Long term follow-up and clinical outcomes of unrelated hematopoetic stem cell transplantation

Summary: Haematopoietic stem cell transplantation (HSCT) provides potential curative treatment for a wide range of otherwise life threatening haematological diseases. Ideally, HSCT is performed using stem cells collected from an HLA-matched sibling but it is not always possible as only approximately 30% of patients will have such a donor. The expansion in alternative donor stem cell sources coupled with the advent of reduced-intensity conditioning (RIC) regimens have contributed to the rise in the number of HSCT in general and in unrelated donor (URD) stem cell transplantation in particular. Outcomes following URD transplants depend mainly upon the indication and urgency of transplant, age and comorbidities of recipients, CMV matching/mismatching between donor and the recipient and degree of HLA matching. High resolution for human leucocyte antigens (HLA) should be performed to choose the most appropriate donor. HLA disparity between donor and recipient increases the risk for graft failure, graft versus host disease (GVHD) and mortality. Early complications develop within the first 100 days after AHSCT and include bone marrow suppression, veno-occlusive disease (VOD), mucositis, acute GVHD, bacterial infections, Cytomegalovirus (CMV), Herpes simplex virus (HSV) and invasive fungal infections (IFI). Late complications develop 100 days after AHSCT and include chronic GVHD, encapsulated bacterial infections, IFI, Varisella zoster virus (VZV) and late CMV infections. Despite the advances in HLA typing and GVHD prophylaxis, GVHD still remains as one of the most common complications of AHSCT. In unrelated transplantation, the incidence of grade II–IV acute GVHD is nearly 40–60% in 100 days, and the incidence of 5-year chronic GVHD is 40%. The most common cause of death after AHSCT is relapse. Oppurtunistic infections including CMV infection and IFI and GVHD are the most causes of transplant related mortality (TRM). This study aims to investigate the effect of ATG prophylaxis, conditioning regimen type, stem cell source, HLA disparity, as well as CMV infection, IFI, acute and chronic GVHD on transplant outcomes, overall survival (OS), progression free survival (PFS) and Graft versus host disease free survival (GFS) in 65 patients, who underwent unrelated donor AHSCT at Istanbul University Istanbul Medical Faculty Hematopoietic Stem Cell Transplantation Unit. Material and methods: Our study group included 65 patients (40 male, 25 female) who underwent unrelated AHSCT from unrelated donor at Istanbul University Istanbul Medical Faculty Hematopoietic Stem Cell Transplantation Unit between April 2003 and March 2020. Data recorded include demographic features, use of ATG as GVHD prophylaxis, conditioning regimen, stem cell source, HLA disparity, the presence of CMV infection, IFI, acute and chronic GVHD as well as overall survival (OS), progression free survival (PFS) and GVHD Free Survival (GFS). OS (months) was defined as the time from AHSCT to death or to the last follow up. PFS (months) was defined as the time from AHCST to disease relapse. GFS (days) was defined as the time from AHSCT to development of acute or chronic GVHD. Grading of acute GVHD was made by Modified Seattle Glucksberg criteria. Staging of chronic GVHD was made according to NIH consensus criteria and patients were classified as mild, moderate and severe GVHD. VOD was diagnosed according to Modified Seattle Criteria. Severity of VOD was determined according to EBMT criteria. After whole-blood collection, blood tubes with EDTA was centrifuged. Blood was collected into EDTA tubes and then centrifuged. Plasma sample was frozen at -20°. CMV quantitation test of plasma relies on extraction using QIAsymphony DSP Virus/Pathogen Mini Kit (QIAGEN, Germany) and QIASYMPHONY SP-AS (QIAGEN, Germany) assay and amplification and quantitation using Artus® CMV QS-RGQ Kit (QIAGEN, Germany) and Rotor-Gene Q (QIAGEN, Germany) assay. The method was real-time PCR (RT-PCR). The sensitivity of Kit was 43 copies/ml and range of quantitation was between 80 to 108 copies/ml. Internal control was added before extraction to detect false negativity. Three levels of probability to the diagnosis of invasive fungal infection was assigned: probable, possible and proven. Proven IFI was diagnosed İf there was histological and/or mycological evidence from tissue biopsies or resection material or detection of the pathogen from normally sterile specimens including blood cultures. Probable IFI was diagnosed in the presence of at least 1 host factor, clinical feature and mycologic evidence (serum galactomannan positivity). Possible IFI was diagnosed in the presence of host factors and findings compatible with IFI (radiological findings). All statistical analysis including clinical and laboratory parameters was performed using SPSS. All p-values<0.05 were considered statistically significant. OS curves of AHSCT were prepared by the 'Kaplan-Meier' method to estimate OS, PFS and GFS. Comparison of OS between the groups was performed using 'log rank' test. Results: A total of 65 patients (25 female (38.5%), 40 male (61.5%) were enrolled. The mean age of donors at time of AHSCT was 33.15±10.49 (SD:18-57). The actual mean age and the mean age at time of AHSCT were 36.26±10.92 (SD: 19-59) and 33.43±11.09 (SD: 16-58), respectively. The mean age between male and female patients showed no statistical significance (p=0.809). The most common diagnoses were acute myeloid leukemia (AML) (35.4%), acute lymphoblastic leukemia (ALL) (18.5%) and myelodysplastic syndrome (MDS) (16.9%). The stem cell source was peripheral blood in 59 patients (90.2%) and bone marrow in 6 patients (9.8%). Pretransplant CMV-IgG was positive in 56 (90.5%) of the recipients and in 39 (62.9%) of the donors. Myeloablative (MA) conditioning regimen and RIC were used in 73.8% and 26.2% of the patients, respectively. ATG containing conditioning regimen was used in 42 patients (64.6%). Of the donors, 61.5% were HLA full matched unrelated and 38.5% were 9/10 matched unrelated. The mean period of follow up was 31.90±41.82 months (SD=1-204). The mean day of CMV reactivation was 52.18±73.606 (SD=10-515). 76.9% (n=50) of the patients developed CMV infection. 36 (55.4%) had only CMV viremia and 14 (21.5%) developed concomitant CMV viremia and CMV disease. In the whole group, 50 patients developed CMV viremia (76.9%) and 14 developed CMV infection (21.5%). The rate of recurrent CMV infection was 38.5%. IFI was documented in 43 patients (66.2%). Of the 43 patients diagnosed with IFI, 2 (3.1%) were proven, 11 (16.9%) were probable and 30 (46.2%) were probable. 33 (50.8%) patients and 30 (46.2%) patients developed acute GVHD and chronic GVHD, respectively. Severe VOD, moderate VOD and mild VOD was diagnosed in 4 patients (6.2%), 4 patients (6.2%) and 5 patients (7.7%), respectively. 13 (20.1%) patients developed VOD. Mortality rate on day 100 after AHCST was 13.8% (n=9). During the mean follow up period of 31.9 months, 17 patients (26.2%) relapsed and 38 (58.5%) patients were deceased. Patients with myeloid malignancies showed a trend for a higher rate of AHSCT from 9/10 HLA matched donors than 10/10 HLA matched donors compared to patients with benign disease or lymphoid malignancies (80%, 50%; 8%, 10%; 12%, 40%, respectively; p=0.061). Patients with 9/10 HLA matched donors showed a significantly shorter period of thrombocyte engraftment (14.26±3.96, 17.56±8.01 days, respectively; p=0.034). Period of thrombocyte engrafment was significantly shorter in patients who received MA conditioning regimen than in patients who recieved RIC regimen (15.56±6.13, 19.59±7.85 days, respectively; p=0.037). Time to development of chronic GVHD (day) was significantly shorter in patients who received MA conditioning regimen (mean 1019.81±437.57, 1846.24±447.64 days, respectively; p=0.010). Patients with benign disease showed a trend for a higher rate of AHSCT with RIC compared to patients with myeloid and lymphoid malignancies (35.3%, 0; 41.23%, 