG, and its relationship to CD4 count using SPSS 16.0 software, and all data were examined using appropriate statistical tools.
Results:
Men were over 60% of the 400 subjects. The average age of these subjects was 35.6 years; the range for this age group was 6 to 63 years. Spleen involvement was found on ultrasonographic examination in 45.1% of subjects, while liver and lymph node involvement was seen in 43.6% of subjects. Substantial correlations between CD4 counts and findings such as periportal & mesenteric lymphadenopathy, localized pancreatic lesion, splenic microabscess, splenomegaly, and hepatomegaly were found. One percent of individuals had lymphoma, which affected the retroperitoneal lymph nodes, pancreas, and liver.
Conclusion:
Present research demonstrates the significance of abdominal ultrasonographic examination in HIV+ patients. CD4 counts have a big impact on how an HIV/AIDS patient’s differential diagnosis is determined. The interpretation of USG results in relation to CD4 levels may aid in accurate diagnosis.
KEYWORDS: CD4 count, HIV, lymphoma, splenic microabscess, ultrasonography
INTRODUCTION
A retrovirus called the human immunodeficiency virus (HIV) attacks CD4 cells and impairs their functionality. Worldwide, the number of immunocompromised patients has increased at an unheard-of rate during the past ten years. According to World organisations, nearly 39 million people worldwide had HIV in 2019 with the majority of them from economically backward nations.[1]
Multiple infections are seen that may range from simple bacterial to live threatening ones are and are frequently linked to HIV/acquired immunodeficiency syndrome (AIDS). The degree of immunodeficiency seen correlates with CD4 counts and is a useful indicator for tracking the course of the illness.[2] A patient’s vulnerability to infection and, consequently, aberrant sonographic results, should rise as their immune system weakens.[2-5]
According to the NACO of India, nearly 3 million people in India are estimated to be living with HIV, with the first case being reported almost 40 years back. Since then, Asia has seen an increase in HIV infection rates, particularly in India. 3 Antiretroviral therapy’s development and widespread use have increased life expectancy and enhanced quality of life, which may be why these abdominal signs are now becoming more common.[6,7] It is well-recognized that HIV/AIDS can affect many organs and cause a wide range of clinical symptoms.
“Ultrasonography (USG)” is a flexible imaging technique that can assess the majority of abdominal pathologies. Compared to USG, other imaging techniques, especially “Computed Tomography (CT)”, helps study pathologies in depth. But, CT is frequently seen as a backup option in abdominal imaging because it uses radiation, is more expensive, is harder to get, and frequently produces results that are comparable to USG. This is especially factual in underdeveloped nations, where the economy makes abdominal ultrasound a feasible modality for the evaluation of HIV-positive people. Imaging is crucial to the treatment plan. The amount of the CD4 + count correlates with the degree of immunodeficiency, making it a useful indicator for tracking the course of the disease. It is anticipated that when immune function declines, infection susceptibility and, in turn, aberrant sonographic findings, should rise8. It is anticipated that when immune function declines, infection susceptibility would rise, and as a result, aberrant sonographic results will do the same.
Accordingly, the objective of the existing study was to consider the efficiency of the US for the diagnosis of abdominal diseases as well as variations in the course of HIV+ illness among people. Also, abdominal pathologies in HIV/AIDS patients were evaluated using USG, and its relationship to CD4 count was further examined.
MATERIALS AND METHODS
The study was designed after taking permission from the ethics committee. The records of the department of radiology were collected from 2019-2022 and the analysis was done for the patients who had the ELISA tests done and confirmed as HIV+. The study included 400 subjects who were referred to the department of radiology for the USG for various reasons of “Abdominal pain, abnormal liver and/or renal function, weight loss, vomiting, diarrhoea, fever, and sepsis”. All the subjects had a positive finding in the abdominal USG. Both the genders and all the adult age groups were considered for the study.
All those who had a negative ELISA report, and positive patients who didn’t consent and had no apparent finding in the abdominal USG were all excluded from the study. All individuals provided informed consent to their voluntary involvement in the study. When patients in the current study were unreactive to TB treatment or when there was a strong suspicion of malignancy, were further referred to the histological examinations.
