Abstract
Purpose
Morbidity from acute graft-versus-host disease (GVHD) limits the success of allogeneic hematopoietic stem-cell transplantation (HSCT) to treat malignancy. Interleukin-7 (IL-7), the principal homeostatic cytokine for T cells, is required for acute GVHD in murine models. In contrast to inflammatory cytokines (eg, IL-2, tumor necrosis factor α), IL-7 has not been studied extensively in the clinical transplant setting relative to its relationship with acute GVHD.
Patients and Methods
We evaluated the association of serum IL-7 levels with acute GVHD in 31 patients who were uniformly treated in a prospective clinical trial with reduced-intensity allogeneic HSCT from human leukocyte antigen–identical siblings. GVHD prophylaxis consisted of cyclosporine and methotrexate. Serum IL-7 levels and lymphocyte populations were determined at enrollment, the day of transplantation before the allograft infusion, and at specified intervals through 12 months post-transplantation.
Results
As expected, IL-7 levels were inversely correlated with T-cell populations (P < .00001). Acute GVHD was significantly associated with higher IL-7 levels at day +7 (P = .01) and day +14 (P = .00003) post-transplantation as well as with the allograft CD34+ cell dose (P = .01). IL-7 levels at day +14 also correlated with the severity of acute GVHD (P < .0001). In logistic regression models, these factors were highly sensitive (up to 86%) and specific (100%) for classifying whether patients developed acute GVHD.
Conclusion
These data support preclinical observations that IL-7 plays a critical role in inducing acute GVHD and provide a rational basis for novel approaches to prevent and treat acute GVHD through modulation of the IL-7 pathway.
INTRODUCTION
The morbidity that may result from acute graft-versus-host disease (GVHD) poses a major barrier to the successful application of allogeneic hematopoietic stem-cell transplantation (HSCT) to treat malignancy.1 On infusion, the allogeneic effector T-cell populations that initiate acute GVHD are exposed to elevated levels of various cytokines, of which many are inflammatory mediators, (eg, tumor necrosis factor α) whose contribution to acute GVHD has been the focus of extensive prior study.2
Allogeneic effector T cells are also directly influenced by homeostatic cytokines, particularly interleukin-7 (IL-7), that promote immune reconstitution after allogeneic HSCT. As the principal T-cell homeostatic cytokine, IL-7 is critical for T-cell development and survival.3 IL-7 is produced constitutively by stromal cells and consumed by the available pool of resting T cells, all of which express the IL-7 receptor (IL-7R) at high levels except for CD4+CD25+ regulatory T cells.4 Systemic IL-7 levels increase during periods of lymphopenia to maintain naïve T-cell homeostasis and support the thymic-independent peripheral expansion and maintenance of mature T cells.5-7 The conditioning regimen administered before allogeneic HSCT causes severe host lymphopenia, thus perturbing T-cell homeostasis and creating a milieu in which infused allogeneic lymphocytes are exposed to elevated levels of endogenous IL-7.5 These conditions promote the expansion and persistence of allogeneic T cells that have the potential to recognize host antigens, become activated, and initiate GVHD.
Murine models from our laboratory and others implicate IL-7 in the development of acute GVHD.8-11 However, a relationship between IL-7 and acute GVHD has not been identified in the limited clinical data that are published to date.12 We evaluated the association of IL-7 levels with the development of acute GVHD in a prospective clinical trial, finding that serum IL-7 levels in the early post-transplantation period were sensitive and specific for predicting the development of acute GVHD.
PATIENTS AND METHODS
Patients
Patients were eligible for this study if they had a high-risk hematologic malignancy and a consenting human leukocyte antigen–matched sibling donor. The study (CC# 03-C-0077) was approved by the National Cancer Institute Institutional Review Board and conducted in accordance with the Declaration of Helsinki.
Treatment
Before the transplant conditioning regimen, all patients received an identical, conventional-dose induction regimen with the goal of achieving a relatively homogeneous patient population with respect to host T-cell lymphopenia. Up to three cycles of this regimen, consisting of etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and fludarabine (EPOCH-F), were administered until the peripheral-blood CD4+ T-cell count was less than 100 cells/μL.13 Patients with CD20+ malignancies also received rituximab with EPOCH-F. All patients then received reduced-intensity conditioning consisting of fludarabine (30 mg/m2/d) and cyclophosphamide (1,200 mg/m2/d) for 4 days. Donor peripheral-blood hematopoietic progenitor cells were mobilized with filgrastim, collected by apheresis, and cryopreserved until the day of transplantation. GVHD prophylaxis consisted of cyclosporine and methotrexate. Cyclosporine was administered twice daily, starting the day before transplantation, maintained at serum levels between 200 and 250 ng/mL through 3 months post-transplantation, then tapered to discontinuation by 6 months post-transplantation in the absence of GVHD. Methotrexate (5 mg/m2) was administered on days +1, +3, +6, and +11 after transplantation. Acute GVHD was confirmed by biopsy of at least one involved site and graded according to the 1994 Consensus Criteria.14
Serum IL-7 Enzyme-Linked Immunosorbent Assay
Serum samples for IL-7 determination were obtained at enrollment before EPOCH-F induction chemotherapy; on the day of transplantation before infusion of the allograft; and at day +7, day +14, 1 month, 2 months, 3 months, 6 months, 9 months, and 12 months post-transplantation. Sera were prepared by centrifugation of blood and stored in 0.5-mL aliquots at −80°C. Samples were analyzed using a high-sensitivity colorimetric enzyme-linked immunosorbent assay (ELISA; R&D Systems, Minneapolis, MN). Along with patient serum samples, each ELISA plate was loaded with serial dilutions of a standard recombinant IL-7 solution that was provided with the ELISA assay, according to the manufacturer's instructions. The results of these samples were used to prepare standard curves for each plate. Patient samples whose IL-7 levels exceeded the range defined by the standard samples were diluted and rerun on another ELISA plate. Plates were analyzed with a microplate reader and SOFTmax Pro 3.0 software (Molecular Devices Corporation, Sunnyvale, CA).
