Subjects received 15.1 ± 2.3 mCi of FDG intravenously after fasting for at least 6 h. Fasting glucose values on the day of scanning were 910 ± 104 mg/L. FDG scans were acquired from the base of the skull to the inguinal region starting 61 ± 5 min after injection using an Advance scanner (General Electric, Milwaukee, Wisconsin, USA). Scans were acquired in two-dimensional mode at 8–10 min/section and attenuation corrected with a 3-min transmission scan. The brain was scanned separately using 10-min emission and 4-min transmission scans. Images were reconstructed using an iterative reconstruction algorithm.
Images were analyzed visually on a computer blinded to the subject cohort and other clinical information. Sites with increased activity beyond that seen physiologically were considered abnormal. In the stop group, baseline scans were interpreted first. Subsequent scans were compared to the baseline scan and each other.
Quantitative analysis was performed by determining the standardized uptake value (SUV) corrected for lean body mass [SUV = (nCi/g in region/nCi injected) × lean body mass] . Regions of interest (ROI) were drawn over the liver, spleen, lung, and gluteal muscle for determination of SUVmean. ROI were drawn over visualized lymph nodes, parotids, tonsils, L5 marrow, and testicles using an automated algorithm to determine the SUVmax and the SUVmean using a threshold of 80% of the maximum counts. When lymph nodes were not visualized, a ROI was drawn over the area where lymph nodes would normally reside. ROI were drawn over the entire brain and the SUVtotal of the brain was obtained. SUVtotal were obtained as described previously . Whole body retention of FDG (SUVtotal_wb) was determined by manually excluding regions of known high physiologic uptake of FDG (kidney, brain, and bladder) and then adding the SUV of all the other pixels greater than 2.5. A minimum SUV value of 2.5 was arbitrarily selected because some normal structures have values in this range and because this value is sometimes used to differentiate between benign and malignant disease. In addition, regions were drawn manually only over all lymph nodes identified visually and the total SUV was obtained (SUVtotal_lymph nodes).
The small number of evaluable subjects in the advanced group precluded statistical comparisons with these. Comparisons between the remaining groups were made using one-way analysis of variance with post-test comparisons using the Student–Newman–Keuls method. Comparisons between two variables were performed using the two-tailed t test. Correlation of viral load with SUVtotal_wb measurements and SUVtotal_lymph nodes (in HIV patients in whom lymph nodes were visualized) were performed using Pearson's Correlation coefficient; all HIV infected patients were included except for two advanced patients (patients 29 and 30), that were excluded for reasons outlined below. Statistical analyses were performed using SigmaStat (Jandel, San Rafael, California, USA).
We enrolled 30 subjects who each underwent an FDG scan; six of these underwent additional scanning after stopping ART (Table 2). Twenty-eight males and two females (mean age 38.9 ± 9.9 years) were enrolled (18 Caucasian, seven Hispanic, five Black). Seven subjects were excluded from further statistical analysis after their FDG scans were performed, although their data is included in Table 2. Two advanced group subjects were excluded because one (patient 29), on re-interviewing, gave a history of being on granulocyte colony stimulating factor  and another because of viral load ≤ 50 000 copies/ml (patient 30). Two control subjects were excluded because of a history of immune thyroiditis (patient 6) and for untreated, latent tuberculosis (patient 5). Both of these subjects had abnormal FDG scans. Two suppressed subjects (patients 17 and 19) who at screening had viral loads < 50 copies/ml actually had 60 and 99 copies/ml on the day of their scans. One of the stop ART subjects (patient 26) was excluded because of viral load of ≥ 50 copies/ml at baseline scanning (72 copies/ml). The mean CD4 T-cell counts and median viral loads for the early (n = 5), advanced (n = 2), and suppressed (n = 12) groups were 332 cell/μl and 86 262 copies/ml, 42 cells/μl and 104 442 copies/ml, and 575 cells/μl and < 50 copies/ml, respectively.