68.8%; 23.5%, 31.2%; p=0.052). The rate of acute GVHD was signicantly increased in patients who underwent MA conditioning regimen compared to RIC (76.5%, 39.6%, respectively; p=0.009). The rate of mortality was significantly higher in patients who underwent MA conditioning regimen compared to RIC regimen (66.7%, 33.3%, respectively; p=0.024). The rate of IFI showed a tendency to be higher in patients who underwent MA conditioning regimen compared to RIC (72.9%, 47.1%, respectively; p=0.053). Hematopoietic Cell Transplantation-specific Comorbidity Index (HCT-CI) was significantly higher in patients receiving ATG prophylaxis compared to patients without prophylaxis (1.85±1.58, 0.78±1.04, respectively; p=0.002). In patients receiving ATG prophylaxis, the rate of use of RIC regimen was significantly higher than patients without prophylaxis (38.1%, 4.3%, respectively; p=0.003). The rate of use of ATG prophylaxis showed a trend to be higher in benign diseases compared to myeloid and lymphoid malignancies (38.1%, 4.3%; 64.3%, 56.5%; 21.4%, 43.5%, respectively; p=0.052). The rate of chronic GVHD showed no significant difference in patients who received ATG compared to patients who did not recieve ATG (42.9%, 52.2%, respectively; p=0.471). On the other hand, the rate of severe chronic GVHD was significantly lower in patients receiving ATG prophylaxis compared to those not receiving ATG prophylaxis (33.4%, 91.7%, respectively; p=0.006). The rate of death was significantly lower in patients receiving ATG prophylaxis compared to those not receiving ATG prophylaxis (47.6%, 78.3%, respectively; p=0.017). The rate of use of defibrotide 40 mg/kg/day in patients diagnosed with VOD was significantly higher in in patients receiving ATG prophylaxis compared to those not receiving ATG prophylaxis (100%, 50%, respectively; p=0.021). In patients with acute GVHD, the number of CMV infections were significantly higher (1.73±1.15, 1.09±1.14, respectively; p=0.03). OS was significantly shorter in patients with acute GVHD compared to patients without acute GVHD (mean: 35.64±8.83 months; 95% CI:76.78-149.67 and 104.12±20.92 months; 95% CI:18.57-52.79, respectively; p=0.017). PFS showed a trend towards being shorter in patients with acute GVHD with respect to patients without acute GVHD (mean: 34.46±8.90 months; 95% CI:17.01-51.91 and 94.50±22.82 months; 95% CI:49.76-139.25, respectively; p=0.058). Our findings demostrate that the rate of acute GVHD was significantly increased in patients who underwent MA conditioning regimen compared to RIC regimen (87.8%, 59.4%, respectively; p=0.009). The rate of CMV infection was found to be significantly increased in the presence of acute GVHD (%90.9, %66, respectively; p=0.007). The rate of CMV disease showed no difference between patients with acute GVHD and those with no acute GVHD (30.4%, 12.5%, respectively; p=0.081). Patients with acute GVHD had a significantly higher rate of chronic GVHD compared those with no acute GVHD (63.6%, 28.1%, respectively; p=0.004). The rate of death was significantly higher in patients with acute GVHD with respect to patients with no acute GVHD (75.8% and 40.6%, respectively; p=0.004). In patients with acute GVHD, the use of defibrotide as VOD prophylaxis was statistically higher compared to patients with no acute GVHD (63.6%, 37.5%, respectively; p=0.035). Also, the rate of IFI was signicantly higher in patients with acute GVHD compared to patients with no acute GVHD (81.8%, 50%, respectively; p=0.007). The incidence of female donor was signficantly higher in acute GVHD group (53.1%, 25.8%, respectively; p=0.027). Patients with myeloid malignancies have a statistically higher rate of acute GVHD compared to patients with lymphoid malignancies and benign hematological diseases (75.8%, 24.2%, 46.8%, 0, respectively; p=0.011). Patients who have minor blood group inconpability with their donors show a significantly lower rate of acute GVHD (p=0.048). In patients with chronic GVHD, the number of CMV infections was significantly higher (1.93±1.14, 