Flow Cytometry
Blood was collected at the same time points as serum IL-7 measurements. Flow cytometric enumeration of lymphocyte populations was performed using whole-blood lysis by a Clinical Laboratory Improvement Amendments–certified laboratory (Science Applications International Corporation, Frederick, MD). The absolute numbers of CD3+, CD4+CD3+ and CD8+CD3+ T cells, and CD3–CD56+ NK cells were calculated using the percentages of these cells in the lymphocyte gate and the absolute lymphocyte count from a CBC (Beckman Coulter Inc, Fullerton, CA).
Donor-Recipient Hematopoietic Chimerism
Chimerism analysis was performed at day +14, 1 month, and 3 months post-transplantation on total peripheral-blood mononuclear cells, using the variable number tandem repeats–polymerase chain reaction method in a Clinical Laboratory Improvement Amendments–certified laboratory (Blood Center of Southeastern Wisconsin, Milwaukee, WI). Chimerism was also determined on blood samples enriched for myeloid (CD15+ or CD33+) and T-lymphoid (CD3+) subsets.
Statistical Analysis
Overall survival (OS), progression-free survival (PFS), and the proportions of patients developing acute and chronic GVHD were estimated using the Kaplan-Meier method. Serum IL-7 levels and peripheral-blood lymphocyte counts were compared with dichotomous clinical parameters including acute GVHD; chronic GVHD; engraftment syndrome requiring corticosteroid administration, defined as noninfectious fever, erythematous rash, and noncardiogenic pulmonary edema resulting in diffuse pulmonary infiltrates and hypoxemia, occurring within 96 hours of neutrophil recovery;15 full donor chimerism, defined as at least 95% donor origin for all cell populations measured, at day +14, 1 month, and 3 months post-transplantation; and cytomegalovirus infection. Other relevant clinical and immunologic parameters were analyzed for potential associations with acute GVHD, as shown in Table 4. Comparisons employed the exact Wilcoxon rank sum tests for continuous parameters and Fisher's exact test for categoric parameters. Logistic regression analysis was used to determine which parameters identified as potentially significant in the univariate analysis were jointly associated with acute GVHD. Correlation analyses used Spearman (nonparametric) analysis, with r less than 0.70 indicating strong correlations and r between −0.5 and −0.7 indicating moderately strong correlations. All P values are two tailed and have not been formally adjusted for multiple comparisons; however, in this exploratory study, P values less than .005 would be readily interpretable as being statistically significant whereas those such that P between .005 and .05 would reflect strong trends.
Table 4.
Results From Univariate Analysis of Factors for Acute GVHD
| Variable | Acute GVHD
|
P | |
|---|---|---|---|
| Yes | No | ||
| Patient age, years | .91 | ||
| Median | 57 | 53 | |
| Range | 32-71 | 31-63 | |
| Donor age, years | .18 | ||
| Median | 53 | 48 | |
| Range | 38-74 | 29-67 | |
| Disease status at transplant* | |||
| CR or PR | 6 | 7 | .71 |
| SD or PD | 6 | 11 | |
| Graft CD3 dose, ×106 cells/kg | .27 | ||
| Median | 284 | 349 | |
| Range | 176-610 | 103-533 | |
| Graft CD34 dose, ×106 cells/kg | .01 | ||
| Median | 9.15 | 5.72 | |
| Range | 3.95-17.40 | 4.0-10.71 | |
| IL-7 at day +7, pg/mL | .01 | ||
| Median | 25.7 | 11.8 | |
| Range | 7.3-73.0 | 3.0-27.8 | |
| IL-7 at day +14, pg/mL | .00003 | ||
| Median | 20.9 | 8.2 | |
| Range | 9.9-74.7 | 1.3-27.2 | |
| Donor sex | .99 | ||
| Male | 5 | 7 | |
| Female | 9 | 10 | |
| Sex mismatch | .99 | ||
| Yes | 8 | 10 | |
| No | 6 | 7 | |
| CMV exposure | .46 | ||
| Yes | 11 | 11 | |
| No | 3 | 6 | |
| Prior rituximab | .72 | ||
| Yes | 9 | 9 | |
| No | 5 | 8 | |
| Donor chimerism at day +14 | |||
| Myeloid > 95% | .67 | ||
| Yes | 10 | 12 | |
| No | 2 | 4 | |
| Lymphoid > 95% | .99 | ||
| Yes | 10 | 13 | |
| No | 4 | 4 | |
Abbreviations: GVHD, graft-versus-host disease; IL, interleukin; CR, complete remission; PR, partial remission; SD, stable disease; PD, progressive disease; CMV, cytomegalovirus.