Visual analysis showed differences in FDG accumulation between certain groups (Table 2, and Fig. 1 upper panel). Control and suppressed subjects did not have abnormal FDG accumulation in nodes, with the exception of faint axillary uptake in one subject in each group. In contrast, the early group had uptake in nodes-typically in more than one region. Splenic uptake was greater than liver in the early group as compared to the control or suppressed group (P ≤ 0.05). The two evaluable subjects in the advanced group had uptake in multiple nodal regions similar to the early group. The advanced patients excluded also had nodal uptake (Table 2). Uptake in small bowel, large bowel, and cecum did not differ among the five groups and was similar to that often seen in patients with a variety of malignant and benign disorders. Mesenteric nodes were not visualized in any subject. The images of brain were grossly normal in all groups, except for subject 27 who showed focal changes consistent with prior head trauma. Of the seven evaluable subjects with early or advanced HIV disease, six had FDG uptake in upper nodes (cervical/axillary) and lower nodes (inguinal/iliac) and one had inguinal nodes only.
The mean and standard deviation of the SUVmax in hypermetametabolic nodes, irrespective of which group the patients were in, was 4.9 ± 1.6 and the maximum SUVmax was 7.3. FDG uptake in normal organs was compared quantitatively among groups using the SUVmean (Table 3) or SUVmax (data not shown). The SUVmean for lung, muscle, marrow, liver, spleen, and brain were similar between control, early, and suppressed subjects. Testicular uptake was higher in the controls than either the suppressed or early group. Uptake in tonsils and parotids was higher in the early group than in the control or suppressed groups (P < 0.05). The uptake in the submandibular salivary glands also appeared visually higher in the early group than in the control or suppressed groups, but was not quantitated because of overall low uptake and inadequate visualization in the latter two groups.
Although no statistical differences in the SUV of liver or spleen were observed, quantitative analysis confirmed the visual findings of higher uptake ratios in the spleen than in the liver in the early group. The spleen:liver SUV ratio was statistically different between the control (1.06 ± 0.1) or suppressed (0.931 ± 0.2) groups versus the early group 1.38 ± 0.4 (P < 0.05), but not different between the control and suppressed groups (P = 0.34).
Total whole body retention (SUVtotal_wb) of FDG using a threshold SUV of 2.5 was not different between controls and suppressed subjects (Table 3). In contrast, significantly higher SUVtotal_wb FDG levels were observed in early subjects compared to the suppressed subjects (P < 0.05). In addition, the early subjects had higher SUVtotal_wb of FDG than the control subjects, although this was not statistically significant. Total whole body FDG retention in the initial studies of all HIV subjects with high viral loads was significantly higher than those with suppresed viral loads: median SUVtotal_wb of 23644 versus 10580 (P = 0.034), respectively. When viral load was correlated with SUVtotal_wb, a significant correlation was observed (r = 0.814, P < 0.0001), but no significant correlations was seen with SUVtotal_lymph nodes (total FDG activity in visualized nodes (Fig. 2).
Four of the five evaluable stop subjects developed new FDG uptake in nodal sites on their second or third scans (Table 2, Fig. 1 lower panel). Subject 24 had a negative baseline scan and showed cervical node uptake at 13 days off ART that persisted on day 20, despite having no increase in viral load above 50 copies/ml. In this subject, viral load subsequently increased to 1269 copies/ml by 28 days and continued to increase thereafter. Subject 23 had persistently low viral loads < 50 copies/ml and negative scans, with SUVs in organs and the whole body on the baseline scan that did not statistically differ from those on his follow-up scans off ART (data not shown).
Lymphoid organs function as a major site of HIV replication and HIV infection results in nodal lymphocyte activation . It is not surprising that FDG, a surrogate marker of glucose metabolism, accumulates in activated lymph nodes since activated lymphocytes exhibit increased glucose utilization [23,24]. FDG accumulation has been noted in disorders associated with increased lymphocyte activation or inflammation [17,25–27]. FDG accumulation in lymph nodes of subjects with HIV infection has been reported . Normal individuals usually have no visible FDG accumulation in nodal tissue, as was typically seen in our controls and other studies.
