Patient and donors
The median follow-up in survivors was eight years (range 1-22 years). However, as expected, the median follow-up differed in the consecutive GVHD prophylaxis cohorts; CsA-MTX/MMF groups (N 271, 13-years (range 6-22), tacro-siro (N 145, 6-years (range 1-12), and PTCy-tacro (N 68, 3-years (range 1-6) [p 0.001]. Since analyzing the impact of the introduction of PTCy as GVHD prophylaxis was a major study objective, outcomes were compared at 3-years post-transplant since this the shortest median survival of the GVHD prophylaxis strategies used over time. Patient and donor characteristics are shown in table 1.
Engraftment and early mortality
The overall CumInc of neutrophil (>0.5x109/L) and platelet recoveries (>20x109/L) were 97% and 90% at 30 days. The median time to neutrophil recovery was 17 days (range 10-76) and platelet recovery 14 days (range 6-104), with 54 patients (11%) never reaching a platelet count below 20x10e/L. Primary graft failure occurred in six patients (N 4 in the PTCy-tacro, and one each in CsA-MTX/MMF and tacro-siro cohorts). Five patients died without engraftment recovery (three due to infection, one of veno-occlusive disease, and one relapsed at +29 days).
Patients who received PTCY-tacro as GVHD prophylaxis had much longer times to engraftment; the median time to neutrophil and platelets engraftment was +22 days (range 16-42) and +26 days (range 12-104) while non-PTCY recipients recovered neutrophils at +17 (range 10-76) and platelets at + 15 days (range 6-73) [p 0.0001] (Figure 1). G-CSF was started in those who had < 0.1 x 109/L neutrophils on day +21.
Acute and chronic graft versus host disease
CumInc of grade II-IV and III-IV aGVHD were 25% and 15% at 100 days and 30% and 17% at 200 days, with a median day of onset of 41 days (range 5-200) and 38 days (range 7-200), respectively.
In UVA variables which had a statistically significant impact on the occurrence of grade II-IV and III-IV aGVHD include for the former female donor to male recipient, lymphoid underlying disease, disease status at SCT and the GVHD prophylaxis used (table2); of note the lower 200-day incidence of grade III-IV aGVHD in the PTCY-tacro group (6% vs.12% vs.23%) (table2 and figure2a).
In the MVA variables which impacted on grade II-IV aGVHD were female-to-male (p 0.03, HR:1.5; CI95%: 1.04-2.18), lymphoid vs myeloid underlying disease (p 0.03, HR:1.4; CI95%: 1.04-2), while GVHD prophylaxis with PTCY-tacro reduced the risk (p 0.03, HR:0.7; CI95%: 0.6-0.9). In addition, PTCY-tacro as GVHD prophylaxis was the only variable protective for grade III-IV aGVHD in MVA (p 0.0001; HR:0.5; CI95%: 0.3-0.7).
The CumInc of overall and moderate-severe cGVHD were 48% and 36%, respectively (the median month of onset of cGvHD was 6 months (range 2-135). In UVA variables which impacted moderate-severe cGVHD (table2) were female-to-male, myeloid underlying disease and GVHD prophylaxis other than PTCy-tacro. In MVA variables which protected against moderate-severe cGVHD were lymphoid vs myeloid underlying disease (p 0.007; HR:0.6; CI 95%: 0.4-0.8) and GVHD prophylaxis with PTCy-tacro (p 0.001; HR:0.7; CI 95%: 0.6-0.8); of note the lower 3-year incidence of moderate-severe cGVHD in the PTCY-tacro group (8% vs.40% vs.38%) (table2 and figure2b).
Non-relapse mortality and relapse
The CumInc of NRM was 10% (95%CI; 2-18%) at 100 days, 23% at 3-years (95%CI; 19-27%), and 29% (95%CI; 25-33%) at 8-years (figure 3a). In the UVA (table3), the variables with an impact on NRM were HCT-CI-age (<2), the EBMT-score (<3) (figure 3b), female-to-male, patient CMV serology, disease status at transplant and type of GVHD prophylaxis (table3). In MVA variables with an impact on NRM included HCT-CI-age >3 (p 0.002; HR:1.9; CI95% 1.25 to 2.8), EBMT score >4 (p 0.005; HR:1.7; CI95% 1.2-2.4), donor female to male recipient (p 0.007; HR:1.64; CI95% 1.15-2.4) and positive patient CMV serology (p 0.03; HR:1.7; CI95%, 1.1-2.8).
The causes of NRM according to GVHD prophylaxis are shown in figure4a, although this variable did not show an impact on NRM as of 3-years of follow-up.
