Autologous stem cell transplantation in HIV-positive patients affected by relapsed/partially responding lymphoma: let it be

KEYWORDS:

• Autologous stem cell transplantation
• relapsed/partially responding lymphoma
• HIV infection
• co-morbidities
• long-term events
• immune recovery

Autologous stem cell transplantation (ASCT) has acquired a well-established role in the setting of HIV-positive patients affected by relapsed or partially responding lymphoma (RPRL) [1–6]. In this population, ASCT has proved to be an effective, feasible, and safe procedure ensuring comparable survival outcomes in both HIV-positive and negative individuals [7].

Nevertheless, a significant difference in ASCT outcome has been reported in the literature, mainly as a result of a selection of the patients included in the studies [1–6]. Focusing in particular on two of these trials, among the 14 patients included in the study by Gabarre and colleagues [4], only 4 patients were in complete remission at 49 months and one presented lymphoma relapse at 36 months after ASCT. On the contrary, Krishnan et al. [5] reported 17 out of 20 patients as alive and in remission at a median follow-up of 31.8 months. It is noteworthy that in the former study [4] all included patients were relapsed or resistant; while in the latter [5], patients with high-risk disease in first complete remission were also eligible for ASCT.

A major breakthrough has been achieved by Diez–Martin et al. [7], who matched a HIV-negative and a HIV-positive cohort of patients on the basis of the following criteria: histology, Ann Arbor stage at diagnosis, non-age-adjusted International Prognostic Index at diagnosis, and disease status at ASCT. No difference in the incidence of relapse, progression-free survival, and overall survival was showed in the two matched populations, differing only for HIV status. Moreover, this study confirmed ASCT safety in the HIV-positive group: the higher incidence of early bacterial infections did not significantly affect overall survival.

However, the feasibility of ASCT in everyday clinical practice is still a controversial and not yet deeply investigated topic. The Gruppo Italiano Cooperativo AIDS e Tumori study [6] has shed some light on this subject. Only 27 out of 50 (54%) unselected enrolled HIV-positive patients affected by RPRL received ASCT. Consequently, a definition of which part of the HIV-positive population suffering from RPRL could actually undergo and derive a survival advantage from ASCT is still lacking. An unmet goal in the care of these patients is the lack of a validated and shared score predicting ASCT mortality risk before transplant. There is no agreement on established clinical tools to sharply separate patients who could benefit from ASCT from those with a higher risk of treatment mortality (non-relapse mortality). Analogous considerations on ASCT feasibility and on the best tools required for patients’ selection can be raised in the elderly population. In fact, in both HIV-positive and elderly population with RPRL, the amount and the severity of co-morbidities may play an important role, affecting the outcome of more intensive chemotherapy treatments. In our daily practice, at ASCT enrollment, we tried to extend to HIV-positive population the mortality score (HCT-CI), introduced by Sorror et al. [8] in the stratification risk of patients submitted to allogeneic hematopoietic cell transplantation. The preliminary analysis of our data (not published) has pointed out that among 74 consecutive unselected HIV-positive patients with RPRL recruited at the National Cancer Institute in Aviano (Italy) from January 2001 to July 2012, 27 (36.5%) showed a significant burden of co-morbidities (as represented by a HTC-CI score ≥5). In our case series, a HTC-CI score ≥5 calculated before salvage therapy start was the only parameter significantly associated with a higher incidence of mortality deriving from treatment toxicity. On the other hand, a favorable outcome in terms of disease-free survival and overall survival was significantly combined only to disease chemosensitivity (data not published). Our preliminary results strongly suggest that the estimation of a transplant benefit cannot ignore both an objective evaluation of co-morbidities and of lymphoma, through adequate prognostic scores, ensuring a pre-transplant assessment of all mortality risks. This evaluation should be crucial for ASCT choice: either a high co-morbidity burden or a disease chemoresistance should be criteria of exclusion from ASCT. Alternative and less toxic conditioning regimens or experimental procedures should be reserved to these ruled out patients, representing a numerically significant portion of HIV-positive population with RPRL. In the lack of alternative criteria to score transplant mortality, we believe that the HCT-CI proposed by Sorror et al. [8] could integrate lymphoma prognostic scores in the ASCT selection for HIV-positive patients suffering from RPRL. Recent reports [9,10] have shown that other scores such as the HIV-score, the age-adjusted International Prognostic Index for non-Hodgkin Lymphoma and in a more comprehensive way, the AIDS-related lymphoma International Prognostic Index correlate with prognosis in HIV-infected patients with lymphoma treated with frontline rituximab-containing chemoimmunotherapy. However, in our experience, in the context of salvage therapy, these prognostic scores fail to identify the proportion of patients with a better prognosis, whereas only the evaluation of disease chemosensitivity and co-morbidities weight has a role.

