Hematopoietic stem cell therapy for malignant diseases

Abstract

Allogeneic bone marrow or blood stem cell transplantation (SCT) has changed its face in the last two decades. The introduction of nonmyeloablative conditioning regimens has reduced procedure toxicity and allowed the application of SCT in patients and conditions in which SCT was not offered in the past. In this review we will summarize the changes and accomplishments achieved in the past years in the field of stem cell transplantation for malignant disorders.

• Donor lymphocyte infusion (DLI)
• graft‐versus‐host disease (GVHD)
• graft‐versus‐leukemia (GVL)
• graft‐versus‐tumor (GVT) effects
• nonmyeloablative conditioning
• nonmyeloablative stem cell transplantation (NST)
• reduced intensity conditioning (RIC)
• stem cell transplantation (SCT)

Introduction

Allogeneic bone marrow or blood stem cell transplantation (SCT) represents an important therapeutic tool for the treatment of a long list of otherwise incurable malignant diseases. Until recently, allogeneic SCT was used primarily to replace malignant, genetically abnormal, or otherwise deficient immunohematopoietic compartments using highly toxic myeloablative preparative regimens that were considered mandatory for eradication of all undesirable host‐derived hematopoietic elements 1. However, animal experiments done in the early 1960s 2, 3 and clinical observations confirmed that immune‐mediated graft‐versus‐leukemia (GVL) effects is highly dominant in the course of SCT 4. Furthermore, 20 years ago, our group documented that alloreactive donor lymphocytes infused after SCT in tolerant hosts (e.g. patients post‐SCT receiving T cells from the original donors), with no additional posttransplant chemotherapy, can eliminate leukemic cells even in patients fully resistant to maximally tolerated doses of chemoradiotherapy 5–7. In this review we will try to give an overview of past accomplishments and recent developments in the field of SCT as used for the treatment of malignant disorders.

Myeloablative regimens

Initial experience using SCT involved the usage of regimens based on myeloablative conditioning in order to achieve three goals: eradication of the basic disease, prevention of rejection (suppression of the host‐versus‐graft reaction), and niche creation in order to make ‘space’ in the recipient's marrow for the transplanted cells. This was employed by administration of high doses of irradiation and/or cytotoxic agents (usually including cyclophosphamide, busulfan, etoposide, or others). Unfortunately, high‐dose chemoradiotherapy was not always sufficient to cure the basic disease and was rather toxic with high transplant‐induced morbidity, mortality, and graft‐versus‐host disease (GVHD) 8, 9. Past experience suggests that an attempt to give more aggressive chemoradiotherapy is likely to cause increased toxicity with minimal or no improved disease‐free survival (DFS). Whereas a certain proportion of patients with relapsing or primary resistant leukemia or other hematological malignancies not expected to be cured with conventional chemotherapy may respond to more intensive, myeloablative chemoradiotherapy followed by ‘rescue’ with autologous bone marrow or blood stem cells, the vast majority are likely to be resistant or to relapse. Bone marrow or blood stem cells obtained from an human leukocyte antigens (HLA)‐matched family member or matched unrelated donor (MUD) are likely to be much more effective due to the GVL effects mediated by alloreactive donor lymphocytes, usually but not always in association with antihost responses causing GVHD 10–12. Taken together, myeloablative conditioning with the goal in mind to eliminate or minimize the number of resistant tumor cells or otherwise abnormal host cells, followed by allogeneic SCT was considered until recently the most powerful treatment available for patients with blood cancer not expected to be cured by any alternative therapy. A similar approach was considered the treatment of choice for a large number of nonmalignant diseases correctable by SCT. The intensity of the regimens used for SCT escalated with time, while procedure‐related toxicity and mortality increased accordingly. The increased incidence and severity of late complications also became an important issue in evaluating the quality of life of long‐term survivors. So, it became clear that newer modalities must be developed in order to improve the cure rate of patients with hematological malignancies, as well as to improve the quality of life of successfully treated patients. In parallel, more and more data was published concerning the effect of allogeneic lymphocytes on host, specifically malignant cells, and the clear association between GVHD and graft‐versus‐tumor (GVT) effects in the clinical setting.