Technique and interpretation for imaging
GE Healthcare LOGIQ 3 and Philips HD11 XE ultrasound machines were used in the current investigation to conduct real-time USG. Convex abdominal probes, linear probes, and transvaginal probes were utilized to assess deeper structures, the superficial structures, and the female pelvic region, respectively. Each study participant’s liver size, echotexture, and focal lesions were first assessed. Hepatomegaly was regarded as being present if >15 cm midclavicular dimension was seen longitudinally. It was also done to compare the liver’s echogenicity to that of the kidney and spleen. The size, patency, and color of the hepatic and portal veins were later established. Intrahepatic biliary radicles >7 mm and common bile duct >2 mm, were considered enlarged. The gall bladder’s (GB) right, sub-costal, and oblique sections were examined, and a GB >3 mm was deemed to be thickened. Gallstones, if any, were noted to be present.
The epigastric region was examined. If the anteroposterior diameter of the head, body, and tail of the pancreas was greater than 3 cm, 2 cm, and 2.5 cm, respectively, the pancreas was regarded as bulky. If the pancreatic duct (PD) measured more than 2 mm, it was deemed dilated. It was noted if there were calcifications.
If renal echogenicity was more than that of the liver or spleen, it was regarded as enhanced. When the diameter of the retroperitoneum and abdominal lymph nodes was greater than 1 cm, they were deemed to be enlarged. Additionally, lymph node hilum and echotexture were recorded.
Each person’s spleen was examined for localized lesions. When the spleen was longer than 12 cm, splenomegaly was suspected.
To assess the abdominal cavity, anus, rectum, caecum, ilium, sigmoid colon, and appendix, transducers were inserted into the right and left iliac fossa. When it was greater than 4 mm, gut wall thickening was taken into account. The evaluation of the urinary bladder, prostate, uterus, and adnexa was done by putting the transducer in a suprapubic area.
CD4 validations
CD4 cell counts were used to group patients. patients whose CD4 levels are less than 500, range between 499 and 201 cells/mm3, and “Class-I (Non-significant), Class-II (Mild to Advanced), and Class-III (Severe)” were each assigned a cell density of 200 cells/mm3, correspondingly.[3,5]
Statistic evaluation
Analysis of associations between CD4 levels and diseases was done using the ANOVA & Chi-square test. Observations have been displayed as percentages and mean values. The cutoff for significance was P < 0.05.
RESULTS
The 400 people who were considered had an average age of 35.6 ± 10.8 years (ranging from 6 to 63 years old), with the majority of patients (35.1%) falling in this age range. In addition, 66.2% of the patients evaluated were men, compared to 33.8% M: F = 2:1. 52.5% of subjects were in Class II, 24.1% were in Class III, and 23.1% were in Class I. Current study found that patients within each of these classifications had significantly different presentations of the various abdominal pathologies. The incidence of all these pathologies was inversely proportional to the CD levels and was evident after <500 cells/mm3 levels. These results did not, however, differ significantly between the CD4 classes taken into account [Table 1].
Table 1.
Comparison of the USG and the CD4 classes
Ultrasonograhic presentations | Class I | Class II | Class III | P |
---|---|---|---|---|
Ascites | 2.1% | 5.7% | 7.3% | 0.266 |
Bowel wall thickening | 1.0% | 3.8% | 6.2% | 0.178 |
Dilated Common bile duct | 1.0% | 0.4% | 3.1% | 0.155 |
Dilated Pancreatic duct | - | 1.4% | 1.0% | 0.514 |
Focal liver lesions | - | 5.7% | 4.1% | 0.064 |
Focal pancreatic lesion | - | - | 2.0% | 0.042 |
Hepatomegaly | 42.8% | 44.1% | 29.4% | 0.045 |
Mesenteric Lymph nodes | 13.1% | 17.4% | 32.5% | 0.001 |
Periportal Lymph nodes | 9.8% | 12.0% | 27.3% | 0.001 |
Portal vein | - | 0.4% | 1.0% | 0.400 |
Renal focal lesions | - | - | 1.0% | 0.208 |
Renomegaly | 3.2% | 2.3% | - | 0.241 |
Retroperitoneal Lymph nodes | 22.9% | 18.3% | 33.5% | 0.076 |
Splenic microabscess | 2.1% | 19.3% | 45.2% | 0.001 |
Splenomegaly | 35.1% | 33.4% | 49.4% | 0.024 |
Uterus and Adnexa | - | 1.3% | - | 0.606 |
Sonographic results
When abdominal sonography was positive, 60.1% of patients had only one system involved, whereas 39.7% had multiple systems involved. The abdominal USG of this group showed involvement of the pancreatic and biliary organs in 1.2% of cases, intestine in 3.7%, kidney in 2.2%, ascites in 5.1% of cases, liver and lymphadenopathy in 43.5% of cases, and spleen in 45.1% of cases. The uterus and adnexa were the least often implicated organs (0.7%). Four of these individuals had lymphoma, and the remaining patients all had infections [Table 1].
