HIV infection induces follicular hyperplasia and the formation of germinal centres in lymph nodes, which is necessary for the efficient trapping of viral particles. Both virions and infected cells are sequestered, and this seems to be associated with relatively low levels of viraemia during clinical latency [1,2]. Follicular dendritic cell (FDC) involution takes place, and the architecture of lymphoid tissue degenerates during intermediate to late HIV infection . The increase in viral load and disease progression seem to correlate directly with the collapse of lymphoid tissue [1,2].
The introduction of so-called highly active antiretroviral therapy (HAART) has changed the prognosis of HIV infection . This is as a result of the partial immune reconstitution achieved by most HIV-infected patients who start HAART. However, data on the anatomical changes in lymphoid tissue of patients under HAART are very few and contradictory. Zhang and coworkers  found a recovery of CD4 cells in lymphoid tissue, and a partial resolution of the pathological changes of the FDC network  one year after commencing HAART. The CD4 T cell counts of the patients in those studies were approximately 200/μl. On the other hand, other studies that included patients with CD4 T cell counts over 500/μl  and patients with a wide range of immunodepression  did not find morphological changes in lymphoid tissue. In such studies, the second biopsy was taken after 6 months of HAART. Theoretically, a partial recovery of lymphoid tissue might be feasible in patients with mild immunosupression after a longer follow-up. This normalization of lymphoid tissue could have been missed by previous studies as a result of an early second biopsy. Because of this, we evaluated the effect of a quadruple antiretroviral regimen after 48 weeks of therapy on the tonsil architecture lymphoid tissue of HIV-infected patients with CD4 T cell counts over or equal to 500/μl.
From June 1997 to February 1998 all asymptomatic HIV-infected patients with CD4 T cell counts over or equal to 500/μl, and a high compliance with the previous follow-up, seen at our unit were offered quadruple antiretroviral therapy. The CD4 T cell counts had to be documented on at least two occasions separated by a minimum of 4 weeks. In addition, eligible patients had to have had no previous antiretroviral therapy or previous treatment only with zidovudine. All patients gave written informed consent to be included in the present survey. The study was approved by the local ethics committee.
There were two therapy groups according to previous exposure to zidovudine: (i) antiretroviral treatment-naive patients received: zidovudine 600 mg/day, lamivudine 300 mg/day, saquinavir hard-gel 1200 mg/day, and indinavir 2400 mg/day; (ii) zidovudine-pretreated patients received: didanosine 400 mg/day, stavudine 40 or 30 mg/day depending on the weight above or below 60 kg, respectively, and saquinavir and indinavir at the same doses as the former group. Scheduled clinical visits were carried out 4 weeks before starting therapy, at baseline, 4 and 12 weeks afterwards, and every 12 weeks thereafter. Blood samples were drawn at every scheduled clinical visit. Tonsil biopsies were obtained by an experienced otorhinolaringologist within a week before starting therapy. A second tonsil biopsy was taken at 48 weeks of therapy.
This study was designed in late 1996. The initial objective was to evaluate the feasibility of HIV eradication administering a very potent antiretroviral therapy. Accordingly, quadruple therapy including two protease inhibitors was used. The combination of two of these drugs had not been explored in depth. The most investigated combination of ritonavir plus saquinavir was tolerated badly. We chose the indinavir plus saquinavir combination, which had been showed to improve the pharmacokinetic profile of saquinavir . The initial objective was abandoned at the end of the study period, when evidence supported latent HIV reservoirs unaffected by prolonged therapy that prevented HIV eradication .