Nonmyeloablative regimens

This novel approach to transplantation was based on the following philosophy and premises: 1) By comparing numerous protocols comprising a wide range of intensities for each of the cytoreductive components used for over 20,000 transplants reported to the International Bone Marrow Transplant Registry, no difference or clear advantage could be documented for different regimens administered as preparation for autologous SCT or allogeneic SCT, including or excluding total body irradiation (TBI) 9; 2) the importance of immune‐mediated reactions between donor‐derived immunocompetent, alloreactive T lymphocytes and host‐type tumor cells has been recognized to be of major therapeutic importance, accounting for the significantly better antitumor effects induced by allogeneic SCT as compared with autologous SCT and transplants from an identical twin (syngeneic) 11; and mostly, 3) by the remarkable therapeutic potential of GVL effects induced by donor lymphocyte infusion (DLI) pioneered by Slavin and colleagues in late 1986 5–7, 13–17. Altogether, it seemed that the therapeutic component of allogeneic SCT could be mainly attributed to GVL effects mediated by T cells rather than to chemoradiotherapy‐induced elimination of all tumor cells. The possibility to completely eradicate tumor cells (hematological malignancies as well as solid tumors) by adoptive allogeneic cell therapy in preclinical animal models 18–20 suggested that alloreactive T lymphocytes of donor origin may be the strongest tool available against tumor cells of hematopoietic origin. Hence, we have developed a working hypothesis suggesting that the main role of the transplant procedure may be in the induction of a state of host‐versus‐graft tolerance for accomplishing durable engraftment of donor‐derived T lymphocytes, thus providing alloreactive T cells the opportunity to recognize and eradicate, over time, late post‐SCT, host‐derived tumor cells or abnormal stem cells, optimally when the recipient is off immunosuppressive treatment. This working hypothesis prompted us and other centers to develop a new approach for the treatment of both malignant and nonmalignant hematological diseases, avoiding myeloablative conditioning, in order to improve the immediate and long‐term outcome of the patients by preventing or minimizing procedure‐related toxicity and mortality. Thus, several protocols were designed based on minimizing the intensity of the conditioning regimen to the range of nonmyeloablative levels, followed by infusion of donor stem cells, preferably granulocyte colony stimulating factor (G‐CSF)‐mobilized blood stem cells enriched with circulating T lymphocytes collected by apheresis, or alternatively with bone marrow cells. The main new component of most of the new regimen was based on the use of intensive immunosuppression with fludarabine pregrafting. Fludarabine‐based protocols initially introduced in MD Anderson in Houston and Hadassah Hospital in Jerusalem 21, 22 were found to constitute the optimal conditioning regimens for recipients of HLA‐matched sibling or matched unrelated donor (MUD) allografts, a finding that has since been confirmed by many transplant centers worldwide 23–25. Fludarabine is usually used in combination with low‐dose busulfan, cytoxan, or melphalan with or without anti‐T cell antibodies (ATG). A different approach was adopted later by the Seattle group. Their initial preparative regimen consisted of low‐dose total body irradiation (TBI), followed posttransplant by the concurrent administration of cyclosporine and mycophenolate mofetil, in an effort to reduce the graft rejection and GVHD rates. Because their preliminary results showed increased rejection rates, fludarabine was later included in the initial preparative regimen of TBI 2 Gy 26, 27. Recently, we have carried out successful trials on novel conditioning protocols with even less cytotoxicity of conditioning than the initially reported conditioning for the ‘classical’ nonmyeloablative SCT (NST) and achieved a good engraftment rate. We named this procedure ‘minimally ablative stem cell transplantation’ (MAST), or ‘microtransplant’ 28. These protocols were successfully used for elderly patients and those with poor performance status. Following NST based on nonmyeloablative yet immunosuppressive conditioning, which is well tolerated even by elderly individuals (no upper chronological age limit), donor lymphocytes can be accepted without rejection and thus can mediate GVL or GVT effects against residual malignant cells that have escaped chemotherapy. Basically, while the patient's own T lymphocytes ignore tumor cells, which is one of the reasons why the tumor continues to grow indefinitely, donor T cells, being foreign to the tumor, may recognize residual malignant cells and destroy them. This is, in essence, the entire rationale behind allogeneic cell therapy, which provides a curative option for patients with otherwise incurable blood cancer or lymphoma.