Our study differs from prior reports of Scharko et al. and Iyengar et al. in several ways [16,17]. Subjects were stratified by CD4 T-cell counts, a well-established marker of disease stage, and for level of viremia. We include a group of patients who were taken off ART and studied serially with FDG. We quantitatively evaluated several extra-nodal tissues to assess how they differed between the various HIV groups. In addition, all HIV subjects were screened to be relatively healthy which was similar to Iyengar et al.'s study, but contrasted with Scharko et al., thus eliminating other confounding processes that could result in abnormal FDG imaging. For example, diarrhea is common in late-stage HIV disease and it is possible that prior reports of late-stage subjects having increased uptake in mesenteric nodes (a finding not seen in our study) may simply have represented the failure to screen out other causes of FDG uptake and not signify a true source of virus replication .
As a group, viremic subjects (advanced and early-stage) showed significant differences in FDG uptake when compared to control or suppressed subjects, supporting the hypothesis that increased FDG uptake was the result of active HIV replication. Our studies, similar to those of Iyengar et al., reported quantitatively on the uptake of FDG, whereas in the study of Scharko et al. there was no quantitative evaluation since attenuation correction was not performed. Compared to Iyengar et al. we did not find a good correlation between SUVtotal_lymph nodes and viral load (Fig. 2) . Nonetheless, there was a good correlation between viral load and elevated whole body FDG retention. The reason for the lack of correlation between lymph node uptake and viral load in our study compared to Iyengar et al. are not completely known but may have been in part related to differences in subject population who had different levels of viral load. In contrast to prior FDG studies in HIV patients, we describe several sites of increased extranodal uptake of FDG not previously reported. T-cell activation in tonsils that normalizes with ART has been reported previously and is likely the explanation for the elevated FDG findings in the early group . Evidence of splenic hypermetabolism was suggested by the 1.4-fold higher uptake in spleen compared to liver in early disease subjects versus the 0.9 to 1.1 ratio in the suppressed or control groups, similar to the findings of Iyengar et al. . The increased uptake in the parotids of the early group may be due to infectious processes, diffuse infiltrative lymphocytosis that sometimes involves the salivary glands in HIV-infected subjects, or increased nodal uptake in nodes within the parotids due to activation because of HIV replication [30,31].
Stage of disease, stratified by CD4 T-cell count, did not appear to affect the pattern of FDG biodistribution. Nodal chain FDG uptake in advanced subjects (mean CD4 cell count, 57/μl) was comparable to viremic early stage individuals (mean CD4 cell count, 465/μl). Furthermore, in evaluable subjects with advanced disease, colonic activity was similar to that often seen in healthy controls as well as patients with various malignancies. We did not see mesenteric or prominent bowel FDG uptake, a finding reported in late-stage disease by others . This discrepancy could reflect differences in cohort selection, as mentioned above.
Despite the ability of ART to suppress HIV viremia below the sensitivity levels of commercially available assays, there is strong evidence that HIV is not eradicated but, rather, continues to replicate at low levels. Likely explanations for this are either that therapy is not potent enough to completely shut down replication, or that areas of on-going replication are at pharmacologically privileged sites where required levels of drug cannot reliably be delivered [32,33]. Because the vast majority of subjects are not able to control HIV replication immunologically, HIV viremia invariable rebounds when ART therapy is withdrawn . In an attempt to locate areas of low-level, on-going replication during effective ART, we performed a series of scans in subjects with suppressed viremia who were withdrawing from ART. We hypothesized that new FDG uptake detected prior to developing viremia in the blood may represent sites of early HIV rebound and presumably the source of subsequent viral dissemination and peripheral viremia. In four of the five evaluable subjects in this stop group, new increased FDG uptake developed in lymph nodes after therapy interruption. However, only one of the subjects (24) had new FDG lymph node uptake prior to the development of viremia. FDG in other tissues remained similar to baseline (data not shown). Interestingly, the subject (23) whose serial scans remained persistently negative after ART withdrawal also failed to demonstrate a rebound in viral load. These findings suggest that FDG nodal uptake is closely linked to, and may actually precede, viral proliferation, and that after ART withdrawal, these nodes are sites of early proliferating virus that activate lymphocytes, thus causing increased FDG uptake. These in vivo results are consistent with the ex vivo finding that nodal lymphocyte activation and viral expression occurs rapidly after ART interruption in previously suppressed subjects .
