Different preparative regimens

In this review, any transplantation that uses a conditioning regimen less than myeloablative is called 'RIST'. A subset of RIST that uses a minimal-intensity regimen is called 'Minimal-Intensity Stem Cell Transplantation (MIST)'. MIST regimens are truly nonmyeloablative, that is, the patient's own hematopoiesis can recover if donor stem cells are not infused. Examples of MIST conditioning regimens are fludarabine (Flu)/cyclophosphamide (Cy)^{5, 6, 16, 17, 18} and 200 cGy TBI with or without Flu.^{2, 19, 20} The other, larger subset of RIST is 'Moderate-intensity Stem Cell Transplantation (MOST)'. Examples of MOST conditioning regimens are Flu or cladribine (Cla)/busulfan (Bu) with or without anti-thymocyte globulin (ATG),^{3, 4, 21} and Flu/melphalan (Mel) with or without alemtuzumab.^{1, 22, 23, 24} For transplantation using intensive, myeloablative conditioning regimens, we use the term 'Myeloablative Stem Cell Transplantation (MAST)'. Figure 1 displays the intensity of these different regimens according to myeloablation and immunoablation.

Figure 1.

Diagram showing intensity of each regimen. The x-axis indicates the degree of immunoablation, and the y-axis of myeloablation. RIST: reduced-intensity stem cell transplantation, MIST: minimal-intensity stem cell transplantation, MOST: moderate-intensity stem cell transplantation, MAST: myeloablative stem cell transplantation.

Full figure and legend (43K)

An interesting early observation was the persistence of host cells for long periods after RIST, resulting in mixed donor/host chimerism (Figure 2). As different conditioning regimens were employed, the speed of conversion to complete donor type chimerism in different hematopoietic lineages was noted to vary among protocols. MAST conditioning regimens almost always cause complete donor chimerism in all hematopoietic lineages promptly after transplants, and the clinical courses following different MAST conditioning regimens are similar. RIST regimens vary widely in the intensity of myeloablation, immunoablation, and antitumor effects. Their toxicities vary widely, as does the rate of achieving complete donor hematopoietic chimerism. The mucosal toxicity of MOST regimens such as Flu/Mel 180 mg/m² is significant,²⁵ whereas a MIST regimen such as Flu/TBI 200 cGy may obviate the need for transfusions.²⁶ GVHD after MIST may also be less intense.²⁷ The characteristics of these regimens are summarized in Table 1.

Figure 2.

Schematic diagram of mixed chimerism. It turned out that stable mixed chimerism, which requires donor lymphocyte infusion (#) to be converted to complete donor chimerism, is relatively rare.

Full figure and legend (12K)

Table 1 - Comparison of transplantation.

Full table

Hematopoietic chimerism, RIST, and DLI

Initial clinical results suggested that persistent mixed chimerism after RIST was inevitable and, thus, additional DLI was considered to be necessary in order to achieve complete donor chimerism. Indeed, some MIST regimens do not completely eradicate host hematopoiesis^{2, 20} and some protocols incorporate DLI from the outset.²⁸ Additional clinical experience suggested, however, that DLI was often not necessary to achieve complete donor chimerism following MOST regimens. Thus, MOST is similar to MAST in not needing DLI. DLI also caused toxicities, including severe GVHD.^{29, 30} Marks et al³⁰ reported a 25% incidence of grades II–IV acute GVHD after DLI, and 15% of grades III–IV. As a consequence, at most institutions, DLI is currently used only following relapse or after an increase of host cells during a period of mixed chimerism. If T-cell depletion is performed in vitro or in vivo, the probability of incomplete donor-type chimerism is higher and, in such cases, DLI may be necessary.^{31, 32}

Kinetics of achieving complete donor chimerism varies among transplant regimens. MAST regimens usually achieve complete donor chimerism in all lineages within 1 month (Figure 3a). In MOST, complete donor chimerism in the myeloid lineage is almost always achieved promptly, although host T cells may persist, causing mixed chimerism for up to 2–3 months^{33, 34, 35} (Figure 3b). With MIST, complete myeloid engraftment can be somewhat slower, taking as long as 3 months to achieve (Figure 3c). In one study, complete donor T-cell chimerism preceded complete donor myeloid chimerism after MIST,⁶ but it is unclear whether this difference is clinically important. Mohr et al³⁶ reported that a variety of factors, including the number of transplanted CD34 cells, the conditioning regimen, and the number of previous chemotherapy regimens may all determine the kinetics of achieving complete donor chimerism (Table 2).

Figure 3.

Diagram of chimerism status after stem cell transplantation using each regimen. (a) Myeloablative stem cell transplantation (MAST). (b) Moderate-intensity stem cell transplantation (MOST). (c) Minimal-intensity stem cell transplantation (MIST) with fludarabine/TBI 200 cGy. A solid line represent for myeloid lineage, a broken line for T-cell lineage.

Full figure and legend (23K)

Table 2 - Factors affecting donor–host chimerism.