Disease status was not assessable for one patient.
RESULTS
Between January 2003 and August 2004, 31 patients were enrolled (Table 1). The median recipient age was 57 years. Donor-recipient sex mismatch was present for 18 recipients. In nine cases, both the donor and the recipient were seronegative for cytomegalovirus; 22 recipients were either seropositive themselves (n = 18) and/or had a seropositive donor (n = 17). Disease status at the time of transplantation was chemotherapy-sensitive (complete or partial remission) in 12 patients, chemotherapy-resistant (stable or progressive disease) in 18 patients, and not assessable in one patient.
Table 1.
Characteristics of Patients Undergoing Reduced-Intensity Allogeneic Hematopoietic Stem-Cell Transplantation
| Characteristic | Value |
|---|---|
| Age, years | |
| Recipient | |
| Median | 57 |
| Range | 31-71 |
| Donor | |
| Median | 50 |
| Range | 29-74 |
| Diagnosis, No. | |
| Non-Hodgkin's lymphoma | 17 |
| CLL/SLL | 4 |
| PLL | |
| Hodgkin's lymphoma | 5* |
| MDS/AML | 3 |
| MPD | 2 |
| Disease status at transplantation, No. | |
| Complete remission | 6 |
| Partial remission | 6 |
| Stable disease | 14 |
| Progressive disease | 4 |
| Not assessable | 1 |
| CD20+ malignancy (rituximab exposure), No. | |
| Yes | 18 |
| No | 13 |
| Sex (donor/recipient), No. | |
| Male/male | 9 |
| Male/female | 3 |
| Female/male | 15 |
| Female/Female | 4 |
| Donor pregnancy, No. | |
| Yes | 15 |
| No | 1 |
| Unknown | 3 |
| CMV status (donor/recipient), No. | |
| Negative/negative | 9 |
| Negative/positive | 3 |
| Positive/negative | 4 |
| Positive/positive | 15 |
| Allograft composition | |
| CD34+ cells/kg recipient weight (×106) | |
| Median | 6.8 |
| Range | 4.0-17.4 |
| CD3+ cells/kg recipient weight (×108) | |
| Median | 3.2 |
| Range | 1.0-5.3 |
Abbreviations: CLL, chronic lymphocytic leukemia; SLL, small lymphocytic lymphoma; PLL, prolymphocytic leukemia; NHL, non-Hodgkin lymphoma; MDS, myelodysplastic syndrome; AML, acute myelogenous leukemia; MPD, myeloproliferative disorder; CMV, cytomegalovirus.
One prior follicular non-Hodgkin's lymphoma.
Transplant Outcomes
Infused allografts contained a median 6.8 × 106 CD34+ cells/kg and 3.2 × 108 CD3+ cells/kg. Median donor T-cell chimerism at day +14 post-transplantation was 98% (range, 60% to 100%), with full donor T-cell chimerism in 23 patients (74%); median donor mononuclear cell chimerism at 1 month and 3 months was 100% (range, 55% to 100%) and 100% (range, 10% to 100%), respectively, with full donor mononuclear cell chimerism in 24 patients (77%) and 22 patients (71%), respectively. There were no primary graft failures.
Grades 1 to 3 acute GVHD (Table 2) occurred in 14 (45%) of 31 patients, with grades 2 to 3 occurring in 13 patients (42%). One case of grade 1 acute GVHD occurred with isolated stage 2 skin involvement. No grade 4 acute GVHD was observed. Grade 2 acute GVHD was predominantly cutaneous; five of six patients with grade 3 acute GVHD had GI involvement. Median time to onset of acute GVHD was 33 days (range, 19 to 49 days) post-transplantation. Chronic GVHD occurred in 21 of 29 assessable patients, of whom five have died. At 2 years post-transplantation, chronic GVHD had occurred in 72% of patients. Two patients who died as a result of progressive lymphoma at days +125 and +166 post-transplantation, respectively, were not assessable for chronic GVHD. With a median potential follow-up of 26.7 months, actuarial PFS and OS probabilities at 2 years were 49.5% (median, 20.6 months) and 64% (median not reached), respectively.
Table 2.