In contrast to the significant difference in whole body FDG retention in subjects with early disease versus control or suppressed subjects, we did not see any significant differences between the baseline and follow-up studies of stop group subjects, although as noted above there were qualitative differences in lymph node chains. This lack of difference may be due to the subtle changes in lymph node uptake that occurred early and/or to the relatively low degree of viremia seen in the stop group compared to the early group (median 11851 copies/ml versus 86262 copies/ml, respectively).
In the suppressed group, the two evaluable LTNP did not have nodal FDG uptake, similar to the controls, and subjects with suppressed viremia on ART. In contrast, the one excluded LTNP subject (baseline viral load ≥ 50 copies/ml) had peripheral node uptake. It is unclear if this pattern was related to low detectable viremia (60 copies/ml) or perhaps to a nodal activation process associated with relative immunological control of viral replication, such as CD8 T-cell proliferation . The lack of increased signal from FDG in the tissue of suppressed or LTNP compared to control subjects (Table 3) show that the method is insensitive to detect low levels of viral replication.
This and previous reports have not addressed fundamental questions of the specificity and sensitivity of FDG to detect specific anatomical sites of HIV replication. These parameters must be defined prior to our being able to recommend FDG scanning for the non-invasive in vitro study of HIV infection. By carefully screening for healthy HIV-infected subjects, we feel that the increased nodal FDG uptake seen in viremic HIV-infected subjects, but not in HIV-negative controls or suppressed subjects, is directly a result of HIV replication. However, although areas of FDG uptake in viremic subjects clearly represent areas of lymphocyte activation, it is unclear if this activation is due to local HIV replication or, alternatively, to systemic factors that induce lymphocyte activation but are not necessarily produced in that given node. Similarly, in a given viremic subject, the biological reason that some nodal chains have FDG uptake and other do not is unclear. In subjects with significant viremia such as those in this study, presumably all nodes contain infected lymphocytes. Differential uptake of FDG may simply be a matter of sensitivity even if not pathologically enlarged, anatomically distinct nodal chains vary in size, and, as seen in nodal evaluation for lung cancer, the detection rate may be size-dependent . Moreover, because the FDG signal intensity is less in HIV infection than in most neoplastic processes, node size probably plays a more important role in sensitivity determination in these patients. Future studies will have to pathologically compare the degree of HIV replication between FDG positive and FDG negative nodes in order to better delineate the utility of this potentially powerful non-invasive method.
In summary, our findings suggest that HIV viremia is associated with increased nodal activation that is detectable by FDG. In subjects with suppressed viral loads there is limited lymphocyte activation and a corresponding lack of FDG accumulation in nodes. The FDG pattern seen in suppressed patients resembles that of HIV-negative controls. Further studies in subjects with advanced disease may be useful to clarify the role of gut or mesenteric nodes as a reservoir of HIV in late-stage infection. In addition, serial evaluation of subjects with HIV before and after starting ART may be useful to determine the speed at which nodal uptake resolves.
We thank Dr. Clara C. Chen for her review of the manuscript and her constructive comments and critique.
1. Pantaleo G, Graziosi C, Demarest JF, Cohen OJ, Vaccarezza M, Gantt K, et al
. Role of lymphoid organs in the pathogenesis of human immunodeficiency virus (HIV) infection. Immunol Rev 1994; 140:105–130.
2. Perelson AS, Neumann AU, Markowitz M, Leonard JM, Ho DD. HIV-1 dynamics in vivo: Virion clearance rate, infected cell life-span, and viral generation time. Science 1996; 271:1582–1586.
3. Wei X, Ghosh SK, Taylor ME, Johnson VA, Emini EA, Deutsch P, et al
. Viral dynamics in human-immunodeficiency-virus type-1 infection. Nature 1995; 373:117–122.