0.97±1.04, respectively; p=0.001). In patients with chronic GVHD, the time to diagnosis of CMV infection was significantly longer than those with no GVHD (66.17±94.16, 32.29±14.64 days; p=0.042). The presence of CMV infection and viremia was significantly increased in patients with chronic GVHD compared to those with no GVHD (96.7%, 60.0%, respectively; both p values=0.0001). The rate of chronic GVHD was significantly decreased in patients with no CMV infection as compared to patients with CMV viremia and CMV viremia+disease (%3.3, %66.7, %30, respectively; p=0.001). The rate of acute GVHD was significantly higher in patients with chronic GVHD than patients with no chronic GVHD (70%, 34.3%, respectively; p=0.0005). Mortality at day 100 after HSCT was significantly lower in patients with chronic GVHD than patients with no chronic GVHD (3.3%, 22.9%; p=0.023). The rate of IFI infection was significantly higher in in patients with chronic GVHD than patients with no chronic GVHD (83.3%, 51.4%, respectively; p=0.007). The number of CMV infections was significantly higher in patients with IFI than those with no IFI (mean:1.70±1.20, 0.86±0.94, respectively; p=0.006). Time to development of CMV infection was significantly shorter in patients with IFI than those with no IFI (mean: 38.17±20.67, 87.36±132.19 days, respectively; p=0.033). Time to development of GVHD was significantly shorter in patients with IFI than those with no IFI (mean: 336.95±95.66, 3389.40±762.90 days, respectively; p=0.002). Patients with IFI show a tendency towards a higher rate of a history of MA conditioning regimen compared to those with no IFI (81.4%, 59.1%, respectively; p=0.053). The rate of CMV disease was significantly higher in patients with IFI than those with no IFI (32.6%, 0, respectively; p=0.007). The rate of CMV viremia+CMV disease was significantly higher in patients with IFI than those with no IFI (32.6%, 0, respectively; p=0.004). The rate of CMV viremia and CMV infection show a trend to be higher in patients with IFI than those with no IFI (83.7%, 63.6%, respectively; both p values=0.069). The rates of acute and chronic GVHD were significantly higher in patients with IFI than those with no IFI (62.8%, 27.3%; 58.1%, 22.7%, respectively; both p values=0.007). The rate of IFI was significantly higher in the presence of severe chronic GVHD compared to moderate and mild chronic GVHD (64.0%, 20.0%, %16, respectively; p=0.027). The rate of IFI was significantly higher in patients receiving VOD prophylaxis than those with no VOD prophylaxis (60.5%, 31.8%, respectively; p=0.029). Time to development of GVHD was significantly shorter in patients with CMV infection than those with no CMV infection (mean:950.53±402.96, 2350.57±474.95 days, respectively; p=0.009). The rate of acute GVHD was significantly higher in patients with CMV infection than those with no CMV infection (60.0%, 20.0%, respectively; p=0.007). The rate of chronic GVHD was significantly higher in patients with CMV infection than those with no CMV infection (58.0%, 6.7%, respectively; p=0.0001). Mortality at day 100 was significantly higher in patients without CMV infection than those with CMV infection (33.3%, 8.0%, respectively; p=0.013). The incidence of IFI showed a tendency to be higher in patients with CMV infection than those with no CMV infection (72%, 47.7%, respectively; p=0.069). OS was significantly longer in patients with HCT-CI<4 compared to patients with HCT-CI≥4 (mean: 76.98±13.84 months; 95% CI: 54.80-106.14 and 18.71±6.53 months; 95% CI: 6.65-32.48; p=0.013). GFS was significantly longer in patients with HCT-CI<4 compared to patients with HCT-CI≥4 (mean:1882.82±448.98 days; 95% CI:1002.81-2762.82 and 65.33±21.76 days; 95% CI: 22.67-107.99, respectively; p=0.049). OS showed a trend to be lower in patients who received MA conditioning regimen compared to patients receiving RIC regimen (mean: 59.56±13.23 months; 95% CI: 36.02-87.69 and 71.43±13.88 months; 95% CI:43.89-98.42, respectively; p=0.053). GFS was significantly shorter in patients who