Increased thymic mass and circulating naive CD4 T cells in HIV-1-infected adults treated with growth hormone

Objective: To determine whether treatment with growth hormone (GH) enhances thymopoiesis in individuals infected with HIV-1.

Cited Here...: Five HIV-1-infected adults were treated with GH for 6-12 months in a prospective open-label study. Immunological analyses were performed before GH treatment and repeated at 3 month intervals after GH initiation. Thymic mass was analysed using computed tomography with quantitative density and volume analysis. Analysis of circulating lymphocytes, including naive and memory T cell subsets, was performed using multiparameter flow cytometry.

Cited Here...: GH treatment was associated with a marked increase in thymic mass in all GH recipients. Circulating naive CD4 T cells also increased significantly in all patients during GH therapy, suggesting an enhancement of thymopoiesis.

Cited Here...: GH has significant effects on the human immune system, including the reversal of thymic atrophy in HIV-1-infected adults. De-novo T cell production may thus be inducible in immunodeficient adults.

Introduction

HIV-1 infection is associated with profound immunodeficiency marked by peripheral T cell depletion and defects in cell-mediated and humoral immunity. Recent progress in antiretroviral therapy (ART) makes the immunosuppression of HIV-1 disease partly reversible, although the degree of immune recovery appears to be more complete in some individuals than in others [^1-4]. As the thymus is the primary site of T cell development, the state of thymic function in HIV-1-infected individuals may be a pivotal factor in determining the potential for full immune reconstitution.

In HIV-1-infected adults, thymopoiesis may be impaired by age-associated atrophy [⁵] or by the immunodestructive effects of HIV-1 [^6-8], and such impairment may itself contribute to regenerative failure or accelerated disease progression [^9,10]. It appears, nonetheless, that a thymic reserve persists into adulthood and can be summoned. The reversal of age-dependent thymic involution has been observed [^11,12], and recent evidence suggests that thymic function may even be recoverable in HIV-1-infected adults [¹³]. Consequently, a therapeutic role may be found for immunomodulating agents that stimulate thymopoiesis and accelerate T cell production in the setting of immunodeficiency.

Animal studies have indicated that growth hormone (GH) plays an important role in mammalian thymopoiesis [¹⁴]. GH may act directly upon immune tissues or its effects may be mediated indirectly through insulin-like growth factor-1 (IGF-1). GH-deficient rodents exhibited thymic hypoplasia that improved with GH replacement [^15,16]. The administration of GH or IGF-1 to older rodents reversed age-related declines in thymopoiesis [^14,17,18] and accelerated immune reconstitution in immunodeficient animals [^17-20]. Although GH-deficient humans do not appear to be immunologically compromised [¹⁴], we hypothesized from these findings in animal models that GH treatment would stimulate thymopoiesis in HIV-1-infected individuals.

Methods

Study subjects

Five HIV-1-positive men, treated with GH for HIV-1-associated fat accumulation, were prospectively enrolled in our study. Details of the parent study have recently been reported [²¹]. Immune analyses were scheduled before GH treatment (baseline) and at 3, 6, 9, and 12 months after GH initiation. The effects of GH therapy were determined by the comparison of post-treatment with pre-treatment values. Comparative historical control data were derived from a separate observational study at our institution designed to monitor immunological changes in HIV-1-infected adults on effective ART. These individuals underwent immune analyses at similar intervals to GH patients but did not receive GH. The historical control subjects were selected to have baseline characteristics similar to the GH cohort. Selection was conducted according to the following pre-defined criteria: age greater than 35 years; CD4 cell count greater than 250 cells/μl; unchanged ART for one year or more; and the availability of data from at least three study visits spanning a one year period. All individuals who met these criteria were included. Study protocols were approved by the Committee on Human Research of the University of California, San Francisco, USA.

Growth hormone treatment

Individuals were treated with recombinant human GH (Serono Laboratories, Norwell, Massachusetts, USA) at an initial dose of 3 mg per day (30-40 μg/kg per day) administered by subcutaneous injection on an outpatient basis. The dose was reduced to 1.5 mg per day after the first 6 months of therapy, except in subject 1, in whom GH dose reduction occurred at month 2 as a result of persistent arthralgias.