NST regimens have been adopted by the most transplant centers worldwide, and currently a significant percentage of patients, mostly the elderly or those with significant comorbidities, are enrolled in NST trials. There is a growing amount of data that shows that NST regimens can produce a high rate of engraftment, not only in the setting of HLA‐identical sibling donors but also with the use of matched unrelated donors, with lower transplant‐related toxicity and in patients previously considered ineligible as transplant candidates.

Key messages

• Allogeneic stem cell transplantation (SCT) is a curative therapeutic tool for numerous otherwise incurable malignant diseases, including hematological malignancies and solid tumor, mainly through a T cell‐mediated immune graft‐versus‐tumor effect.

• SCT by its self may be considered as a platform for cancer immunotherapy.

• Nonmyeloablative preparative regimens produce a high rate of engraftment, with lower transplant‐related toxicity and in patients previously considered ineligible transplant candidates, including elderly patients.

Differences between the nonmyeloablative regimens

For reasons of simplicity, in this review, any allo‐SCT transplantation with a regimen less intense than the standard myeloablative preparative regimens is named nonmyeloablative (NST). However, the various NST regimens differ in the degree of the myeloablation (suppression of the recipient's bone marrow hematopoietic production) and immunoablation (suppression of the recipient's ability to reject the graft). A subset of NST uses minimal myelosuppressive conditioning. Examples of minimally myelosuppressive (MAST) regimens are regimens consisting of fludarabine and cyclophosphamide, or TBI 2 Gy, or with very low dose of busulfan. Another larger subset of NST uses moderate intensity regimens. Examples of moderately myelosuppressive regimens are regimens consisting of fludarabine in combination with intermediate dose busulfan or melphalan, with or without the addition of immunosuppressive antibodies such as ATG, or MabCampath. Fast recovery of autologous hematopoiesis always occurs after minimally ablative regimens if donor cells are not infused, whereas this is not the case after moderately ablative regimens. Although minimally ablative regimens are minimally toxic, persistent mixed chimerism is very common, and withdrawal of posttransplant immunosuppression, or DLI, is sometimes necessary for conversion to complete donor chimerism. In contrast, clinical experience from the use of moderately ablative regimens showed that conversion to complete donor chimerism occurs spontaneously in most cases without any posttransplant immune manipulation.

Did nonmyeloablative preparative regimens reduce transplant‐related toxicity?