Liver and biliary
Enlargement was seen in 40.2% of the subjects evaluated, making it the most often seen anomaly in adults between the ages of 31 and 40. Hepatomegaly was more prevalent in patients with CD4 >200 cells per mm3, and the incidence of the condition varied substantially between the three classes (P < 0.046). In 12% of subjects, the liver’s echogenicity was found to be elevated, and in 5% of those evaluated, the liver’s echotexture was found to be altered.
4.0% of the patients whose livers were examined had focal hypoechoic findings. Liver abscesses were seen in 3.7% of subjects with hypoechoic lesions. Although there was no association between CD4 levels and hepatic lesions (P = 0.064), lesion were frequently seen associated with lower CD4 levels. 1.2% of the cases taken into consideration had biliary tract anomalies, and 1.4% of cases had GB calculi [Table 1].
Spleen findings
45.2% of patients had splenomegaly or localized hypoechoic lesions. 37.7% of patients had splenomegaly. 49.4% 33.4%, and 35.1%, of patients in Class III, Class II, and Class I, groups, had splenomegaly. Splenomegaly was positively associated with lower CD levels. Compared to patients in the Class I group, microabscesses were seen significantly in the other two classes [p < 0.001] [Table 1].
Pancreas
In 1.4% of instances, there were pancreatic anomalies. Pathologies were significant between the classes of CD levels [Table 1].
Lymph nodes
43.5% of the patients who were investigated showed abdominal lymphadenopathy. The most common affected area (22.8%) was the retroperitoneal region. No significant variance between classes of CD4 for Retroperitoneal lymphadenopathy (P = 0.077). Nevertheless, subjects with CD4 <200 cells/mm3 had considerably higher frequency anomalies linked to periportal and mesenteric lymph nodes (P = 0.001 and P = 0.002) [Table 1].
Kidney
9.0% of patients with both normal and larger kidneys showed elevated renal echogenicity. In 2.1% of the patients evaluated, renal abnormalities were seen. 2 percent of them (P = 0.24; renomegaly not significantly correlated with CD4 levels) had it. The majority of the patients were identified as having pyelonephritis [Table 1].
Various systems
Among 3.7% of the individuals who underwent examination, thickening of the intestinal wall was also noted. All cases of intestinal wall thickening had caecum involvement as the abnormality’s underlying cause. In our study, TB was discovered to be the cause of the thickening of the gut wall in every instance. Two of these cases had the development of intestinal masses and ileal thickening. In 5.3% of the instances lacking internal septations and echoes, ascites were observed [Table 1].
DISCUSSION
There is a global pandemic of HIV/AIDS. With about 2.1 million HIV-positive individuals living there, India has the third-largest HIV epidemic in the world, with a prevalence of about 0.3%.[1] Within the Indian population, AIDS cases are increasing dramatically, and proportionally, so are the numerous spectrum of disorders affecting various other systems. The Current study evaluated a variety of sonographic anomalies that can be seen in HIV/AIDS patients’ abdomens. According to the study, CD4 levels were substantially correlated with periportal & mesenteric lymphadenopathy, localized pancreatic lesion, splenic microabscess, splenomegaly, and hepatomegaly. This suggests that ultrasonographic findings and immunosuppression are directly related. The numbers of CD4 cells did not significantly correspond with the other sonographic anomalies, though.