4. Ho DD, Neumann AU, Perelson AS, Chen W, Leonard JM, Markowitz M. Rapid turnover of plasma virions and CD4 lymphocytes in HIV-1 infection. Nature 1995; 373:123–126.
5. Silvestri G, Feinberg MB. Turnover of lymphocytes and conceptual paradigms in HIV infection. J Clin Invest 2003; 112:821–824.
6. Moanna A, Dunham R, Paiardini M, Silvestri G. CD4+ T-cell depletion in HIV infection: killed by friendly fire? Curr HIV/AIDS Rep 2005; 2:16–23.
7. Zaunders JJ, Cunningham PH, Kelleher AD, Kaufmann GR, Jaramillo AB, Wright R, et al
. Potent antiretroviral therapy of primary human immunodeficiency virus type 1 (HIV-1) infection: Partial normalization of T lymphocyte subsets and limited reduction of HIV-1 DNA despite clearance of plasma viremia. J Infect Dis 1999; 180:320–329.
8. Fleury S, Pantaleo G. T cell regeneration in HIV-infected subjects under highly active antiretroviral therapy (Review). Int J Mol Med 1999; 4:91–97.
9. Stellbrink HJ, Zoller B, Fenner T, Lichtenberg G, van Lunzen J, Albrecht H, et al
. Rapid plasma virus and CD4+ T-cell turnover in HIV-1 infection: Evidence for an only transient interruption by treatment. AIDS 1996; 10:849–857.
10. Wood GS. The immunohistology of lymph nodes in HIV infection: a review. Prog AIDS Pathol 1990; 2:25–32.
11. Smith PD, Meng G, Salazar-Gonzalez JF, Shaw GM. Macrophage HIV-1 infection and the gastrointestinal tract reservoir. J Leukocyte Biol 2003; 74:642–649.
12. Bissel SJ, Wiley CA. Human immunodeficiency virus infection of the brain: Pitfalls in evaluating infected/affected cell populations. Brain Pathol 2004; 14:97–108.
13. Muciaccia B, Uccini S, Filippini A, Ziparo E, Paraire F, Baroni CD, et al
. Presence and cellular distribution of HIV in the testes of seropositive subjects: an evaluation by in situ PCR hybridization. FASEB J 1998; 12:151–163.
14. Pomerantz RJ. Residual HIV-1 RNA in blood plasma of patients taking suppressive highly active antiretroviral therapy. Biomed Pharmacother 2001; 55:7–15.
15. Scharko AM, Perlman SB, Hinds PW, Hanson JM, Uno H, Pauza CD. Whole body positron emission tomography imaging of simian immunodeficiency virus-infected rhesus macaques. Proc Natl Acad Sci USA 1996; 93:6425–6430.
16. Scharko AM, Perlman SB, Pyzalski RW, Graziano FM, Sosman J, Pauza CD. Whole-body positron emission tomography in patients with HIV-1 infection. Lancet 2003; 362:959–961.
17. Iyengar S, Chin B, Margolick JB, Sabundayo BP, Schwartz DH. Anatomical loci of HIV-associated immune activation and association with viraemia. Lancet 2003; 362:945–950.
18. Migueles SA, Sabbaghian MS, Shupert WL, Bettinotti MP, Marincola FM, Martino L, et al
. HLA B*
5701 is highly associated with restriction of virus replication in a subgroup of HIV-infected long term nonprogressors. Proc Natl Acad Sci USA 2000; 97:2709–2714.
19. Zasadny KR, Wahl RL. Standardized uptake values of normal tissues at PET with 2-[fluorine-18]-fluoro-2-deoxy-D-glucose: variations with body weight and a method for correction. Radiology 1993; 189:847–850.
20. Akhurst T, Ng VV, Larson SM, O'Donoghue JA, O'Neel J, Erdi Y, et al
. Tumor burden assessment with positron emission tomography. Clin Positron Imaging 2000; 3:57–65.
21. Mayer D, Bednarczyk EM. Interaction of colony-stimulating factors and fluorodeoxyglucose F-18 positron emission tomography. Ann Pharmacother 2002; 36:1796–1799.
