Generation of HIV-1–specific CD8⁺ cell responses following allogeneic hematopoietic cell transplantation

Recent studies have shown that autologous hematopoietic cell transplantation (HCT) is feasible for treatment of HIV-associated lymphoma when HIV is controlled with highly active antiretroviral therapy (HAART).¹  This study investigates a nonmyeloablative (NM) HCT regimen, NMHCT,²  combined with uninterrupted HAART to determine the effect on the control of HIV replication and on the development of HIV-1–specific immune responses.

Eligibility included HIV-1–infected patients with hematologic malignancies on suppressive HAART. The transplantation protocol was approved by the Institutional Review Board of the Fred Hutchinson Cancer Research Center (FHCRC). Written informed consent was obtained in accordance with the Declaration of Helsinki.

The preparative regimen, 90 mg/m² total dose of fludarabine and 200 cGy total body irradiation, was administered as described.²  Patients received granulocyte colony-stimulating factor (G-CSF)–mobilized peripheral blood stem cells (PBSCs) from an HLA allele–matched related or unrelated donor. Prophylaxis for graft-versus-host disease (GVHD) consisted of cyclosporine (CSP) and mycophenolate mofetil (MMF).^2,3  HAART was given through conditioning and after HCT (Table 1).

Plasma HIV RNA levels, peripheral blood mononuclear cell (PBMC) HIV proviral DNA levels, and PBMC quantitative HIV cultures were obtained at least weekly.^4-6 

Peptides based on optimal EBV- and CMV-derived HLA class I–restricted epitopes,⁷  obtained from Mabtech (Mariemont, OH) were tested individually and in pools. The ELISpot assay has been described.⁸  A positive response was defined as a 2-fold or greater increase in the mean number of spot-forming cells (SFCs) of experimental wells compared with negative controls, provided the mean SFC/10⁶ cells in experimental wells was more than 50 after subtraction of negative controls.

HIV-1 coreceptor CCR5Δ32 genotype was determined using DNA restriction fragment length polymorphism analysis as previously described.^8,9 

The choice of HAART took into consideration potency, adverse effects, and potential drug interactions. Patent 1 was switched to efavirenz to provide a more effective regimen and to avoid possible nevirapine-mediated hepatotoxicity associated with immune reconstitution.¹⁰  Protease inhibitors were not used, to avoid drug interactions causing toxic levels of CSP and the azoles.¹  Although efavirenz induces P450 enzymes, we reasoned that the dosage of affected drugs could be adjusted upward. The risk of abacavir hypersensitivity was extremely unlikely, since neither patients nor donors expressed HLA-B*5701.¹¹ 

After transplantation, full donor CD33⁺ chimerism was established in both patients. The CD3⁺ subset of patient 1 converted from 51% donor at day +28 to 100% donor by day +80 after withdrawal of immune suppression for treatment of recurrent leukemia, detected at day +42. In patient 2, the proportion of donor CD3⁺ cells continues to increase. Reconstitution of CD4⁺ and CD8⁺ subsets (Table 1) was typical for NMHCT recipients, half of whom receive prolonged immunosuppressive therapy.²  Patient 2 has not experienced opportunistic infections or GVHD and remains alive more than 180 days after HCT. Patient 1 developed CMV reactivation on day +83, successfully treated with foscarnet. GVHD grade 3 developed on day +56, after sudden withdrawal of immune suppression to treat recurrent leukemia, and responded to additional immune suppression. On day +100, patient 1 was discharged home on MMF, tacrolimus, and tapering doses of prednisone. Seven months after HCT he developed severe bronchiolitis obliterans (BO) unresponsive to additional immune suppression. He died 3 months later of pulmonary failure related to BO, Pseudomonas aeruginosa sepsis, and mucormycosis of his sinuses while on multiple immune suppressive agents.