The human immune system deteriorates as a result of HIV infection, which also leads to abdominal diseases. The CD4 level are a key measure of host immunity, and it has been demonstrated that CD4 <500 cells/mm3 increase the risk of acquiring neoplasms and gastrointestinal infections. AIDS is clinically characterized as a further decline in CD4 <200 cells/mm3, which increases these diseases. Most of the patients evaluated for the current study had CD4 counts between 200-500 cells/mm3. Earlier, patients were divided into several groups depending on their CD4 levels by a few authors.[3,5] These authors stated that, in contrast to our analysis, the preponderance of the patients evaluated had CD4 <200 cells/mm3. The majority of the patients with abdominal anomalies were male, with M: F = 2:1. In line with this, Mahmoud et al.[6] and Bhabhor et al.[3] reported that males were most prone to experience the anomalies. However, more female patients with abdominal anomalies were found by Obajimi et al.[7] The current research results are in good agreement with earlier research, which found that HIV patients tend to be between 35 and 40 years old on average.[3,7,8] The higher proportion of patients between the ages of 31 and 40, or in their fourth decade, may be related to an increase in sexual activity during this time.
In the current study, a significant number of patients had splenomegalies and splenic microabscesses. Those with CD4 counts with HIV frequently experienced this. Current results so agree with those published in numerous earlier research.[3,5,7,9,10] Without splenic focal lesions, splenomegaly in HIV patients may be a sign of other infective pathologies. Similar to Bhabhor et al.,[3] TB continues to be the most frequent reason for splenic microabscesses in this study, frequently affecting HIV patients with CD4 <200 cell/mm3. The current study’s large patient population with microabscess of the spleen may reflect India’s higher tuberculosis prevalence.[5,7,9]
Patients with HIV/AIDS frequently exhibit hepatomegaly, which can be caused by a non-specific infection response.[4-6] A frequent abnormality found in the Current study was hepatomegaly, which had a strong correlation with CD4 levels. As a result, hepatomegaly was documented in numerous earlier investigations, however, none of them found a connection between hepatomegaly and CD4 levels.[3,4-6,11] It was seen that a limited number of subjects with changed echotexture and increased hepatic echogenicity, similar to Bhabhor et al.[3] findings. HIV-positive patients’ diffuse hyperechoic hepatic sonographic patterns could be brought on by hepatic granulomatosis or fatty infiltration. The liver can be non-invasively examined by USG to distinguish between infected and non-infectious localized lesions. In the Current investigation, the majority of patients with focal hypoechoic liver lesions had abscesses diagnosed, with the exception of one patient who had cancer. According to these findings, focal hypoechoic liver lesions that did not substantially correlate with CD4 levels were observed by Pawar et al.[4] There is evidence that several fungal infections affect the liver in AIDS patients.[11] It was seen that one patient with a hypoechoic liver abscess that was caused by a candida infection and CD4 levels of less than 200 cells/mm3.
The current study detected GB wall thickening and gall stones in HIV patients, albeit the figures recognized in this investigation were smaller than those of earlier publications by Pawar et al.[4] There was no discernible relationship between CD4 numbers and GB wall thickening, gallstones, or both. Similar findings in HIV patients were reported by Obajimi et al.[7]
In the Current analysis, lymphadenopathy was another frequent observation, with periportal and mesenteric lymphadenopathy significantly associated with CD4 levels. Similar to this, numerous studies revealed that among HIV/AIDS patients, abdominal lymphadenopathy was the most frequently discovered aberrant finding that was substantially related to CD4 levels and served as an independent predictor of abdominal TB.[3-5,12]
The most popular imaging technique for evaluating renal impairment in HIV is USG. The incidence of renal involvement in the current study was lower (1%) and manifested as pyelonephritis. According to study observations, renomegaly was found in 19% and 20% of patients, respectively, by Di Fiori et al.,[13] which was not linked to CD4 levels. In reality, the majority of the patients in the studies had CD4 levels of 200 to 349 cells/mm3.[14] Renomegaly was described by Adeyekun et al.[15] in fewer participants (8%) and their presence did not substantially connect with CD4.