Relationship of Acute and Chronic GVHD to Serum IL-7 Levels
| GVHD | No. of Patients | % | Organ Involvement (No. of patients)
|
Median IL-7 Level (pg/mL)
|
|||
|---|---|---|---|---|---|---|---|
| Skin | Gut | Liver | Day +7 | Day +14 | |||
| Acute | |||||||
| None | 17 | 55 | 11.8 | 8.2 | |||
| Grade I | 1 | 3 | 1 | 0 | 0 | 23.9 | 14.0 |
| Grade II | 7 | 23 | 6 | 3 | 0 | 32.8 | 23.0 |
| Grade III | 6 | 19 | 3 | 5 | 2 | 25.7 | 27.8 |
| Grade IV | 0 | 0 | |||||
| Chronic | |||||||
| Yes | 21 | 68 | 12.82 | 12.42 | |||
| No | 8 | 26 | 24.85 | 12.27 | |||
| Not assessable | 2 | 6 | — | — | |||
Abbreviations: GVHD, graft-versus-host disease; IL, interleukin.
Serum IL-7 Levels and Peripheral-Blood Lymphocyte Counts
Median IL-7 levels rose from 12.1 pg/mL (range, 0 to 46.9 pg/mL) at enrollment to a peak of 37 pg/mL (range, 13.3 to 79.2 pg/mL) on day 0 after the transplant conditioning regimen, but before the infusion of the allograft (Fig 1A; Table 3). Median IL-7 levels subsequently declined to 12.0 pg/mL (range, 1.3 to 74.7 pg/mL) by day +14 after transplantation and remained relatively stable thereafter. Inverse correlations were present between serum IL-7 levels and circulating CD3+ T-cell counts at baseline (r = −0.52; P = .005), and at day +7 post-transplantation (r = −0.82; P < .0001). Absolute lymphocyte, CD4, and CD8 counts were also inversely correlated with serum IL-7 levels at day +7 (Table 3).
Fig 1.
(A) Median serum interleukin-7 (IL-7) levels and blood lymphocyte counts from baseline through 12 months post-transplantation. (B) Median serum IL-7 levels according to subsequent development of acute graft-versus-host disease (aGVHD). Error bars denote interquartile ranges. ALC, absolute lymphocyte count.
Table 3.
Correlation of IL-7 Levels With Lymphocyte Counts Over Time
| Level | Baseline | Day 0 | Day +7 | Day +14 | 1 Month | 2 Months | 3 Months | 6 Months | 9 Months | 12 Months |
|---|---|---|---|---|---|---|---|---|---|---|
| IL-7, pg/mL | ||||||||||
| Median | 12.1 | 37.0 | 13.5 | 12.0 | 9.4 | 6.5 | 7.6 | 9.8 | 9.0 | 15.0 |
| Range | 0-46.9 | 13.3-79.2 | 3.0-73.0 | 1.3-74.7 | 1.3-34.9 | 1.5-70.0 | 0.2-68.0 | 0-43.5 | 0-71.2 | 3.2-33.7 |
| ALC, cells/μL | ||||||||||
| Median | 871 | 3 | 76 | 493 | 1,138 | 1,028 | 1,014 | 1,298 | 1,600 | 1,220 |
| Range | 95-43,817 | 0-322 | 6-525 | 9-2,904 | 149-3,172 | 145-5,564 | 110-3,606 | 147-4,140 | 207-3,909 | 229-5,494 |
| R | −0.49 | −0.28 | −0.83 | −0.35 | −0.25 | −0.22 | −0.08 | −0.24 | −0.40 | −0.22 |
| P | .008 | .14 | < .0001 | .06 | .18 | .27 | .66 | .30 | .06 | .36 |
| CD3, cells/μL | ||||||||||
| Median | 629 | 3 | 60 | 307 | 773 | 777 | 745 | 1003 | 1339 | 985 |
| Range | 5-2,858 | 0-69 | 5-464 | 6-1,428 | 128-2,927 | 112-5,369 | 97-3,516 | 95-3,908 | 172-3,729 | 200-5,257 |
| R | −0.52 | −0.28 | −0.82 | −0.36 | −0.39 | −0.22 | −0.14 | −0.25 | −0.38 | −0.33 |
| P | .005 | .14 | < .0001 | .05 | .04 | .28 | .47 | .29 | .07 | .15 |
| CD4, cells/μL | ||||||||||
| Median | 294 | 3 | 45 | 206 | 405 | 310 | 352 | 439 | 494 | 440 |
| Range | 5-1,652 | 0-65 | 5-385 | 5-810 | 75-1,145 | 78-901 | 40-1,493 | 35-1,457 | 70-1,176 | 33-1,032 |
| R | −0.58 | −0.22 | −0.86 | −0.33 | −0.36 | −0.19 | −0.08 | −0.34 | −0.20 | −0.25 |
| P | .001 | .24 | < .0001 | .07 | .05 | .36 | .66 | .14 | .34 | .29 |
| CD8, cells/μL | ||||||||||
| Median | 269 | 0 | 16 | 95 | 326 | 364 | 356 | 557 | 660 | 397 |
| Range | 1-1,763 | 0-22 | 0-88 | 1-963 | 27-2,123 | 31-4,484 | 57-2,990 | 57-2,447 | 98-2,932 | 165-4,258 |
| R | −0.54 | −0.35 | −0.65 | −0.48 | −0.32 | −0.22 | −0.16 | −0.22 | −0.46 | −0.22 |
| P | .003 | .06 | .0002 | .006 | .09 | .27 | .39 | .36 | .02 | .34 |
NOTE. Correlation coefficients and P values shown are for comparison with interleukin-7 (IL-7) levels at specified time points.
