Relapse remains one of the leading causes of treatment failure after allogeneic HCT and typically carries a poor prognosis. DLI is a strategy that is offered to relapsed patients in this situation but this form of adoptive immunotherapy can incur significant toxicity as acute GVHD and marrow aplasia are the leading causes of nonrelapse mortality after DLI12
. Further, although DLI has been extremely effective in the treatment of CML, this treatment modality has been less effective in the treatment of other hematologic malignancies13–15
Our Phase I study demonstrated the feasibility and safety of allogeneic CIK cell infusions in patients with relapsed hematologic malignancies. This Phase I dose escalation trial reached a planned maximal cell dose of 1 × 108 CD3+ cells/kg as the maximum tolerated dose. We observed only 2 cases of acute GVHD, grade II, that responded to topical or systemic corticosteroids and one patient developed limited chronic GVHD. The most serious adverse events seen were at the lowest cell dose whereas 2 patients developed transient ventricular arrhythmias. The etiology of the arrhythmias was never elucidated.
The low incidence of GVHD in this clinical study parallels the low GVHD incidence seen in preclinical models as we have previously demonstrated that the adoptive transfer of allogeneic CIK cells in a rodent model induced minimal GVHD9
. With the use of bioluminescence imaging, it was shown that luciferase-expressing CIK cells generated from splenocytes exhibited traffic patterns similar to conventional T cells. However, compared to the conventional T cells, the CIK cells infiltrated GVHD target tissues much less, demonstrated a slower division rate, were less susceptible to apoptosis and produced high amounts of interferon gamma, a cytokine known to confer a protective effect against acute GVHD16
As this was a Phase I feasibility study in a very heterogeneous population of patients, it is difficult to draw conclusions regarding efficacy. For the 12 patients who received CIK infusions while in CR, the median time to progression was 6 months. This remission duration is notable considering that this was a high risk population with the most durable remissions observed in patients with lymphoid malignancies. From the time of CIK infusion, the median EFS and OS were 4 months and 28 months, respectively. Although almost all patients in this study underwent some form of cytoreduction prior to CIK infusion, five patients had a longer time to progression/relapse after CIK infusion compared to time to progression/relapse immediately after allogeneic HCT. CIK cell infusions had no impact on patients with rapidly progressive disease at the time of infusion. The impact of CIK infusion on donor chimerism was not assessable in most patients since eleven patients already exhibited full donor chimerism at the time of CIK infusion. However, one of 5 patients with mixed chimerism converted to full donor chimerism approximately 2 months after CIK infusion without the development of GVHD.
A previously reported Phase I study of similar design also described clinical responses in patients who had received allogeneic CIK infusions after post-HCT relapse17
. Introna et al.
administered allogeneic CIK infusions in 11 patients with hematologic malignancies and reported the achievement of a completed remission in 3 patients with 1 remission lasting for over 2 years at the time of the report. Two of the 3 patients converted to full donor chimerism but also developed extensive cutaneous GVHD that commenced soon after the documented clinical response. Unlike our study where patients received only 1 CIK cell infusion, most patients received multiple sequential infusions with cell doses ranging from 3 × 106
to 15 × 106
With regards to feasibility of culture and expansion, we generated a median 12 fold expansion of CD3+
cells and 31 fold expansion of CD3+
cells, similar to the expansion data reported by Introna et al
. Additionally, we characterized the percentage and number of CD3+
cells with 53% of cells expressing this phenotype after culture. In vitro
antitumor activity against various tumor targets was also confirmed in our study and as reported previously by our group and others9, 17–18
Other groups have manipulated CIK cells with the intent of improving specificity and enhancing cytotoxicity against various tumor targets. The cytotoxicity of CIK cells was significantly increased against B-NHL targets when co-cultured with the anti-CD20 antibodies, rituximab and GA10119
. Other investigators have incorporated bispecific antibodies with the goal of redirecting CIK cells and increasing killing against primary ovarian cancer cells and against B-cell ALL20–21
. The expansion of cord-blood derived CIK cells represents another promising source of adoptive immunotherapy with potential application for patients with relapsed malignancies after umbilical cord blood transplantation22
. We have also utilized CIK cells to deliver an oncolytic virus to the tumor bed in a preclinical model with remarkable efficacy23
. This approach is being developed for clinical translation.
In summary, we have presented the feasibility and safety of allogeneic CIK cell infusions in patients with relapsed hematologic malignancies. Although CIK cells require expansion under specific culture conditions, we and other groups have shown the feasibility of this approach and have obtained similar results regarding low observed toxicity and high cell yields. Although it is difficult to assess clinical responses in this setting due to the variable and high risk nature of the patient population under treatment, it should be noted that several responses extended past one year in this high risk patient population. As seen with DLI, CIK cells appear to induce the longest response duration in patients who had minimal residual disease at the time of infusion. The comparable efficacy of CIK cells to DLI in the relapsed setting is difficult to determine but our data and those of others have confirmed lengthy clinical responses accompanied with a low incidence of acute and chronic GVHD compared to DLI17
. Thus, our results suggest that CIK cells exerted a biological effect and warrants further investigation. We are conducting two follow-on studies with infusion of CIK cells after allogeneic HCT with reduced intensity conditioning. One study involves high risk MDS patients who receive CIK cells pre-emptively on day +42 and a second trial is eligible to patients with CLL who demonstrate mixed chimerism. In this second trial, both chimerism and molecular disease burden will be assessed.