Histology and immunohistochemistry
Tissues were immediately placed in buffered formalin. After 24 h fixation, they were embedded in paraffin using routine methods. Sections 4 μm thick were deparaffinized and rehydrated. Histological evaluations were first made on sections stained with haematoxylin-eosin. Only biopsies with an adequate amount of lymphoid tissue were selected for immunohistochemical and morphometric analyses. For immunohistochemistry, deparaffinized tissue sections were subjected to the antigen retrieval method using 1 mM ethylenediamine tetraacetic acid (pH 8.0) buffer in a pressure cooker. Sections were immersed in 3% hydrogen peroxide in water and then covered with 10% normal swine serum (Vector Laboratories, Burlingame, CA, USA) in Tris-buffered saline, to block non-specific reactivity. Incubation with primary antibody was performed overnight at 4°C in a humid chamber. The dilution and source of anti-human monoclonal antibodies (mAb) were as follows: CD4 (clone1F6; prediluted; Master Diagnóstica, Granada, Spain), CD8 (clone 4B11; 1 : 40; Novocastra Laboratories, Newcastle, UK), CD20 (clone L26; prediluted; Concepta, Barcelona, Spain), CD45RA (clone 4KB5; 1 : 40; Dako, Glostrup, Denmark), CD45RO (clone A6; prediluted; Concepta), follicular dendritic cell (clone CNA.42; 1 : 40; Dako), and proliferating cell Ki-67 antigen (MIB-1; prediluted; Concepta). Biotinylated secondary antibody and avidin-biotin-peroxidase complex (LSAB2 Kit, Dako) were applied following the manufacturer's recommendations. Diaminobencidine (Signet Laboratories, Dedham, MA, USA) was used for chromogenic development. Double Ki-67 antigen/CD45RA immunostaining was achieved by sequential immunoperoxidase (Ki-67 antigen) and immunoalkaline phosphatase (CD45RA, Dako) methods. Alkaline phosphatase was developed with fast red substrate (Dako).
Tonsil tissue sections were examined blindly to evaluate qualitative structural and immunohistochemical changes by two independent pathologists (M.A.J., D.S.), who were not aware either of the dates of tonsil biopsy, the clinical situation, the CD4 T cell counts nor the viraemia changes of the patients. On haematoxilin-eosin sections the histological picture was scored 1+ to 3+, according to the following definitions: score 1+, marked lymphocyte depletion, accumulation of plasma cells, and prominent vascularity, germinal centres poorly defined or absent; score 2+, partial lymphocyte depletion, small germinal centres with incomplete FDC network; score 3+, no lymphocyte depletion, prominent germinal centres with well-formed FDC network. Each mAb stained section was also scored 1+ to 3+. The scoring of immunohistochemical changes was defined as follows: score 1+, less than 25% of the cells were stained; score 2+, between 25 and 50% of the cells were stained; score 3+, over 50% of the cells were stained. Quantitative image analyses were performed semi-automatically. Bright-field microscope images were directly captured with a digital camera (Kodak MDS120, Life Technologies, Merelbeke, Belgium). Ten high power fields (×40 objective) from T zones were captured for each tonsil tissue section. Images were then analysed with the NIH Image 1.57 software (National Institutes of Health, Bethesda, MD), using available cell scoring macros. The total amount of lymphoid tissue and follicular zones were measured by planimetry from low power fields (×10 objective). The areas were marked on the monitor and quantitated in square pixels with the NIH Image 1.57 software by using the available area measurement tool.
Peripheral blood determinations
CD4 and CD8 T cell counts were determined by standard flow cytometry. The plasma HIV-1-RNA concentration was measured using a quantitative polymerase chain reaction method (HIV Monitor Test Procedure, Amplicor PCR Diagnostic; Hoffman-La Roche, Basel, Switzerland). The analytical sensitivity of the assay was 20 RNA copies/ml. All procedures were performed according to the manufacturers’ recommendations.
Continuous variables are expressed as median (range) and categorical variables as absolute numbers (percentage). The Wilcoxon test was used for comparisons of continuous variables with repeated measures. The scores of the two pathologists were compared using Cohen's kappa test. Statistical analyses were performed by using the SPSS software package (SPSS, Chicago, IL, USA).