Only 9.1% of individuals did the Current study find evidence of elevated renal echogenicity. Contrary to Current findings, Blessing et al.[5] found that the presence of CD4 cells was substantially linked with the presence of elevated renal cortical echogenicity in 37.3% of patients. Similarly to this, Adeyekun et al.[15] found that renal cortical echogenicity increased in 41.7% of patients, although they did not find a correlation between the anomaly and CD4 counts. The majority of the study’s participants had low CD4 counts (200 cells/mm3), which may be the cause of the increased prevalence of renal cortical echogenicity that was seen in these studies. High viral loads and low CD4 counts have been demonstrated to be good predictors of the pathology, even though many factors influence the augmentation of renal echogenicity in HIV/AIDS patients.[13]
HIV infection reduces cellular immunity, intestinal peristalsis, and stomach acid output. These elements encourage enteritis and typhlitis, which are linked to opportunistic pathogen infections like TB. Gastrointestinal problems and CD4 levels did not, however, significantly correlate in individuals. According to Harriet et al.[16] and Blessing et al.,[5] the presence of thicker intestinal walls in HIV cases is strongly linked with CD4 cell counts. In line with earlier research, the Current study found ascites in 5.3% of the patients evaluated, and CD4 counts did not substantially correlate with its existence.[16] One patient who had been given an inflammatory illness diagnosis had tubo-ovarian involvement.
Patients with HIV/AIDS have a higher chance of acquiring cancer because of their compromised immune systems. To distinguish cancer from its imitators, it is crucial to identify HIV/AIDS-related cancers early and accurately, as well as the distinctive aspects of their imaging. The most frequent cancer in HIV patients that characterizes AIDS is lymphoma.[8,9] In this investigation, retroperitoneal lymph nodes, pancreas, and liver were all affected by lymphoma in 1% of individuals.
In the current investigation, only a few patients (0.5%) had isolated hypoechoic lesions inside the pancreatic. It’s interesting to note that all of the cases had CD4 <200 cells/mm3 and were identified as lymphomas. Furthermore, a significant relationship between CD4 levels and the occurrence of localized hypoechoic lesions were seen (P = 0.042). In contrast, Blessing et al.[5] 6% of patients had a hypoechoic pancreas without any focal lesions. Furthermore, neither pancreatomegaly nor hypoechogenicity was associated with CD4 count. Similarly to this, Lito et al.[17] found that HIV-infected patients had pancreatic neuroendocrine tumors that did not meet the criteria for AIDS. This was probably caused by the mild immunosuppression that these patients experienced (median CD4 count: 497 cells/mm3). In 1% of the individuals evaluated, the Current study also found dilated PD, which may have been caused by pancreatic atrophy.
The lack of microbiological proof of splenic and liver abscesses, and abdominal TB, in all subjects limits the scope of the Current investigation. Additionally, the Current study may have been biased because all of the subjects included had abdominal USG results that were positive and we were previously aware of their CD4 numbers. As a result, results could vary if HIV-infected persons were investigated without taking into account their CD4 numbers. Lastly, due to the limitations of the Current investigation, it was unable to compare HIV patients with lesser CD4 counts who had no abdominal USG pathologies.
CONCLUSION
For assessing numerous abdominal diseases, ultrasonography is a widely used, safe, and affordable technique. In HIV/AIDS patients, specific abdominal diseases can be assessed using USG can improve diagnosis precision. Narrowing the differential diagnoses will be made easier by correct elucidation of abdominal USG in combination with an understanding of probable pathological manifestations that may arise in HIV patients with certain CD4 counts. As a result, USG will make it easier to evaluate HIV/AIDS patients in places with inadequate resources, which are frequent in India.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