Abbreviation: ALC, absolute lymphocyte count.
IL-7 and Acute GVHD
The IL-7 levels at day +7 (P = .01) and day +14 (P = .00003) post-transplantation were associated with the subsequent development of acute GVHD (Table 4; Fig 1B). Among patients who developed acute GVHD, median IL-7 levels were 25.7 and 20.9 pg/mL at days +7 and +14, respectively. Patients who did not develop acute GVHD had median IL-7 levels of 11.8 and 8.22 pg/mL at days +7 and +14, respectively. Higher IL-7 levels at day +14 were strongly associated with more severe grades of acute GVHD (P < .0001; Table 2). Among other clinical parameters (Table 4), only the allograft CD34+ cell dose was also strongly (P = .01) associated with acute GVHD. Acute GVHD was not associated with prior rituximab administration (P = .72). The achievement of complete donor myeloid or lymphoid chimerism by day +14 post-transplantation was not associated with the development of acute GVHD. Serum IL-7 levels were not significantly associated with full donor chimerism through 3 months post-transplantation (data not shown) or with the development of chronic GVHD. IL-7 levels at day +7 or day +14 post-transplantation demonstrated no significant association with PFS or OS, although there was a modest trend toward better OS among patients with serum IL-7 levels less than 24.4 pg/mL at day +7 (P = .07). Disease status at transplantation was not associated with either IL-7 levels (data not shown) or the development of acute GVHD (P = .71).
Logistic regression analysis confirmed that the IL-7 level at day +14 was the strongest single parameter associated with acute GVHD. Using a cutoff value of 13 pg/mL, the IL-7 level at day +14 predicted the subsequent development of acute GVHD with a sensitivity of 85.7% and specificity of 88.2%. The positive and negative predictive values for this single-variable model were 85.7% and 88.2%, respectively. Alternate predictive models that included the CD34+ cell dose and the IL-7 level at day +7 were also evaluated. The close correlation between IL-7 levels at day +7 and day +14 (r = 0.69; P < .0001) precluded both parameters from appearing jointly in a model. IL-7 at day +7 alone was less accurate than IL-7 at day +14 alone (data not shown). When the CD34+ cell dose was included with the IL-7 level at either day +7 or day +14, the resulting models predicted acute GVHD with only two errors (Tables 5 and 6), with sensitivity of 84.6% or 85.7%, respectively, and with 100% specificity for either model.
Table 5.
Logistic Regression Analysis for Development of Acute Graft-Versus-Host Disease
| Model | Parameter Estimate | P | Odds Ratio | 95% CI | No. of Patients |
|---|---|---|---|---|---|
| #1 | |||||
| Intercept | −3.26 | .006 | 31 | ||
| IL-7 day +14 | 0.21 | .01 | 1.24 | 1.04 to 1.46 | |
| #2 | |||||
| Intercept | −11.38 | .01 | 29* | ||
| IL-7 day +7 | 0.25 | .02 | 1.28 | 1.04 to 1.59 | |
| CD34+ cell dose | 0.88 | .03 | 2.41 | 1.10 to 5.28 | |
| #3 | |||||
| Intercept | −6.90 | .004 | 31 | ||
| IL-7 day +14 | 0.18 | .009 | 1.20 | 1.05 to 1.38 | |
| CD34+ cell dose | 0.50 | .04 | 1.66 | 1.01 to 2.70 |
NOTE. Intercept is a logistic regression term indicating a constant value in deriving the equation to determine the probability of the outcome of interest.
Abbreviation: IL, interleukin.
IL-7 data missing at day +7 for two patients.
Table 6.
Classification Results by Model
| Predicted | Observed aGVHD
|
|
|---|---|---|
| Yes | No | |
| Model 1 | ||
| Yes | 12 | 2 |
| No | 2 | 15 |
| Model 2 | ||
| Yes | 11 | 0 |
| No | 2 | 16 |
| Model 3 | ||
| Yes | 12 | 0 |
| No | 2 | 17 |
Abbreviation: aGVHD, acute graft-versus-host disease.
Because engraftment syndrome (ES) may result from alloreactivity after allogeneic HSCT,15 the relationship between IL-7 at day +14 and acute GVHD was re-examined with all cases of ES reclassified as acute GVHD, finding a similarly strong association between IL-7 at day +14 and acute GVHD in univariate analysis (P = .00001).