Twenty patients were offered quadruple therapy during the study period, 17 of them accepted. Six patients were excluded from the analysis of lymphoid tissue changes because of insufficient or inadequate tissue sampling in at least one of the two biopsies. Two of them were not included as a result of deformation of the tonsil tissue during the biopsy. The remainder were excluded because some of the samples mainly consisted of mucosa and a small quantity of lymphoid tissue. The remaining 11 patients were evaluable for the study of changes in tonsil lymphoid tissue. Eight patients had a sexual transmission route, four were promiscuous heterosexual individuals and four were homosexual men, and three were intravenous drug users (IDU). Nine patients were men. The median age of the patients was 35 (29–65) years. The baseline median CD4 T cell counts were 658 (431–1178) cells/μl. The median log plasma HIV-RNA levels of the 10 patients with detectable viraemia was 3.9 (2.43–5.25) copies/ml. Seroconversion had been documented in one homosexual man and one IDU, 9 and 10 years before entry into the study, respectively. In the rest of the patients the median period of time since the first known positive HIV antibody test and inclusion in the study was 8 (2–11) years. The six excluded patients were three male IDU and three promiscuous heterosexual women, with a median period of time since the first known HIV antibody test and the performance of the baseline tonsil biopsy of 8.5 (3–10) years. Their baseline CD4 T cell counts and log plasma viraemia were similar to those of the included patients.
Seven patients were treatment-naive and received the zidovudine-based regimen. Four patients were zidovudine-experienced and were assigned to the stavudine plus didanosine-containing regimen. Most of the patients achieved maximal viral suppression and, in spite of high CD4 T cell counts, most patients showed higher CD4 T cell counts at 48 weeks than at baseline (Table 1). There were no differences in the response of zidovudine-naive compared with zidovudine-experienced patients. Some adverse effect was detected in all patients during their follow-up; however, these were mainly mild and only one patient discontinued indinavir because of renal lithiasis. He then continued with the same nucleoside analogues and saquinavir 1800 mg per day. One excluded patient was zidovudine-experienced. The changes in CD4 T cell counts and viraemia after HAART of the six excluded patients were similar to those of the study patients.
In general, baseline tonsil biopsies showed moderate to severe lymphoid tissue HIV-related injuries. A variable degree of cellular depletion, plasma cell accumulation and prominent vessels were thus observed in all cases, and in most of them germinal centres were absent (Fig. 1a). Vaguely formed follicular structures were seen in four patients. Three of the six excluded patients had an evaluable baseline biopsy that showed a histological score of 1 in all of them. Follow-up tonsil biopsies demonstrated some degree of improvement in all of the evaluable cases (Table 1), except for patient no. 4, in whom only cellularity improved. Lymphoid cell numbers increased in all patients, and germinal centres appeared in seven cases that did not show them at baseline (Fig. 1b). The recovery of germinal centres was confirmed by the immunohistochemical observation of both CD20+ cells (Fig. 1c,d) and FDC (Fig. 1e,f). This recovery of germinal centres was quantified by planimetry, measuring both poorly defined and well-developed follicular formations. The median log area of follicular zones was 4.72 (0–5.6) at baseline and 5.36 (5.02–5.83) square pixels at 48 weeks (P = 0.04). The median log area of total lymphoid tissue was 5.9 (5.55–6.07) square pixels at baseline and 5.97 (5.72–6.25) square pixels at 48 weeks (P = 0.3).
The semiquantitative evaluation of the immunohistochemical markers is shown in Table 1. The scores of the pathologists agreed in 96.2% of cases (P = 0.94).
The quantitative analysis of selected markers in T-dependent zones is shown in Table 2. An increase in CD45RA+ cells was observed, moreover most of the Ki67+ cells in T zones were CD45RA+Ki67+ cells (Fig. 1g,h). However, the proportion of these proliferating CD45RA+Ki67+ cells in T zones did not differ significantly between baseline and 48 weeks of HAART. CD20+ cells in T zones remained constant from baseline to follow-up sections. CD8 cell counts decreased and CD4 cells increased significantly.