Thymic analysis

All measurements of the thymus were performed by a radiologist blinded to the treatment status of the individuals being analysed. Computed tomography (CT) images of the thymus were acquired as previously described [¹³], and a thymic index (TI) was assigned by a radiologist according to the following grading scale: 0, no soft tissue; 1, minimal soft tissue, barely recognizable; 2, minimal soft tissue, more obvious; 3, moderate soft tissue; 4, moderate soft tissue of greater extent, almost mass-like; and 5, mass-like appearance, suggesting hyperplasia or thymoma. Computer-based density and volume analysis of the thymus was performed by transferring CT data to a GE Advantage Windows workstation (versions 2.1 and 4.0, GE Medical Systems Advanced Windows Workstation Training Program, Milwaukee, WI, USA). Contours of the anterior mediastinum were outlined using a trace beginning at the level of left brachiocephalic vein, extending beneath the sternum, to the lungs laterally, posteriorly to the anterior margins of the great vessels and aortic arch, and ending inferiorly along the anterior portion of the pulmonary artery and right ventricular outflow tract. A small field of view was used to maximize the precision of the trace. The pulmonary parenchyma and sternum were excluded. The anterior mediastinal and prevascular lymph nodes were not excluded but were inspected to confirm that their size was unchanged between examinations. The density estimation was reported in Hounsfield units (HU), referenced with respect to water (structures with density greater than water have HU > 0 and those less dense than water have HU < 0). Region-of-interest HU measurements of subcutaneous axillary fat (cursor width > 1 cm) were performed identically at each timepoint. Agreement between the radiologist-assigned TI and the computer-based density measurement was strong (r = 0.9;P = 0.0004).

Laboratory measurements

Circulating naive and memory CD4 and CD8 T cells were measured on whole-blood specimens using multiparameter flow cytometry. Analyses were conducted in the General Clinical Research Center Core Immunology Laboratory by personnel blinded to treatment status. Whole blood (100 μl) was incubated with fluorochrome-conjugated antibodies directed against phenotypic surface antigens (CD45-FITC/CD4-PE/CD8-ECD/CD3-PC5 or CD45RA-FITC/CD62L-PE/CD4-ECD/CD8-PC5; Beckman Coulter, Fullerton, CA, USA). Red blood cells were lysed after antibody incubation using the semi-automated TQ-Prep workstation (Beckman Coulter) according to the manufacturer's protocol. Data acquisition and analyses were performed using a Coulter Epics XL flow cytometer with System II software, v.3.0 (Beckman Coulter). Complete blood count, serum IGF-1 levels and HIV-1 viral load measurements in GH recipients were measured as described previously [²¹].

Statistics

Analyses were conducted using StatView version 5.0 software (SAS Institute, Cary, NC, USA). Paired t-tests were used to evaluate differences over time compared with baseline values in the human subjects. Unpaired t-tests were used to evaluate differences between baseline characteristics in the GH and control cohorts. A repeated measures mixed effects model was used to compare the GH-treated and control groups with respect to changes in naive CD4 cell percentage over time with Proc Mixed (SAS System for Windows, Version 8.2; SAS Institute).

Results

The effects of GH on thymopoiesis were investigated in a subset of five HIV-1-infected individuals enrolled in a prospective open-label study examining the effects of GH on HIV-1-associated fat accumulation [²¹]. Cohort characteristics are displayed in Table 1. The participants were all men with a mean baseline age of 52 years and a mean CD4 T cell count of 419 cells/μl. All participants had been on stable ART for 18 months or more at enrollment. Four out of five patients were being treated with a stable protease inhibitor (PI)-based regimen with complete or near-complete virological suppression (mean duration of PI-based therapy 2.9 years). The baseline ages, viral loads, and CD4 cell counts were not significantly different between GH and control subjects. Baseline naive CD4 and CD8 cell percentages were lower in GH recipients (P = 0.02 and P = 0.03, respectively). Five of the six control subjects were taking stable PI-based ART (mean duration of PI-based therapy 1.5 years). All control subjects had complete or near-complete virological suppression, which did not change over the study period.

Table 1

GH was administered for 6-12 months. As expected, serum IGF-1 levels increased with GH treatment from a baseline of 163 to 703 ng/ml at 3 months (P = 0.02), 671 ng/ml at 6 months (P = 0.005), and 594 ng/ml at 12 months (P = 0.005).