Full table

Does GVHD decrease with RIST compared to MAST?

The severity of GVHD is related to the intensity of the conditioning regimens and, thus, some authors predicted milder GVHD after RIST.³⁷ In addition, stable mixed chimerism after RIST infers the tolerance of donor cells to host elements, and one study reported a low GVHD rate in the setting of mixed chimerism.³⁸ In a second study, early complete donor T-cell chimerism was associated with higher incidence of grades II–IV GVHD.³⁵ The incidence of GVHD thus appears to be less in MIST than in MAST,²⁷ but equivalent between MAST and MOST.^{10, 39, 40}

In animal models, the induction of GVHD requires host antigen-presenting cells (APCs).^{41, 42, 43} It is thus possible that mixed chimerism in APC populations may induce more GVHD than complete donor chimerism.⁴⁴ Such mixed chimerism might help to explain why GVHD is similar after MOST despite the decreased intensity of conditioning.

Given the risk of GVHD after MOST regimens, some centers have added ATG or anti-CD52 antibody (alemtuzumab) to the preparative regimen. Slavin et al⁴ added ATG to Flu/Bu, and reported excellent results, but the Dresden group observed comparable engraftment rate without ATG, suggesting that it was not necessary.²¹ ATG or alemtuzumab has now been used in many RIST regimens.^{3, 22, 24, 29, 32, 39, 45, 46, 47, 48} These agents have a very long half-life and, thus, suppress not only host immunity responsible for graft rejection but also act as 'in vivo T depletion' of the graft. As the results of studies vary widely, it is unclear whether these agents affect rates of acute GVHD,^{32, 46, 47, 48} chronic GVHD,^{32, 46, 47, 48} donor engraftment,^{32, 36, 47} infections,^{32, 46} or relapse.^{46, 49} Although some studies claim that the use of ATG is associated with a poor prognosis,^{45, 46} other studies have not found such an association.^{32, 48} In a few studies, the use of ATG increased relapse.^{46, 49} A reasonable approach would be to use ATG only in patients at low risk for relapse. The best use of those potent agents in the context of RIST will need to account for dose, timing of administration, and disease status.

RIST for acute myeloid leukemia (AML)

To date, many studies of RIST have included a mixture of diseases. In addition to the difference in efficacy of chemotherapeutic agents for different malignancies, the potency of GVL effects also varies from disease to disease. Early reports of RIST focused on patients with, primarily, myeloid diseases, because the GVL effect was generally thought to be stronger in myeloid diseases.^{1, 2, 3, 4, 21, 23}

In AML, most published reports of RIST have used MOST regimens. In one report of 19 elderly patients (median age 64)⁵⁰ with advanced AML (median blast percentage was 50%), 1-year overall survival (OS) was 68% and 1-year disease-free survival (DFS) was 61%. Martino et al⁵¹ reported a 66% 1-year DFS in 37 patients (17 AML and 20 myelodysplastic syndrome (MDS)). Sayer et al⁵² reported 1-year OS and DFS of 47% in 113 patients, most of whom received Flu/Bu. Hamaki et al⁴⁹ used a Flu or Cla/Bu with or without ATG regimen in 36 AML and MDS patients, with 1-year DFS of 64% in high-risk patients and of 85% in low-risk patients. Using unrelated donors and Flu/Mel conditioning, Wong et al²⁵ reported 1-year OS and DFS of 44 and 37%. The St Louis group used unrelated donors and a 550 cGy TBI and Cy regimen and reported 72% 1-year survival in first CR AML.^{53, 54}

The only available results with MIST in AML are from the Seattle group.²⁰ A total of 18 patients in CR1 received 200 cGy TBI with or without Flu and transplants from human leukocyte antigen (HLA)-matched sibling donors. The high relapse rate of 39% and 1-year DFS of 42% suggested that a more intensive regimen was needed.

The MDACC group⁴⁰ first used a MIST regimen (FAI) and then a MOST regimen (Flu/Mel) for AML. As expected, the F/M regimen produced more regimen-related toxicity, but less relapse than the FAI regimen (relapse rate was 61% with FAI and 30% with FM). Overall, long-term survival results were comparable between regimens (the 3-year OS was 35% with FM and 30% with FAI).

These studies mentioned above as well as other studies from which AML cases can be extracted are summarized in Table 3.^{55, 56, 57}

Table 3 - Representative reports of RIST for AML/MDS.

Full table

Acute lymphoblastic leukemia (ALL)

After MAST, the GVL effect is less potent in ALL compared to other diseases.⁷ A GVL effect in ALL may be evident only with significant GVHD, as opposed to other diseases such as CML.^{58, 59} The results of RIST for ALL are also not that impressive.^{59, 60} Martino et al⁵⁹ reported a 2-year OS of 31% in 27 cases of advanced ALL, with a relapse rate in patients with GVHD. Another report claimed that RIST is effective treatment only for early stage ALL (CR1) but not for advanced ALL (Table 4).^{60, 61} Currently, it is believed that RIST may not be sufficient as a treatment of ALL.