The overall CumInc of relapse was 9% (95%CI; 5-11%) at 100 days, 28% (95%CI; 24-32%) at 3-years and 30% (95%CI; 26-34%) at 8 years (figure3c). In the UVA the DRI and response at SCT had an impact on relapse (table3), while only the DRI high-very high (p 0.0001; HR:3; CI95%: 2.1-4.26) was shown as an independent factor in the MVA (figure3d).
Overall survival and progression free survival
OS was 87% (95%CI; 85-89%) at 100 days, 56% (95%CI; 52-57%) at 3-years and 43% (95%CI; 45-49%) at 8-years. The variables with an impact on the 3-year survival in UVA were HCT-CI-age (<2), DRI (low-intermediate), female to male, type of GvHD prophylaxis, first and second CR at SCT, while the EBMT-score showed a trend (<3 vs >4 p 0.09) (table 3). In MVA HCT-CI-age <2 (p 0.021, HR:0.7, CI95%: 0.6-0.9), low-intermediate DRI (p 0.0001, HR:0.6, CI95%: 0.5-0.7), female donor to male recipient (p 0.23, HR:1.4, CI95%: 1.1-1.7) and type of GVHD prophylaxis (negative impact if CsA-MTX/MMF were used; p 0.014, HR:1.7, CI95% 1.1-2.7, with no differences between tacro-siro and PTCy-tacro in the risk of 3-year OS).
The causes of mortality distribution over time are shown in figure4b. Secondary malignancies (SM) were the cause of 13 deaths (4%), nine patients in the CsA-MMF/MTX and four patients in the tacro-siro group.
The PFS was 81% (95%CI; 79-83%) at 100 days, 47% (95%CI; 45-49%) at 3-years and 39% (95%CI; 40-44%) at 8-years. The variables that had an impact on the 3-year PFS in UVA were HCT-CI-age (<2), DRI, female to male recipient, type of GVHD prophylaxis and disease status at SCT (table3), while EBMT-score showed only a trend. In MVA variables with an impact on the 3-year PFS included low-intermediate DRI (p 0.0001, HR:0.5, CI95%: 0.44-0.7) and type of GVHD prophylaxis (with a negative impact of CsA-MTX/MMF vs. other; p 0.005, HR:1.7; CI95%: 1.2-2.7; with no differences between tacro-siro and PTCy-tacro in the risk of 3-year PFS).
LONG TERM COMPLICATIONS
To study late complications, we performed a landmark analysis of patients who were alive and progression-free 1-year post-RIC alloHSCT, considering day 0 as day +365, and the results refer to follow-up which equals days post-transplant -365 days. Three hundred thirty-five patients were included in the 1-year landmark analysis. Late complications which occurred in these patients were identified by review of each patient’s electronic medical records. Only new-onset complications which met the following criteria were included: (i) required surgery; (ii) required long-term medication; (iii) required any hospitalization; (iv) led to any physical disability; and (v) any secondary neoplasm(s). Infectious complications, relapse-related complications and chronic GVHD were not analyzed.
The 7-year landmark probability or CumInc of OS, PFS, NRM, and relapse were 60% (95% CI 57-63%), 56% (95% CI 53-59%), 15% (95% CI 11-19%) and 27% (95% CI 23-31%), respectively. One hundred ninety-seven patients (58%) developed at least one late complication (range 0-7), which occurred at a median of 37 months (range 4-237) beyond 1-year post-HSCT.
The most common complications, and their numbers and frequencies were as follows. Ophthalmological complications occurred in 20% of the patients (n=66), of which 91% consisted of cataracts. Skeletal events were diagnosed in 54 cases (16%), the most common being osteoporosis-related fractures, which represented 74% of these events, followed by 20% of avascular necrosis. Cardiac complications (ischemic heart disease, congestive heart failure or atrial fibrillation) occurred in 45 patients (15%), while new onset metabolic disorders (diabetes, arterial hypertension or hypercholesterolemia) occurred in 38 cases (11%). Liver or gastrointestinal disorders were diagnosed in 25 cases (7.5%), while non-infectious pulmonary complications occurred in 7% (n=22). Other complications included deep vein thrombosis (n=17, 5%), chronic renal failure (n=12, 4%), severe psychosis ((n=9, 3%), and non-GVHD-defining autoimmune disease (n=6, 2%).
A secondary malignancy (SM) developed in 44 patients (13%), of whom seven developed two SMs. The most common SM were solid tumors (n=46, 88%), followed by a hematologic malignancy in 5 cases (12%). The most common solid tumors were skin cancers (n=16), head and neck (n=6), prostate (n=5), colon or stomach (n=5), lung (n=3), gynecologic (n=3), breast (n=2), liver (n=2), pancreatic (n=1) and urinary bladder cancer (n=1).