The present study shows an unexpected range of moderate to severe lymphoid tissue lesions in mildly immunosuppressed HIV-infected patients. This lymphoid tissue damage was partly restored after 48 weeks of HAART. Moreover, this beneficial effect is caused by a composite of FDC network regeneration and repopulation of lymphoid tissue, particularly with naive cells.
Tonsil biopsies have been used to investigate HIV replication in lymphoid tissue because of the ease of access that permits repeated sampling. Comparable viral expressing-cells and trapped virions have been reported in bilateral biopsies of palatine tonsils [11,12]. They have also been valuable to assess modifications in CD4 cells under HAART . In the present study, a sampling bias is an unlikely reason for the observed changes, as the severity of baseline tonsil lymphoid tissue injury was quite uniform and the amount of lymphoid tissue in the baseline and follow-up biopsies was similar. Lymphoid tissue present in the tonsil biopsies was sufficient to allow histological evaluation of representative B and T areas and precise measurement of the area occupied by B-dependent zones and immunostained cell quantification in at least 10 high power microscopic fields of T-dependent zones. Other limitations of the present study are the small size of the study population and a possible selection bias, as patients were selected according to a high compliance with the previous follow-up. This is the reason for the overrepresentation of the sexual transmission group compared with the cohort of HIV-infected patients followed at our unit, where IDU account for 60% of the patients . On the other hand, this bias may account for the long period of time from the first known positive HIV test or seroconversion to inclusion in the study.
The baseline lymphoid tissue structural abnormalities contrast with the level of peripheral CD4 T cells. However, these severe lesions are seen in a wide range of CD4 T cell counts, from 200 to 500/μl . On the other hand, CD4 T cell counts and lymphoid tissue damage do not seem to have a strict correlation. A study of patients with CD4 T cell counts ranging from 6 to 616/μl thus revealed follicular hyperplasia in all but one AIDS patient . One possible explanation is that, at a given point, lymphoid tissue lesions could be a result of the time that tissues have been exposed to the continued deleterious effect of ongoing HIV replication. In this regard, the patients reported here were known to be seropositive for a long period of time.
After one year of effective HAART, the baseline pathological changes were partly reversed. Global cellularity increased and well-developed germinal centres were observed; this improvement was confirmed by the recovery of the FDC network and B cells. We quantified an increase in CD4 cells and naive cells, whereas CD8 cells diminished. However, the fraction of proliferating naive cells did not change significantly. These results agree with the observations of Zhang and colleagues [5,6] in more immunodepressed patients who, after a year of HAART, regained CD4 cells and CD4 naive cells , and recovered the FDC network .
Two previous studies did not find modifications in lymphoid tissue after 6 months of HAART [7,8]. Tenner-Racz and colleagues  found before starting therapy, among patients with CD4 T cell counts over 500/μl, follicular hyperplasia in enlarged lymph nodes and normal lymphoid tissue in small lymph nodes. The baseline CD4 T cell counts in peripheral blood and lymphoid tissue were similar to uninfected controls, whereas CD8 cell lymphoid tissue levels were higher. They also showed an elevated ratio of proliferating CD4 and CD8 cells in patients compared with controls. These abnormalities did not change after 6 months of therapy. Orenstein et al.  examined inguinal lymph nodes of patients in a wide range of the spectrum of HIV infection, from asymptomatic to symptomatic, and from very low CD4 T cell counts to near normal. As mentioned above, the authors found follicular hyperplasia in all but one AIDS patient. Only in this patient did they find a histological recovery after therapy; in the rest of the patients they found no changes. Therefore, no normalization of the lymphoid tissue would be expected if at baseline the morphology is fairly well preserved, or near normal, or if follicular hyperplasia is seen; perhaps because more time on therapy would be needed. In this sense, we found a reduction in CD8 cell lymphoid tissue infiltration in samples obtained after 12 months of HAART, but this reduction was not observed after 6 months of HAART by Tenner-Racz et al. .