REFERENCES
- 1.UNAIDS Report on the Global AIDS Epidemic. 2019:16. [Google Scholar]
- 2.Brunetti E, Brigada R, Poletti F, Maiocchi L, Garlaschelli AL, Gulizia R, et al. The current role of abdominal ultrasound in the clinical management of patients with AIDS. Ultraschall Med. 2006;27:20–33. doi: 10.1055/s-2005-858649. [DOI] [PubMed] [Google Scholar]
- 3.Bhabhor K, Shastri M, Desai E. Abdominal ultrasound findings in HIV infected patients and its correlation with CD4+counts. Int J Sci Res. 2017;6:1632–7. [Google Scholar]
- 4.Pawar SS, Biradar SP, Gajbhiye MI, Khadse GJ. Correlation of hepatobiliary ultrasound findings with CD4 count and liver function tests in HIV positive patients. Med J West India. 2013;41:86–8. [Google Scholar]
- 5.Blessing OI, Tobechukwu TM, Ehimwenma O. Trans- abdominal ultrasonic findings correlated with CD4+counts in adult HIV-infected patients in Benin, Nigeria. SA J Radiol. 2009;13:34–40. [Google Scholar]
- 6.Mahmoud MZ, Abdulla AA, Elgyoum AM. AIDS effects in liver and biliary tract by using Ultrasound in Sudan. Afro Egypt J Infect Endem Dis. 2012;2:133–9. [Google Scholar]
- 7.Obajimi MO, Atalabi MO, Ogbole GI, Adeniji-Sofoluwe AT, Agunloye AM, Adekanmi AJ, et al. Abdominal ultrasonography in HIV/AIDS patients in South-Western Nigeria. BMC Med Imaging. 2008;8:5. doi: 10.1186/1471-2342-8-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Terzic D, Brmbolic B, Singer D, Dupanovic B, Korac M, Selemovic D, et al. Liver enlargement associated with opportunistic infections in patients with human immunodeficiency virus infection. J Gastrointestin Liver Dis. 2008;17:401–4. [PubMed] [Google Scholar]
- 9.Tshibwabwa ET, Mwaba P, Bogle-Taylor J, Zumla A. Four-year study of abdominal ultrasound in 900 Central African adults with AIDS referred for diagnostic imaging. Abdom Imaging. 2000;25:290–6. doi: 10.1007/s002610000035. [DOI] [PubMed] [Google Scholar]
- 10.Porcel-Martin A, Rendon-Unceta P, Bascunana-Quirell A, Amaya-Vidal A, Rodriguez-Ramos C, Soria de la Cruz MJ, et al. Focal splenic lesions in patients with AIDS:Sonographic findings. Abdom Imaging. 1998;23:196–200. doi: 10.1007/s002619900322. [DOI] [PubMed] [Google Scholar]
- 11.Grumbach K, Coleman BG, Gal AA, Arger PH, Mintz MC, Arenson RL, et al. Hepatic and biliary abnormalities in patients with AIDS:Sonographic-pathologic correlation. J Ultrasound Med. 1989;8:247–54. doi: 10.7863/jum.1989.8.5.247. [DOI] [PubMed] [Google Scholar]
- 12.Yee JM, Raghavendra N, Horii SC, Ambrosino M. Abdominal sonography in AIDS. A review. J Ultrasound Med. 1989;8:705–14. doi: 10.7863/jum.1989.8.12.705. [DOI] [PubMed] [Google Scholar]
- 13.di Fiori JL, Rodrigue D, Kaptein EM, Ralls PW. Diagnostic sonography of HIV-associated nephropathy:New observations and clinical correlation. AJR Am J Roentgenol. 1998;171:713–6. doi: 10.2214/ajr.171.3.9725302. [DOI] [PubMed] [Google Scholar]
- 14.Janakiraman H, Abraham G, Matthew M, Kuruvilla S, Panikar V, Solomon S, et al. Correlation of CD4 counts with renal disease in HIV positive patients. Saudi J Kidney Dis Transpl. 2008;19:603–7. [PubMed] [Google Scholar]
- 15.Adeyekun AA, Unuigbe EI, Onunu AN, Azubike CO. Renal sonographic parameters in human immunodeficiency virus- infected subjects and relationship to CD4 cell count. Saudi J Kidney Dis Transpl. 2011;22:1164–8. [PubMed] [Google Scholar]
- 16.Harriet NK, Michael GK, Chris K, William W, Robert C. Abdominal sonographic findings in severely immunosuppressed human immunodeficiency virus-infected patients treated for tuberculosis. J Tuberc Res. 2014;2:65–74. [Google Scholar]
- 17.Lito P, Pantanowitz L, Marotti J, Aboulafia DM, Campbell V, Bower M, et al. Gastroenteropancreatic neuroendocrine tumours in patients with HIV infection:A trans-Atlantic series. Am J Med Sci. 2009;337:1–4. doi: 10.1097/MAJ.0b013e31817d1cb7. [DOI] [PubMed] [Google Scholar]