DISCUSSION
Multiple cytokines contribute to the initiation, severity, and persistence of acute GVHD.2 Prior studies have mainly focused on inflammatory cytokines; however, murine models indicate that IL-7 is necessary for the initiation of GVHD.10,11 We prospectively studied patients who underwent allogeneic HSCT in a uniform manner to generate preliminary clinical data relative to the relationship between serum IL-7 levels and acute GVHD. These data demonstrated that higher serum IL-7 levels in the early post-transplantation period were strongly associated with both the subsequent development and the severity of acute GVHD. The magnitude of the effect that we observed, even with a relatively small number of patients, supports the validity of these data and is consistent with the central role of IL-7 in T-cell homeostasis.
Our study confirms that serum IL-7 levels correlate inversely with the numbers of circulating T cells. The patients in this study received induction chemotherapy before reduced-intensity conditioning to produce deliberate host lymphodepletion. This design helped to clarify the relationship of IL-7 to acute GVHD by minimizing the influence of other possible confounding factors. First, host T cells were uniformly depleted to levels that are comparable with those achieved with myeloablative conditioning regimens, thus potentiating the homeostatic induction of IL-7 in the peritransplantation period. Second, unlike some nonmyeloablative regimens, this reduced-intensity platform brought about complete donor chimerism in most patients in the early post-transplantation period. This is relevant because states of mixed donor chimerism are associated with a decreased incidence of acute GVHD,16 although no significant relationship was detected between chimerism and GVHD in our analysis.
The CD34+ cell dose was the only other parameter that demonstrated a significant association with acute GVHD in univariable and multivariable analyses. Stem-cell dose has been identified as a risk factor for acute GVHD in some studies, although the physiologic basis for this relationship is undefined.17,18 In our analysis, CD34+ cell dose had a modest impact on the risk of acute GVHD, compared with the effect of serum IL-7 levels. However, there is evidence that IL-7 can influence CD34+ T-cell precursors,19,20 and a hypothesis that might explain our findings is that higher IL-7 levels potentiated T-cell proliferation and maturation among patients with higher CD34+ cell doses.
Several patients on our trial developed ES. The biology of ES is poorly defined, and its distinction from acute GVHD is somewhat controversial. When we re-examined the effect of IL-7 with all cases of ES reclassified as acute GVHD, the association between IL-7 and acute GVHD was strengthened. This interesting observation supports a potential biologic relationship between ES and acute GVHD after allogeneic HSCT.
Further studies are needed to define the mechanisms underlying our findings. One possible explanation is that elevated serum IL-7 levels are a biologic marker for patients who are in the preclinical stages of acute GVHD. T cells that bind antigen and undergo activation downregulate IL-7R, leading to reduced IL-7 consumption.21 Thus, more pronounced elevations in serum IL-7 levels might be expected in patients with incipient acute GVHD in whom alloreactive donor T cells have become activated. Such a phenomenon is consistent with the findings of a murine model of allogeneic HSCT, in which alloreactive T cells lacked IL-7R expression.22 However, less than 1% of the T-cell clones within a graft are potentially alloreactive,23 so this mechanism may not adequately explain the relationship between IL-7 and acute GVHD.
Another possible explanation is that higher IL-7 levels directly lead to increased incidence and severity of acute GVHD. IL-7 lowers the threshold to develop acute GVHD, given that signaling through both the T-cell receptor (TCR) and IL-7R simultaneously is synergistic for activation and expansion of T cells. In murine models, costimulation with recombinant human IL-7 (rhIL-7) and anti-TCR antibody significantly increased secretion of the inflammatory cytokine interferon-γ by spleen-derived T cells from experimental bone marrow transplant recipients, and administration of rhIL-7 after allogeneic HSCT increased the severity of GVHD in a donor T-cell-dependent manner, possibly by augmenting both effector cell number and reactivity towards subdominant antigen.8,24 IL-7 also prevents T cells from undergoing activation-induced cell death, and so may provide a critical survival signal for effector T cells that initiate and sustain acute GVHD. Alternatively, IL-7 enhances the development and function of dendritic cells,3 which would favor greater presentation of alloantigens after allogeneic HSCT.
Our observations have significant clinical implications. Early prediction of acute GVHD based on serum IL-7 levels as a validated biologic marker could allow an individualized, risk-adapted strategy for prophylaxis of acute GVHD. Other assays related to inflammatory cytokines have been evaluated to identify patients at risk for developing acute GVHD,25-29 but these have not achieved widespread clinical use, possibly because of limitations concerning their predictive power, complexity, availability, or cost. In contrast, the approach used in this study employs technologies that are comparatively simple, inexpensive, and highly reproducible. If serum IL-7 levels directly influence the development of acute GVHD, then the IL-7 signaling pathway may provide novel targets for treatment and even prevention of acute GVHD.30 Our findings also suggest potential pitfalls in clinical trials of rhIL-7 as an immune adjuvant after allogeneic HSCT, and it needs to be determined whether the association between IL-7 and acute GVHD in our study can be generalized to other conditions that modulate the risk of GVHD (eg, allograft T-cell depletion and conditioning regimen intensity) because murine studies indicate that the safety of rhIL-7 administration in this setting may depend on such factors.8,9,22,31 We are initiating a multi-institutional study of the association of IL-7 and acute GVHD in varying allogeneic HSCT settings that vary by conditioning regimen and stem-cell source.
