RESULTS
Twenty-eight patients underwent LDLT for HB during the study period. A
total of 21 (75%) patients were classified as PRETEXT IV, 4 (14.3%)
patients as PRETEXT III and 3 (10.7%) patients were rescue LT. All of
them received pre-LT Qtx: 13 (46.4%) received PLADO, 10 (35.8%)
received alternative regimens, and for 5 (17.8%) patients the Qtx
regimen was not informed. Eighteen patients were event-free, and were
included in the Event-No group. A total of 10 patients experienced an
event (recurrence or death), and were included in the Event-Yes group
(Table 1).
Comparing the two groups, the statistically different pre-LDLT variables
were: time from diagnosis to the LDLT > 12 months
(p=0.09), post-Qtx AFP levels (p=0.07), and AFP post-Qtx reduction> 70% (p=0.05). The only statistically different
post-LDLT variable among groups was the presence of vascular invasion in
the explant analysis (p=0.002) (Table 1). Mean hospital stay was of 13.5+ 7.8 days in the Event-no group vs 12.5 + 4.6 days in the
Event-yes group (p=0.96).
Post-LDLT overall 5-y survival probability was 66% (Figure 1A) and 5-y
event-free survival probability was 63.9% (Figure 1B). In those
patients who achieved an AFP reduction post-Qtx ≥ 70%, 5-y event-free
survival was of 81.3% vs. 40% in those who did not (p=0.02) (Figure
2A). Patients in which the time from HB diagnosis to LDLT was> 12 months had a 5-y event-free survival rate of
40% vs 77% for those who performed the LDLT in <12 months
from the diagnosis (p=0.01; Figure 2B). Patients who received the PLADO
pre-LDLT Qtx regimen (high-risk SIOPEL) had a 5-y event-free survival of
76.9% versus 50% for those who received other Qtx regimens
(p=0.14) (Figure 2C). The 5-y event-free survival in primary LDLT was
67.7% versus 33.3% for those who underwent rescue LT (p=0.04)
(Figure 2D). In the liver explant analysis, the presence of vascular
invasion also impacted in the 5-y event-free survival: no vascular
invasion (82.4%) vs microvascular invasion (50%) vs macrovascular
invasion (20%) (p=0.04).
In order to evaluate the performance of the test (AFP reduction ≥70%)
with the occurrence of the event (recurrence/death), an ROC curve
analysis was made, and the calculated AUC was of 0.8 (Figure 3).
The pre-LDLT factors identified that were associated with the occurrence
of the event (recurrence/death) underwent a Cox-Regression analysis. The
hazard-ratios (HR) were calculated in order to create a scoring system
(Figure 4). Each risk factor (AFP reduction < 70%, time from
diagnosis to LDLT > 12 months, rescue LT) represents
1 point in the pre-LDLT scoring system to predict the probability of the
event: no risk factor present (15.4%), one risk factor present (33.3%)
and > 2 risk factors present (66.7%), p=0.02
(Figure 5).