22. Pantaleo G, Graziosi C, Butini L, Pizzo PA, Schnittman SM, Kotler DP, et al
. Lymphoid organs function as major reservoirs for human immunodeficiency virus. Proc Natl Acad Sci USA 1991; 88:9838–9842.
23. Bental M, Deutsch C. Metabolic changes in activated T-cells — an NMR study of human peripheral-blood lymphocytes. Magnetic Resonance Med 1993; 29:317–326.
24. Brand KA, Hermfisse U. Aerobic glycolysis by proliferating cells: A protective strategy against reactive oxygen species. FASEB J 1997; 11:388–395.
25. Zhuang HM, Alavi A. 18-fluorodeoxyglucose positron emission tomographic imaging in the detection and monitoring of infection and inflammation. Semin Nucl Med 2002; 32:47–59.
26. Kubota R, Kubota K, Yamada S, Tada M, Ido T, Tamahashi N. Microautoradiographic study for the differentiation of intratumoral macrophages, granulation tissues and cancer cells by the dynamics of fluorine-18-fluorodeoxyglucose uptake. J Nucl Med 1994; 35:104–112.
27. Nowak M, Carrasquillo J, Yarboro C, Bacharach SL, Whatley M, Valencia X, et al
. A pilot study of F-18FDG positron emission tomography(FDG-PET) to assess the distribution of activated lymphocytes in systemic lupus erythematosus (SLE). Arthritis Rheumatism 2002; 46:S25.
28. O'Doherty MJ, Barrington SF, Campbell M, Lowe J, Bradbeer CS. PET scanning and the human immunodeficiency virus-positive patient. J Nucl Med 1997; 38:1575–1583.
29. Dyrhol-Riise AM, Voltersvik P, Olofsson J, Asjo B. Activation of CD8 T cells normalizes and correlates with the level of infectious provirus in tonsils during highly active antiretroviral therapy in early HIV-1 infection. AIDS 1999; 13:2365–2376.
30. Vargas PA, Mauad T, Bohm GM, Saldiva PHN, Almeida OP. Parotid gland involvement in advanced AIDS. Oral Dis 2003; 9:55–61.
31. Kazi S, Cohen PR, Williams F, Schempp R, Reveille JD. The diffuse infiltrative lymphocytosis syndrome: Clinical and immunogenetic features in 35 patients. AIDS 1996; 10:385–391.
32. Pomerantz RJ. Reservoirs, sanctuaries, and residual disease: The hiding spots of HIV-1. HIV Clinical Trials 2003; 4:137–143.
33. Lambotte O, Deiva K, Tardieu M. HIV-1 persistence, viral reservoir, and the central nervous system in the HAART era. Brain Pathol 2003; 13:95–103.
34. Davey RT, Bhat N, Yoder C, Chun TW, Metcalf JA, Dewar R, et al
. HIV-1 and T cell dynamics after interruption of highly active antiretroviral therapy (HAART) in patients with a history of sustained viral suppression. Proc Natl Acad Sci USA 1999; 96:15109–15114.
35. Orenstein JM, Bhat N, Yoder C, Fox C, Polis MA, Metcalf JA, et al
. Rapid activation of lymph nodes and mononuclear cell HIV expression upon interrupting highly active antiretroviral therapy in patients after prolonged viral suppression. AIDS 2000; 14:1709–1715.
36. Migueles SA, Laborico AC, Shupert WL, Sabbaghian MS, Rabin R, Hallahan CW, et al
. HIV-specific CD8(+) T cell proliferation is coupled to perforin expression and is maintained in nonprogressors. Nat Immunol 2002; 3:1061–1068.
37. Gupta NC, Graeber GM, Bishop HA. Comparative efficacy of positron emission tomography with fluorodeoxyglucose in evaluation of small (< 1 cm), intermediate (1 to 3 cm), and large (> 3 cm) lymph node lesions. Chest 2000; 117:773–778.
Keywords:© 2006 Lippincott Williams & Wilkins, Inc.
fluorodeoxyglucose; F-18; positron emission tomography; PET; HIV; HAART