Effect of growth hormone on human thymus

There was a marked increase in thymic tissue in all subjects after 6 months of GH therapy. All GH recipients had thymic atrophy before the initiation of therapy as evidenced by the near-complete replacement of the thymus by fat on CT scan (mean TI = 1) (Fig. 1, left panels;Table 1). Repeat analysis after 6 months of GH revealed a prominent increase in dense thymus tissue in all GH recipients (mean TI = 4, P = 0.0002 compared with baseline) (Fig. 1, right panels;Table 1). Quantitative density and volume measurements at 6 months demonstrated a mean increase in thymic density of 86 HU (Fig. 2a) (P = 0.0008), and a mean increase in thymic volume of 88% (Fig. 2b) (P = 0.06). Increased thymic density and volume were evident as early as 3 months after GH initiation, and remained increased at the completion of therapy (Table 1;Fig. 2). No similar changes were seen in the thymus tissue of historical control subjects (data not shown). To determine whether the reversal of thymic atrophy was caused by a generalized lipolytic effect of GH, quantitative density analysis was performed on the axillary adipose tissue of each GH recipient. No increase in axillary adipose density was detected (mean change of +3 HU at 6 months and -9 HU at 1 year, Fig. 2a).

Fig. 1

Fig. 2

Effect of growth hormone on circulating lymphocytes

GH therapy was associated with an increase in the percentage and absolute number of naive CD4 T cells (Table 1). When compared with baseline values, the mean absolute gain in naive CD4 cell percentage was 6% at 6 months (P = 0.05), 10% at 9 months (P = 0.02), and 12% at 12 months (P = 0.03) (Fig. 3a). The naive CD4 cell percentage began to rise 3-9 months after GH initiation, and continued to rise over the final 6 months of therapy (P = 0.03, month 12 compared with month 6). Subject 5, who had a normal CD4 cell count, demonstrated the least gain in naive CD4 cell percentage. Naive CD8 T cells did not increase with GH treatment (Fig. 3b). No significant changes in naive CD4 cell percentage or naive CD8 cell percentage were seen in historical control subjects who were maintained on effective ART for a similar period of time (Fig. 3c and d, respectively). Using a longitudinal repeated measures model, the change in naive CD4 cell percentage differed significantly between the GH and control subjects (parameter estimate -8.12, P = 0.008). The magnitude of the association remained strong when the model was controlled for differences in baseline naive CD4 cell percentage between the two groups. GH administration was not associated with any significant changes in the percentage or absolute number of total CD4 or CD8 T cells during the study period (Table 1). Circulating B and natural killer (NK) cells were analysed prospectively in two individuals and were unchanged (data not shown). No significant changes were seen in other hematopoietic lineages measured by the complete blood count (data not shown).

Fig. 3

Effect of discontinuation of growth hormone

To date, two study participants have been off GH therapy for one or more years. Repeat thymus CT scans, performed after GH discontinuation, revealed a decrease to the pre-GH thymic density in both individuals (Fig. 4). Gains in naive CD4 T cells remained stable.

Fig. 4

Adverse effects of growth hormone

Two out of five patients in our treatment cohort experienced significant adverse effects from GH, including arthralgias and glucose intolerance. There was no discernible effect of GH on HIV-1 viral loads. Subject 2 developed worsening virological failure 6 months into the study. His viral load subsequently became undetectable after empiric modifications of his antiretroviral regimen, including the addition of a non-nucleoside reverse transcriptase inhibitor. Other GH recipients remained on a stable antiretroviral regimen throughout the study period and had no significant changes in viral loads.

Discussion

We show here, in a prospective analysis, that treatment of HIV-1-infected adults with GH is associated with increased thymic tissue as measured by thymic CT with quantitative density and volume analysis. Increased thymic mass was accompanied by an increase in circulating CD4+CD45RA+CD62L+ naive T cells, suggesting that GH enhances thymopoiesis. These findings are consistent with previous studies documenting the capability of GH to stimulate thymopoiesis in animals [^14-20], and establish that GH also has significant effects on the human immune system. Furthermore, these results suggest that thymic atrophy may be reversed pharmacologically in HIV-1-infected adults. In our experience, including longitudinal analysis of thymic function in over 45 HIV-1-infected adults (L.A. Napolitano, J.M. McCune, unpublished observations), such a reversal of thymic atrophy in HIV-1-infected adults on stable ART is unprecedented.

Although it is possible that the observed changes in thymic density were related to processes other than true thymic hyperplasia [²²], we believe that this is unlikely. First, the changes were specific to the thymus and were not detected in axillary adipose or in any other regions of the thorax. Second, increases in thymic density were temporally associated with increases in circulating naive CD4 T cells. Finally, we have observed that GH treatment also induces hyperplasia of the human thymus of the SCID-hu Thy/Liv mouse [^23,24], and preliminary studies suggest that the primary action of GH is directed towards prethymic hematopoietic cells or thymic stroma (L.A. Napolitano, data not shown). In conjunction with related reports in the literature [^17,18,25], these observations suggest that GH may enhance the differentiation, proliferation, or thymic engraftment of pre-thymic hematopoietic progenitor cells and thereby indirectly promote thymopoiesis. Additional investigations are in progress to elucidate more precisely the mechanisms underlying GH effects on human thymopoiesis.