The regeneration of lymphoid tissue histology reported here is caused, at least partly, by the repopulation of the lymphoid tissue with immune cells and the recovery of the FDC network. Both CD4 and CD45RA+ cell counts increased in T-dependent zones. These zones are colonized early in life by naive T lymphocytes, and in adulthood they seem to be the main reservoir of these cells. An expansion of naive cells was demonstrated at follow-up biopsies. However, a proliferation of pre-existing naive cells, measured by the counts of CD45RA+Ki67+, was not observed. A proliferative naive cell expansion might have occurred before or, alternatively, these cells might have migrated from the thymus. The regained T and B lymphocytes are necessary for the maturation of FDC, which may proceed from bone marrow precursors or from FDC remnants in the lymphoid tissue .
Our results suggest that early in the course of HIV, disease may not be just a matter of CD4 T cell counts. Two seroconverters in this study had a long duration of infection. In most of the remaining patients, the period of time from the first known positive HIV test to the tonsil biopsy was very long. It could thus be speculated that the duration of HIV infection might have an impact on lymphoid tissue damage not predictable by peripheral estimations. On the other hand, this study shows that serious lymphoid tissue lesions can be recovered with HAART.
1. Pantaleo G, Fauci AS. Immunopathogenesis of HIV infection. Annu Rev Microbiol 1996, 50: 825–854.
2. Pantaleo G, Graziosi C, Demarest JF. et al
. HIV infection is active and progressive in lymphoid tissue during the clinically latent stage of disease. Nature 1993, 362: 355–358.
3. Pantaleo G, Graziosi C, Demarest JF. et al
. Role of lymphoid organs in the pathogenesis of human immunodeficiency virus (HIV) infection. Immunol Rev 1994, 140: 105–130.
4. Pallela Jr FJ, Delaney KM, Moorman AC. et al
. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatients Study Investigators.
N Engl J Med 1998, 338: 853–860.
5. Zhang ZQ, Notermans DW, Sedgewick G. et al
. Kinetics of CD4+ T cell repopulation of lymphoid tissues after treatment of HIV-1 infection. Proc Natl Acad Sci U S A 1998, 95: 1154–1159.
6. Zhang ZQ, Schuler T, Cavert W. et al
. Reversibility of pathological changes in the follicular dendritic cell network with treatment of HIV-1 infection. Proc Natl Acad Sci U S A 1999, 96: 5169–5172.
7. Tenner-Racz K, Stellbrink HJ, van Lunzen J. et al
. The unenlarged lymph nodes of HIV-1-infected, asymptomatic patients with high CD4 T cell counts are sites for virus replication and CD4 T cell proliferation. The impact of highly active antiretroviral therapy.
J Exp Med 1998, 6: 949–959.
8. Orenstein JM, Feinberg M, Yoder C. et al
. Lymph node architecture preceding and following 6 months of potent antiretroviral therapy: follicular hyperplasia persist in parallel with p24 antigen. Restoration after involution in an AIDS patient.
AIDS 1999, 13: 2219–2229.
9. McCrea J, Buss N, Stone J, et al
. Indinavir–saquinavir single dose pharmacokinetic study. 4th Conference on Retroviruses and Opportunistic Infections
. Washington, DC, 1997 [Abstract 608].
10. Haase AT, Hery K, Zupancic M. et al
. Quantitative image analysis of HIV-1 infection in lymphoid tissue. Science 1996, 274: 985–989.
11. Cavert W, Notermans DW, Staskus K. et al
. Kinetics of response in lymphoid tissues to antiretroviral therapy of HIV-1 infection. Science 1997, 276: 960–964.
12. Wong JK, Hezareh M, Gunthard HF. et al
. Recovery of replication-competent HIV despite prolongued suppression of plasma viremia. Science 1997, 278: 1291–1295.
13. Macías J, Pineda JA, Leal M. et al
. Influence of hepatitis C virus infection on the mortality of antiretroviral treated patients with HIV disease. Eur J Clin Microbiol Infect Dis 1998, 17: 167–170.