For both patients at all time points after HCT the plasma HIV RNA remained undetectable and no HIV was detected by viral cultures of PBMC (Table 1). In patient 1, PBMC proviral DNA was detected at baseline before HCT and at day +28, but thereafter became undetectable when he converted to 100% donor chimerism. In contrast, proviral DNA was detected at all time points evaluated in patient 2, who continued to have mixed donor-host chimerism of the T-cell subset. Both patients' donor cells expressed wild-type CCR5 coreceptor, not the CCR5Δ32 allele, which is associated with resistance to HIV infection.¹² 

These data show that HAART can be administered throughout NMHCT, including during conditioning, with control of plasma viral replication. Neither patient developed HIV-related complications, and, although our first patient did not survive, his death was due to transplantation-related complications of GVHD. In our experience before the advent of HAART, myeloablative HCT resulted in acceleration of HIV disease, persistent culturable virus, and high levels of HIV p24 antigen after transplantation.¹³  Other studies also demonstrated that myeloablative therapy alone did not eliminate viral reservoirs.^14-19  In 2002, Kang et al reported on 2 patients given NMHCT with gene-modified sibling donor PBSC.²⁰  HAART was discontinued 1 week before, then resumed after HCT. One patient developed an acute retroviral syndrome, with a 6-log rise in HIV load, which declined rapidly after reinstitution of HAART. He subsequently developed central nervous system toxoplasmosis and died of progressive Hodgkin lymphoma. The second patient developed no HIV-related complications and was alive at the time of report. Detailed immunologic studies were not reported.

To determine whether HIV-naive donor cells could target HIV-epitope specificities, we examined T-cell responses to multiple optimal epitopes restricted by the patients' HLA class I alleles (Table 2). PBMCs obtained from patient 1 before transplantation reacted with 6 of 26 epitopes tested that spanned 6 proteins, including Nef, Vpr, Pol, Env, Gag, and Tat. Response frequencies to individual epitopes ranged from 100 to 805 SFC/10⁶ PBMCs with a total response of 1545 SFCs/10⁶ PBMCs. HIV-1–specific CD8⁺ T-cell responses were not detected on day +56, but were detected on day +81. The total magnitude of response frequency was 664 SFCs/10⁶ PBMCs with response being targeted to 8 of 26 epitopes examined. The pattern of HIV-epitope specificities of posttransplantation donor cells was markedly different compared with pretransplantation recipient T cells. The inter-pretation of HIV-specific CD8 T-cell responses in patient 2, who has mixed donor-host CD3⁺ chimerism, is more complicated. Before transplantation, PBMCs reacted with 4 epitopes limited to Nef and Gag. After transplantation, 2 of these same epitopes were recognized, probably by residual host CD8 cells. In addition, T-cell responses were elicited by 2 new epitopes in RT and 1 new epitope involving Env.

Allogeneic HCT provides a novel platform to study the development of HIV-1–specific T-cell responses generated from HIV-1–naive donor cells in the setting of chronic controlled infection. We found that new HIV-1–specific CD8⁺ T-cell responses were generated early after HCT, despite the absence of plasma HIV-1 RNA, suggesting that plasma viremia is not necessary for the development of an HIV-1 T-cell response and that T-cells can be primed by HIV antigens expressed in lymphatic tissue with limited HIV replication. The observed shift in recognized epitopes demonstrates that naive donor cells are capable of generating de novo HIV-1–specific immune responses under these conditions. Allogeneic HCT also provides a platform to study the effects of conditioning and the establishment of a new immune system, including antigen-presenting and -responding cells, on the latent HIV-1 reservoir. The gradual loss of detectable proviral DNA after HCT in patient 1, who achieved full donor chimerism, suggests that the pool of latently infected lymphocytes declined after HCT and that the priming of HIV-1–naive T-cells occurred with limited and localized HIV replication and antigen expression. These findings may provide insights for development of HIV-1 vaccines or immune-based therapies.

The authors thank Deborah Sessions for assistance in preparation of the manuscript.

The study was supported by grants CA78902, CA18029, CA15704, AI64061, AI57005, and K08 AI 059173 from the National Institutes of Health (Bethesda, MD).

National Institutes of Health

Contribution: A.W. conceived and designed the study and wrote the manuscript; U.M. designed and performed research assays and analyzed data; R.H. wrote the manuscript; J.M. performed research assays and analyzed data; T.M. designed and performed research assays and analyzed data; S.R. provided analytical agents and analyzed data; R.C. designed and performed research assays and analyzed data; F.A. provided resources and critical review of the manuscript; L.C. designed and performed research assays and analyzed data; and R.S. provided resources and supervised the study.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Ann Woolfrey, MD, Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue N, Seattle, WA 98115; e-mail: awoolfre@fhcrc.org.