GH therapy of HIV-1-infected individuals was associated with a gain in naive CD4, but not naive CD8, T cells. This finding is consistent with previous studies [^13,26,27], which suggested that the regeneration of phenotypically naive CD4 cells is dependent upon the thymus, whereas naive phenotype CD8 T cells might also arise in extrathymic sites. Our findings contrast with a previous study [²⁸], which found no effect of GH on circulating naive T cells in HIV-1-infected adults. However, GH treatment in that study was given for 12 weeks, whereas our subjects received 6-12 months of therapy. As the increase in naive CD4 T cells seen in our subjects did not occur until 3-9 months into GH therapy, it is likely that the limited duration of GH treatment or follow-up may account for the absence of naive cell changes in the earlier study.

Although we cannot exclude the possibility that post-ART changes contributed to naive cell gains in the GH cohort, the simultaneous finding of increased thymic mass suggests that the increase in naive CD4 cell percentage represents enhanced T cell production. When present, post-ART naive cell gains appeared to occur most rapidly within the first year of effective therapy [^1,3,29], and the low baseline naive CD4 cell counts in the GH subjects (on effective therapy for an average of 3 years) suggested that post-ART naive gains in this group were minimal. Furthermore, no significant changes in naive CD4 cell percentage were seen in untreated controls who had a similar ART history and no increase in thymic mass.

We do not believe that these findings support the general use of GH in the setting of immunodeficiency at this time. Certain limitations of this study, including the small number of treated subjects and the lack of a randomized control arm, require that these data be interpreted with caution. Whereas GH appeared to enhance thymopoiesis in this small cohort, it is possible that our study subjects experienced a more pronounced response to GH therapy because of their age, sex, disease status, or underlying condition of fat accumulation. Additional caution is warranted by the potential for adverse effects of GH therapy. GH use is contraindicated in a variety of conditions and, despite appropriate screening, two out of five patients (40%) in this study experienced significant drug toxicity requiring the discontinuation of GH. The role of GH therapy in HIV-1 disease thus requires further examination in the research setting. A larger, randomized study is underway to evaluate further the role of GH in the management of HIV-1 disease.

Nonetheless, the prospect of utilizing immunomodulatory therapy to enhance de-novo T cell production invites additional consideration regarding the potential application of such therapy to the clinical management of HIV-1 disease. Theoretically, some individuals might benefit from GH-mediated enhancement of thymopoiesis, whereas the potential toxicity of GH may outweigh its advantage in others. For instance, the initiation of effective ART has been associated with immune restoration, decreased morbidity, and decreased mortality in a large number of HIV-1-infected patients, and it is not clear whether GH treatment would offer any additional benefit to these individuals. Although it is possible that the enhancement of thymopoiesis might lead to a broader, more effective, T cell repertoire, the true clinical benefit of such gains could be difficult to establish in individuals with a normal or near-normal CD4 cell count. Similarly, the benefits of enhanced thymopoiesis may not be easily detected in individuals with poorly controlled HIV-1 viremia and ongoing T cell destruction. At present, we believe that HIV-1-infected patients on effective ART, with radiological evidence of thymic atrophy, and with a low CD4 cell count have the greatest chance of deriving immunological benefit from GH-mediated enhancement of thymopoiesis. In particular, individuals on effective ART who have not experienced a gain in CD4 T cells despite prolonged virological suppression (e.g. see Teixeira et al. [⁴]) may especially benefit from GH therapy. However, it is possible that such intervention may be ineffective if thymic or bone marrow reserves cannot be restored because of irreversible destruction caused by HIV-1 or advanced age. Additional studies are warranted to determine which patients might receive clinical benefit from GH therapy.

Conclusion

We have shown that GH treatment of HIV-1-infected adults is associated with a significant increase in thymic tissue and circulating naive CD4 T cells, suggesting that GH enhances human thymopoiesis. These data offer promising evidence that thymic involution may be reversed pharmacologically in human adults. If so, de-novo T cell production might be inducible in immunodeficient individuals.

Acknowledgements

The authors would like to thank the individuals who participated in this study, Project Inform, and other members of the Bay Area AIDS community, Cheryl Stoddart, Mary Beth Moreno, Jeff Harris, Mary Beth Hanley, Richard Pearce, Krishna Komanduri, and Barbara Chang for their kind assistance; UCSF/Macy's Center for Creative Therapies; the Gladstone/SFGH Core Immunology Laboratory, and the SFGH GCRC. GH was generously provided by Serono Laboratories.