A few retrospective analyses comparing the outcome after allo‐SCT with NST preparative regimen versus conventional ablative conditioning are briefly summarized below. Analysis of this retrospective data leads to the conclusion that reduction of the intensity of preparative regimen is associated with significant decrease in regimen‐related toxicity (RRT), transplant‐related mortality (TRM), as well as in the incidence of GVHD. In acute leukemia patients, the reduction in toxicity was not translated into improved outcome due to an increase in relapse rate following NST conditioning. However, it should be taken into consideration that these were retrospective comparative studies in which the patients undergoing NST were preselected and tended to be older and sicker, and thus it is highly expected that such patient population will have inferior results. In contrast, in studies including patients with different hematological malignancies, the decreased TRM was translated into improved overall survival (OS), and DFS in patients treated with reduced intensity regimens. The major studies will shortly be described. Alyea et al. 29 performed a retrospective analysis of the outcome of 152 patients above the age of 50 years, with hematological malignancies undergoing NST or myeloablative allo‐SCT from HLA‐identical siblings and MUD. NST patients were more likely to have unrelated donors, a prior transplant, and active disease at transplantation. Despite these adverse characteristics in the NST group, the incidence of moderate to severe acute GVHD (aGVHD) was similar in the two groups, while nonrelapse mortality (NRM) was lower in the NST group. OS was improved in NST group at 1 year and at 2 years following SCT. Diaconescu et al. 30 compared the morbidity and mortality after a nonmyeloablative versus myeloablative allo‐SCT from HLA‐matched related donors in patients with hematological malignancies. In this study, 73 patients with a median age of 54 years received a NST regimen consisting of TBI 2 Gy either alone or with fludarabine, while 73 patients with a median age of 48 years received standard myeloablative conditioning (Cyclophosphamide‐TBI (Cy‐TBI), or Busulfan‐Cy (Bu‐Cy)). Again, NST patients were at higher risk for TRM than patients receiving myeloablative conditioning due to older age, longer time from diagnosis to transplant, higher pretransplantation Charlson comorbidity index (CCI) scores, and more often a prior transplant. Once more, despite these adverse prognostic factors, they experienced significant less RRT. This was translated into less NRM in the NST group. In multivariate analysis, the strongest factor predicting lessened RRT and NRM was the use of nonmyeloablative conditioning, while higher CCI scores predicted higher NRM. Sorror et al. 31 compared the morbidity and mortality after a nonmyeloablative versus myeloablative allo‐SCT from matched unrelated donors in patients with hematological malignancies. Sixty patients with a median age of 54 years received a NST regimen consisting of fludarabine and TBI 2 Gy. The group of NST patients was compared retrospectively with a group of 74 consecutive patients of median age of 41 years, who underwent allo‐SCT during the same time period. Once more, even though NST patients had significant more comorbidities, were older, and more often had failed a previous ablative allo‐SCT, they experienced less RRT. Furthermore, the incidence of grade II–IV aGVHD was lower in nonablative than ablative patients, in part because of significantly less grade III–IV aGVHD in the NST group. The 1‐year NRM was also significantly lower in the NST group in comparison with the group of patients who underwent ablative conditioning. Multivariate analysis showed higher comorbidity scores to result in increased toxicity and mortality.

Many others (including Valcarcel et al., Canals et al., Perez‐Simon et al., Mielcarek et al., Vela‐Ojeda et al., Massenkeil et al., Couriel et al., Dreger et al., and Aoudjhane et al. 32–40) have shown similar results while retrospectively comparing the results of myeloablative to nonmyeloablative conditioning regimens in various hematological malignancies. It is to be noted that to date, no prospective study comparing these two conditioning methods has been published.

Stem cell transplantation in high‐risk groups

The development of NST in preparation for SCT revolutionized the field and due to the decreased RRT and TRM, it resulted in extended application of transplant procedures in patients previously considered as unsuitable candidates because of age or concurrent comorbidities. After the initial encouraging results, some transplant centers tested the possibility to expand the upper age limit for SCT. Studies reporting the results of allo‐SCT with a NST regimen in patients with a median age above 55–60 years are relatively limited, and some are briefly summarized below. Wong et al. (41) explored the feasibility of unrelated donor allo‐SCT after reduced intensity conditioning in 29 patients above the age of 55 years with myeloid malignancies, most of them in an advanced stage of disease. Grade II–IV aGVHD was observed in 41%, while chronic GVHD (cGVHD) developed in 63% of the patients respectively. With a median follow up period of 27 months the probability in 1‐year of OS, event‐free survival (EFS), and NRM were 44%, 37%, and 55%, respectively.

We reported our initial experience of related and unrelated donor SCT (mostly with fludarabine‐based conditioning regimens) for patients above the age of 60 years 42. Eighteen patients with a median age of 63 with various hematological malignancies were included. Only one of these patients was in complete remission (CR) at the time of transplant, while the rest had active disease, most of them refractory to previous salvage chemotherapy. Most of these patients had other significant comorbidities than being aged, and three of them were transplanted from partially mismatched donors. All the patients experienced trilineage engraftment, and severe aGVHD was observed in three patients, while six patients developed cGVHD. With a median follow‐up of 13 months, the OS and TRM were 30%, and 33%, respectively. In a recent study in our institution, we reviewed our experience with the use of a single NST regimen in allogeneic SCT from related and unrelated donors in 37 patients aged 55 years and above with various hematological malignancies. The vast majority of the patients were heavily pretreated, with pretransplantation advanced disease. With a median follow‐up period of 12 months, the 1‐year OS and DFS were 55% and 55%, respectively, while the overall nonrelapse mortality was 35%. Patients transplanted from related and unrelated donors had similar outcome 42.