In conclusion, the observations in this clinical study substantiate the findings of preclinical murine models and are consistent with prior knowledge regarding the biologic role of IL-7 in T-cell homeostasis and the pathogenesis of acute GVHD. Serum IL-7 levels quantify the risk of acute GVHD and may enable a highly individualized clinical approach to managing acute GVHD after allogeneic HSCT. The IL-7 pathway represents a promising target for the development of novel agents to prevent and treat acute GVHD.
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
The author(s) indicated no potential conflicts of interest.
AUTHOR CONTRIBUTIONS
Conception and design: Robert M. Dean, Terry Fry, Crystal Mackall, Seth Steinberg, Ronald Gress, Michael R. Bishop
Financial support: Michael R. Bishop
Administrative support: Jeanne Odom
Provision of study materials or patients: Terry Fry, Fran Hakim, Daniel Fowler, Jason Foley, Michael R. Bishop
Collection and assembly of data: Robert M. Dean, Terry Fry, Seth Steinberg, Fran Hakim, Jeanne Odom, Jason Foley, Michael R. Bishop
Data analysis and interpretation: Robert M. Dean, Terry Fry, Crystal Mackall, Seth Steinberg, Fran Hakim, Jeanne Odom, Jason Foley, Ronald Gress, Michael R. Bishop
Manuscript writing: Robert M. Dean, Seth Steinberg, Daniel Fowler, Ronald Gress, Michael R. Bishop
Final approval of manuscript: Robert M. Dean, Terry Fry, Crystal Mackall, Seth Steinberg, Fran Hakim, Daniel Fowler, Jeanne Odom, Jason Foley, Ronald Gress, Michael R. Bishop
Acknowledgments
We thank Lisa Rybicki for her technical assistance in preparing the figures.
published online ahead of print at www.jco.org on November 10, 2008.
Supported by the Intramural Research Program of the Center for Cancer Research, National Cancer Institute, National Institutes of Health.
Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
Clinical trial information can be found for the following: NCT00055744.
REFERENCES
- 1.Rowlings PA, Przepiorka D, Klein JP, et al: IBMTR Severity Index for grading acute graft-versus-host disease: Retrospective comparison with Glucksberg grade. Br J Haematol 97:855-864, 1997 [DOI] [PubMed] [Google Scholar]
- 2.Ferrara JL, Reddy P: Pathophysiology of graft-versus-host disease. Semin Hematol 43:3-10, 2006 [DOI] [PubMed] [Google Scholar]
- 3.Fry TJ, Mackall CL: Interleukin-7: From bench to clinic. Blood 99:3892-3904, 2002 [DOI] [PubMed] [Google Scholar]
- 4.Seddiki N, Santner-Nanan B, Martinson J, et al: Expression of interleukin (IL)-2 and IL-7 receptors discriminates between human regulatory and activated T cells. J Exp Med 203:1693-1700, 2006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bolotin E, Annett G, Parkman R, et al: Serum levels of IL-7 in bone marrow transplant recipients: Relationship to clinical characteristics and lymphocyte count. Bone Marrow Transplant 23:783-788, 1999 [DOI] [PubMed] [Google Scholar]
- 6.Fry TJ, Connick E, Falloon J, et al: A potential role for interleukin-7 in T-cell homeostasis. Blood 97:2983-2990, 2001 [DOI] [PubMed] [Google Scholar]
- 7.Napolitano LA, Grant RM, Deeks SG, et al: Increased production of IL-7 accompanies HIV-1-mediated T-cell depletion: Implications for T-cell homeostasis. Nat Med 7:73-79, 2001 [DOI] [PubMed] [Google Scholar]
- 8.Sinha ML, Fry TJ, Fowler DH, et al: Interleukin 7 worsens graft-versus-host disease. Blood 100:2642-2649, 2002 [DOI] [PubMed] [Google Scholar]
- 9.Gendelman M, Hecht T, Logan B, et al: Host conditioning is a primary determinant in modulating the effect of IL-7 on murine graft-versus-host disease. J Immunol 172:3328-3336, 2004 [DOI] [PubMed] [Google Scholar]
- 10.Chung B, Dudl EP, Min D, et al: Prevention of graft-versus-host disease by anti IL-7Ralpha antibody. Blood 110:2803-2810, 2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Chung B, Dudl E, Toyama A, et al: Importance of interleukin-7 in the development of experimental graft-versus-host disease. Biol Blood Marrow Transplant 14:16-27, 2008 [DOI] [PubMed] [Google Scholar]
- 12.Abu-Ghosh A, Goldman S, Slone V, et al: Immunological reconstitution and correlation of circulating serum inflammatory mediators/cytokines with the incidence of acute graft-versus-host disease during the first 100 days following unrelated umbilical cord blood transplantation. Bone Marrow Transplant 24:535-544, 1999 [DOI] [PubMed] [Google Scholar]