A particular mention has to be done for immunodeficiency-associated Burkitt Lymphoma (BL), accounting for 30% of lymphomas seen in HIV-positive patients. Owing to the high proliferation index and rapid doubling time, BL is best treated with short-duration, dose-intensive chemotherapeutic regimens that have minimum treatment delays and incorporate central nervous system prophylaxis and rituximab [11]. Moreover, very interesting results have been obtained by Dunleavy and colleagues [12] with the use of EPOCH-R (Etoposide, Prednisolone, Oncovin, Cyclophosphamide, Hydroxydaunorubicin, Rituximab)-based treatments in HIV-positive BL showing how a therapeutic approach, based on a long chemotherapy exposure time and not on an increase of the dose, could be similarly highly effective, allowing a reduction in treatment toxicity. However, effective rescue therapies are lacking for patients with BL that relapse or are refractory. In this population the role of ASCT is unclear. General implications deriving from the analysis of available data in the literature may be summarized as follows: (1) up-front ASCT may be a clinical option for high-risk BL patients achieving a disease complete response or at least a partial response after first-line chemoimmunotherapy; (2) an early ASCT after a second-line intensive chemoimmunotherapy may be a desirable but not always achievable goal for relapsed BL in HIV-positive patients; (3) chemorefractoriness or the presence of a progressive disease despite optimal chemotherapy should be a criterion of patient inclusion in a clinical trial with novel agents or submission to an allogeneic transplantation.

Once ASCT has been established as a very effective procedure inducing long-term lymphoma remission, the next step is the investigation of long-term events developed by HIV-positive patients after ASCT and their impact on survival. Consequently, the definition of a follow-up strategy for HIV-positive long-term survivors after ASCT is another important issue still to be addressed. On this topic, data from our group [13] are the first reported in the literature, though in a population limited in number. We focused on the appearance of the following events: lymphoma relapse, second cancer and opportunistic infections (OIs). Our choice was driven by the supposed increased frequency of these events in the HIV-positive population compared to the HIV-negative counterpart. Among 26 HIV-positive patients achieving a complete lymphoma remission, 9 patients developed OIs at a median time of 0.4 years, 5 secondary malignancies (with a majority of non-AIDS-defining cancers linked to a viral pathogenesis) at a median time of 2.2 years and 2 lymphoma relapses at 4.3 and 3.1 years, respectively, from ASCT. Accordingly, our data support surveillance of OIs early after ASCT (when OIs are more frequent) and of second cancers and lymphoma relapses later from ASCT. Moreover, in our experience, OIs and second cancers were treated and managed successfully and their appearance did not impact on survival while lymphoma relapse had a causative role in one long-term death.

Another important issue still to be investigated is the long-term prevalence of the metabolic syndrome in HIV-positive patients after ASCT, its association with cardiovascular (cerebrovascular, coronary, and peripheral) events and ultimately its impact on survival. Our efforts should be directed toward screening and early management of reversible features of metabolic syndrome anticipating severely disabling and fatal events.

Two reports from our group focused on immune recovery after ASCT in the HIV-positive population: the former [14] demonstrated a not-significant difference compared with the HIV-negative population; the latter [13], restricting to CD4⁺ T-cell dynamics, showed a significantly higher long-term amount compared with CD4⁺ T-cell count at cancer diagnosis. These data exclude a role of ASCT in the amplification of preexisting immune deficit tied to HIV infection and suggest a cooperation of Highly Active Antiretroviral Therapy and ASCT in immune restoration. Moreover, when, in our recent study [13], we looked for a correlation between immune recovery dynamics and events occurrence, we found a significant association between a lower CD4⁺ T-cell count at 6 months and 1 year and onset of OIs, taking place at a median time of 0.4 years from ASCT. On the contrary, no correlation was showed with second malignancies or lymphoma relapse. These results support a favoring role of a lower CD4⁺ T-cell count on OIs appearance. If taken as a whole, immune recovery is a very fascinating area of investigation that still remains under studied. In particular, at the moment, we have no data on the quality of recovery of different subsets of CD4⁺ T-cells, CD8⁺ T-cells, and B cells and on an increased heterogeneity of T or B-cell repertoire.

In conclusion, data are mature to let ASCT be the first option in the setting of chemosensitive HIV-positive patients with RPRL, though with a marginal role of co-morbidities. In the presence of a long enough follow-up time and the expectation for this population to grow-up in number, an important goal to be achieved for the coming years should be to collect long-term secondary events and to figure out their impact on patient survival. The quality of immune restoration developing in HIV-positive population after ASCT deserves further investigation.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.