In a study by Bertz et al. 43 19 patients above the age of 60 years with acute myeloid leukemia underwent reduced intensity conditioning (RIC) allo‐SCT from matched related and unrelated donors. The median age of the patients was 64 years. Only three patients had CR at the time of transplantation. None of the patients experienced graft rejection, while 13 out of 19 achieved CR. Severe aGVHD and cGVHD developed in 10 and 10 patients, respectively.

Shimoni et al. 44 recently reported the results from two different institutions on the outcome of patients above the age of 55 years with various hematological malignancies, after RIC allo‐SCT from matched unrelated donors. Again, aGVHD and cGVHD incidence was comparable to the reported incidence in younger patients undergoing allo‐SCT as well as the OS, DFS, and NRM.

In conclusion, it was shown that the reduction of transplant‐related toxicity achieved by NST made SCT available for this high‐risk patient group that was previously considered as an absolute contraindication for SCT. We expect that, with the improvement of medical care prolonging life and the continuing improvement of the general well being in the elderly population, more and more septuagenarians and even octogenarians will undergo SCT, as malignant diseases including hematological malignancies are very common in this population.

Stem cell transplantation and cell therapy in patients with solid tumors

One of the important changes in SCT in the last few years was the GVT effect, against solid tumors such as renal cell carcinoma, breast cancer, germ cell tumor, ovarian cancer, and lung cancer.

One of the first reported cases suggestive of GVT in patients with solid tumors was by Eibl et al. 45. They reported a 32‐year‐old woman with inflammatory breast carcinoma that underwent SCT from her HLA‐identical sibling. During the evolution GVHD, cytotoxic T lymphocytes were grown and tested in a chromium release assay against B and T lymphocytes of the patient and donor and against a panel of breast cancer cell lines. Resolution of liver metastases was observed simultaneously with clinical GVHD in the first weeks after transplant. In addition, minor histocompatibility antigen (MiHA)‐specific and major histocompatibility complex (MHC) class I antigen‐restricted cytotoxic T lymphocytes recognizing breast carcinoma target cells were isolated from the blood of the patient. Pretreatment of such target cells with tumor necrosis factor (TNF)‐alpha but not with interferon (IFN)‐alpha or IFN‐gamma increased the susceptibility of these cells to lysis by cytotoxic T lymphocytes.

Following this experience, trials of allogeneic SCT in various types of solid tumors were done, including refractory metastatic renal cell carcinoma (RCC) 46, 47, chemoresistant ovarian tumors 48, and metastatic breast cancer 49, 50.