received MA conditioning regimen compared to patients receiving RIC regimen (mean:1019.81±437.57 days; 95% CI: 162.16-1877.47 and 1846.24±447.64 days; 95% CI: 968.85-2723.63, respectively; p=0.010). GFS was significantly shorter in patients with CMV infection compared to patients with no CMV infection (mean:950.53±402.96 days; 95% CI:160.72-1740.33 and 2350.57±474.95 days; 95% CI:1419.65-3281.49, respectively; p=0.009). GFS was significantly shorter in patients with CMV disease compared to patients with no CMV disease (mean: 141.66±41.62 days; 95% CI:60.08-223.23 and 1955.42±474.35 days; 95% CI:1025.68-2885.16, respectively; p=0.005). GFS was significantly shorter in patients with CMV viremia+disease compared to patients with CMV viremia and patients with no CMV infection (mean: 141.66±41.62 days; 95% CI:60.08-223.23; 1055.54 ±463.29 days; 95% CI:147.48-1963.59 and 2400.82±462.99 days; 95% CI:1493.36-3308.24, respectively; p=0.017). GFS was significantly shorter in patients with IFI compared to patients with no IFI (mean: 336.95±95.66 days; 95% CI: 149.44-524.46 and 3389.40±762.90 days; 95% CI:1894.11-4884.68, respectively; p=0.002). OS was significantly shorter in patients with acute GVHD compared to patients with no acute GVHD (mean: 35.64±8.83 months; 95% CI:76.78-149.67 and 104.12±20.92 months; 95% CI:18.57-52.79, respectively; p=0.017). PFS showed a tendency to be shorter in patients with acute GVHD compared to patients with no acute GVHD (mean: 34.46±8.90 months; 95% CI:17.01-51.91 and 94.50±22.82 months; 95% CI:49.76-139.25, respectively; p=0.058). Discussion: Unrelated donor transplantation is a feasible option when an HLA-matched sibling is not available. With current methodologies for molecular HLA typing and supportive care tools, patient outcomes are comparable to results achieved when a sibling donor is available. Although the morbidity, especially from acute GVHD, is higher with matched unrelated donors as compare to sibling donors, significant improvements have occurred over time and recent studies show that the 5 year overall survival approaches that of matched sibling donor transplantation. CMV infection and IFI are important causes of morbidity and mortality after AHSCT. GVHD increases transplant related mortality and hence decrease quality of life and may result in decrease in OS. Our results demostrate that CMV infection, IFI, acute and chronic GVHD influence the clinical course and survival after AHSCT. Our study provides the real life long follow up data of patients who underwent AHSCT from unrelated donors. Studies of IFI after AHSCT mostly include data from patients with proven or probable IFI. Since these studies exclude patients with possible IFI, they do not clearly reflect the real incidence of IFI. Our study included patients with possible IFI after AHSCT and therefore may reflect the true incidence of IFI and its influence on clinical outcomes. Our study not only provides data regarding the incidence of acute and chronic GVHD and their effects on clinical course but also provides insights to the factors influencing GFS. In our study group of patients undergoing unrelated HSCT, our findings suggest that recipients with HCT-CI≥4 are associated with worse OS and GFS. Our data also demonstrate that ATG prophylaxis decrease the rate of severe chronic GVHD and the rate of mortality. Our results confirm that MA conditioning regimen is associted with acute GVHD, increase in mortality and short GFS. Although our findings do not demonstrate a direct effect of CMV infection on OS,we have observed that CMV induces acute and chronic GVHD and thus decreases quality of life and is associated with short GFS. Our observations have also confirmed that IFI is associated with recurrent CMV infection, acute and chronic GVHD and short GFS. Our results also suggest that acute GVHD is associated with CMV infection, recurrent CMV infection, chronic GVHD and poor OS while chronic GVHD is associated with CMV infection, recurrent CMV infection and IFI and result in a decrease in quality of life.