- 13.Bishop MR, Hou JW, Wilson WH, et al: Establishment of early donor engraftment after reduced-intensity allogeneic hematopoietic stem cell transplantation to potentiate the graft-versus-lymphoma effect against refractory lymphomas. Biol Blood Marrow Transplant 9:162-169, 2003 [DOI] [PubMed] [Google Scholar]
- 14.Przepiorka D, Weisdorf D, Martin P, et al: 1994 consensus conference on acute GVHD grading. Bone Marrow Transplant 15:825-828, 1995 [PubMed] [Google Scholar]
- 15.Spitzer TR: Engraftment syndrome following hematopoietic stem cell transplantation. Bone Marrow Transplant 27:893-898, 2001 [DOI] [PubMed] [Google Scholar]
- 16.Childs R, Clave E, Contentin N, et al: Engraftment kinetics after nonmyeloablative allogeneic peripheral blood stem cell transplantation: Full donor T-cell chimerism precedes alloimmune responses. Blood 94:3234-3241, 1999 [PubMed] [Google Scholar]
- 17.Przepiorka D, Smith TL, Folloder J, et al: Risk factors for acute graft-versus-host disease after allogeneic blood stem cell transplantation. Blood 94:1465-1470, 1999 [PubMed] [Google Scholar]
- 18.Remberger M, Mattsson J, Hassan Z, et al: Risk factors for acute graft-versus-host disease grades II-IV after reduced intensity conditioning allogeneic stem cell transplantation with unrelated donors: A single centre study. Bone Marrow Transplant 41:399-405, 2008 [DOI] [PubMed] [Google Scholar]
- 19.Schmitt C, Ktorza S, Sarun S, et al: CD34-expressing human thymocyte precursors proliferate in response to interleukin-7 but have lost myeloid differentiation potential. Blood 82:3675-3685, 1993 [PubMed] [Google Scholar]
- 20.Gutiérrez-Frías C, Sacedon R, Hernandez-Lopez C, et al: Sonic hedgehog regulates early human thymocyte differentiation by counteracting the IL-7-induced development of CD34+ precursor cells. J Immunol 173:5046-5053, 2004 [DOI] [PubMed] [Google Scholar]
- 21.Park JH, Yu Q, Erman B, et al: Suppression of IL7Ralpha transcription by IL-7 and other prosurvival cytokines: A novel mechanism for maximizing IL-7-dependent T cell survival. Immunity 21:289-302, 2004 [DOI] [PubMed] [Google Scholar]
- 22.Alpdogan O, Schmaltz C, Muriglan SJ, et al: Administration of interleukin-7 after allogeneic bone marrow transplantation improves immune reconstitution without aggravating graft-versus-host disease. Blood 98:2256-2265, 2001 [DOI] [PubMed] [Google Scholar]
- 23.Smith JB: Frequency in human peripheral blood of T cells which respond to self, modified self and alloantigens. Immunology 50:181-187, 1983 [PMC free article] [PubMed] [Google Scholar]
- 24.Levy RB, Jones M, Hamilton BL, et al: IL-7 drives donor T cell proliferation and can costimulate cytokine secretion after MHC-matched allogeneic bone marrow transplantation. J Immunol 154:106-115, 1995 [PubMed] [Google Scholar]
- 25.Vogelsang GB, Hess AD, Berkman AW, et al: An in vitro predictive test for graft versus host disease in patients with genotypic HLA-identical bone marrow transplants. N Engl J Med 313:645-650, 1985 [DOI] [PubMed] [Google Scholar]
- 26.Holler E, Kolb HJ, Moller A, et al: Increased serum levels of tumor necrosis factor alpha precede major complications of bone marrow transplantation. Blood 75:1011-1016, 1990 [PubMed] [Google Scholar]
- 27.Theobald M, Nierle T, Bunjes D, et al: Host-specific interleukin-2-secreting donor T-cell precursors as predictors of acute graft-versus-host disease in bone marrow transplantation between HLA-identical siblings. N Engl J Med 327:1613-1617, 1992 [DOI] [PubMed] [Google Scholar]
- 28.Middleton PG, Taylor PR, Jackson G, et al: Cytokine gene polymorphisms associating with severe acute graft-versus-host disease in HLA-identical sibling transplants. Blood 92:3943-3948, 1998 [PubMed] [Google Scholar]
- 29.Lin MT, Storer B, Martin PJ, et al: Relation of an interleukin-10 promoter polymorphism to graft-versus-host disease and survival after hematopoietic-cell transplantation. N Engl J Med 349:2201-2210, 2003 [DOI] [PubMed] [Google Scholar]
- 30.Hofmeister R, Khaled AR, Benbernou N, et al: Interleukin-7: physiological roles and mechanisms of action. Cytokine Growth Factor Rev 10:41-60, 1999 [DOI] [PubMed] [Google Scholar]
- 31.Mackall CL, Fry TJ, Bare C, et al: IL-7 increases both thymic-dependent and thymic-independent T-cell regeneration after bone marrow transplantation. Blood 97:1491-1497, 2001 [DOI] [PubMed] [Google Scholar]