We will highlight some important and interesting points that arose from these studies. Childs et al. 46 treated patients with refractory metastatic RCC with matched donor by NST. Nine of the 19 patients were alive 287–831 days post‐SCT. In ten patients the disease regressed: three had a CR, and seven had a partial response (PR). Regression of metastases was delayed, occurring at a median of 129 days after transplantation, and often followed the withdrawal of cyclosporine and the establishment of complete donor‐T‐cell chimerism. These results were similar to those of Nakagawa et al. 47. Bay et al. 48 presented their preliminary results of five patients with malignant ovarian carcinoma refractory to chemotherapy. Four of the patients presented with aGVHD or cGVHD associated with tumor regression of at least 50% as measured by CA‐125 levels and computerized tomography (CT) scans. Of the four transplantation survivors, three received a nonmyeloablative regimen which did not seem to reduce treatment effectiveness. DLI was administered to two patients. These infusions seemed to promote GVHD which was able to control disease progression for one patient and had no apparent effect on the other. In another study Ueno et al. 49 transplanted 8 patients with metastatic breast cancer (BC) and 15 with metastatic RCC with allogeneic SCT after a RIC regimen. Disease response was seen in ten patients (45%), with three CR, two PR, and five minor responses. Kanda et al. 51 suggested that pancreatic cancer may be sensitive to GVT effects after allogeneic reduced‐intensity SCT. An objective response on CT scan was observed in two patients, and another had a tumor marker response. Marked tumor shrinkage was observed in one of the remaining patients after donor lymphocyte infusion. Once more, as shown by Childs 46, these antitumor effects appeared after the effect of the conditioning regimen had disappeared. Pedrazzoli et al. 52 carried out a pilot trial of allogeneic transplantation after a RIC regimen in patients with refractory malignancies. Patients who had a poor performance status (PS) prior to SCT experienced grade 4 hematological toxicities and grade ≥3 organ toxicities and died of either treatment‐related complications or disease progression within 100 days of transplantation. By contrast, 10 of 11 patients who had an Eastern Cooperative Oncology Group performance status score (ECOG PS) of 0–1 prior to SCT experienced only short‐lasting neutropenia and thrombocytopenia and no organ toxicity. Among patients with a follow‐up >100 days, two CR and three transitory PR were recorded. This study emphasizes an important point by showing that SCT as a therapeutic strategy may provide little benefit in patients with poor PS and rapidly progressing disease. Additionally, one case report documenting GVT with non‐small‐cell lung carcinoma 53 appears in the literature.

In vitro data suggestive of T cell‐dependent GVT were described by Eibl et al. 45 and Kurokawa et al. 54. Both showed, using cell cultures, that donor‐derived tumor‐reactive cytotoxic T lymphocytes can be isolated and stimulated against the patient's tumor without cross‐reactivity against nonmalignant host cells.

In summary, the response of solid tumors for SCT differs from hematological diseases in its time frame. While the response of hematological malignancies is quite fast, solid tumor initial response may take months. This naturally changes the patient population that may be referred for such a procedure. It seems that only patients with life expectancy of at least a year and good PS will have a chance of SCT benefit. This should call for a change in the treatment attitude of oncologists, in order to facilitate early referral to SCT.

Future directions

Four important issues are still unsolved in the relationship with allogeneic SCT: 1) the toxicity of acute and chronic GVHD and the lack of methods to separate beneficial GVL and GVT from untoward GVHD, leading to transplant‐related morbidity and mortality; 2) absence of family or unrelated HLA‐matched donor for approximately 50% of patients in need of SCT, thus requiring safer protocols for transplantation of haploidentically mismatched allografts; 3) high incidence of relapse, especially in patients with advanced disease, and especially those with lymphoid malignancies and metastatic solid tumors; 4) high incidence of infectious complications especially following transplantation of haploidentically mismatched allografts.

Future developments are anticipated to improve further the outcome with respect to the GVHD issue including: 1) selective in vitro depletion of donor cells with antihost alloreactivity should aid in the fast restoration of recipient immunity without increasing the incidence of GVHD 55; and 2) posttransplant immunotherapeutic manipulations with infusion of ‘ex vivo expanded’ activated Natural Killer (NK) cells, or T cells reactive against minor H Antigens restricted to hematopoietic lineage, or the use of monoclonal antibodies such as anti‐CD20 might be proved effective methods to dissociate the beneficial GVT from the deleterious GVHD effect 56–59. The issue of finding a donor for all patients in need may be solved by expanding the unrelated donor registries, cord blood bank, and improving the results of mismatched related transplantations. The latter may be done by applying selective in vivo depletion of host cells with antidonor alloreactivity and might be proved an effective method to allow transplants across major HLA‐barriers with the use of reduced‐intensity conditioning regimens 60. Other new methods for induction of unresponsiveness to haploidentically mismatched allografts may become available too, especially when the mechanism of fetal‐maternal tolerance is better elucidated. As for the anticipated high incidence of relapse in patients with advanced and resistant disease, it appears that the future lies in development of more innovative anticancer modalities rather than clinical application of more intensive and hazardous chemoradiotherapy. Development of more effect and more selective modalities against the malignant cells, while avoiding or minimizing antihost responses, was and will remain the most important next major challenge.