Table 4 - Representative reports of RIST for ALL.

Full table

Other malignancies

CML is one of the diseases for which RIST may be most effective, because the GVL effect is very potent in CML.^{7, 9} CML patients have a large tumor burden, however, and thus some cytoreduction is required.⁶² The best results in CML have been reported after a MOST regimen using Flu/Bu with ATG.⁶³

For some lymphoid diseases, myeloablation may not be needed. In lymphomas, particularly low-grade-lymphomas, early complete donor chimerism may not be critical, because malignancy progresses slowly, and cytoreduction can be achieved with specific agents such as Rituximab.¹⁷ The MDACC group has published the results of MIST in low-grade lymphomas with excellent long-term disease control.^{17, 18}

Does relapse increase after RIST compared to MAST?

Intuitively, less intensive conditioning regimens would cause less cytoreduction of tumor and may result in higher relapse rates. The relapse rates after MOST, however, are not much different from those observed after MAST (Table 2). In one study, the relapse rate was related to the incidence of GVHD, regardless of the intensity of conditioning, again suggesting the importance of GVL as the primary antitumor effect of allogeneic HSCT.³⁹ The relapse rate of AML after MIST, however, appears very high,^{2, 20} suggesting a need for moderate intensity in conditioning for AML. If the GVL effect acts as consolidation or maintenance therapy after the induction of the conditioning regimen, that induction should be of sufficient intensity to control the disease until the consolidation becomes effective.

Both GVL effect and GVHD are observed after RIST, and an optimal balance between these two phenomena is still elusive. We have recently reported more relapse associated with grade II GVHD than with grade I GVHD.⁵⁸ One explanation for this seeming paradox is that systemic steroid treatment may compromise GVL. Thus, prediction of steroid responsiveness at the time of GVHD onset may help tailor immunosuppression in selected cases in order to preserve GVL. Basic and clinical research in this direction is warranted.

Stable mixed chimerism infers an incomplete eradication of host elements, which would imply an increased risk of relapse. In addition, during stable mixed chimerism, donor-derived T cells must be tolerant of host hematopoietic cells, and this lack of alloreactivity may offer less protection against relapse. Mixed chimerism in natural killer (NK) cells may be associated with graft rejection after RIST,³⁵ but it is not known whether such chimerism affects relapse or GVHD. Early complete T-cell donor chimerism is associated with less relapse, less graft rejection, and better survival in one study.⁶⁴ In fact, most reports indicate fewer relapses in complete donor-type chimeras,^{65, 66, 67, 68, 69, 70, 71, 72, 73} although this finding is not universal.^{74, 75} Representative studies are summarized in Table 5.

Table 5 - Comparison of reports that claimed mixed chimerism predisposed relapse (A) or not (B).

Full table

Some studies have shown that the treatment before transplant may affect the engraftment and relapse rate after RIST.⁷⁶ Additional treatment after remission may suppress the disease, as well as host hematopoiesis and host T cells. If the suppression of host T cells results in earlier and more stable engraftment of donor cells, particularly donor T cells, early donor-type chimerism may be achieved, and relapse rate may decrease.⁶⁴ Although intensive treatment before transplant tends to increase regimen-related toxicity (RRT) after MAST, RRT is lower after RIST and the concern decreases. Pre-transplant therapy before MAST does not improve post transplant survival in first-remission AML,⁷⁷ but the impact of pre-transplant therapy for RIST needs to be assessed.

Conclusions and future directions

To date, most studies of RIST for acute leukemia are small, single-institution studies or retrospective multi-institution studies. Larger, prospective, multicenter studies are needed, particularly in AML where a comparison of RIST and MAST for patients who are eligible for full-intensity HSCT. RIST is not advised as treatment for ALL outside of a clinical trial.

The use of RIST is also being explored for myelofibrosis,^{78, 79, 80} solid tumors,^{5, 81, 82} aplastic anemia,^{83, 84} paroxysmal nocturnal hemoglobinuria (PNH),^{85, 86, 87} thalassemia,⁸⁸ sickle cell anemia,^{88, 89} benign nonhematological conditions such as autoimmune diseases.^{90, 91} Future studies may also attempt to 'target' the kinetics of complete donor T-cell chimerism according to the need for immediate disease control. The indications for RIST may therefore expand beyond those of MAST. The variety of RIST regimens may permit risk stratification according to the severity and the refractoriness of each disease. In addition, RIST using alternative donor sources, such as cord blood^{92, 93} or HLA-mismatched donors,^{94, 95} deserves investigation.

To take advantage of strong GVL potential in unrelated transplantation,^{96, 97, 98} RIST from an unrelated donor may be preferred to a matched sibling donor, if GVHD can be better controlled. Research in this area enhances efforts to separate GVL from GVHD. The creation of an optimal regimen for each disease entity will also likely include disease-specific agents in the future. Such approaches need to be validated in large, prospective, multi-institutional studies with significant length of follow up in order to make firm conclusions regarding the optimal RIST regimens in each disease.

References

References