72 patients were included with the average age of 60.3 years and 61.1% were male, of which all had renal involvement and 48.6% had cardiac amyloidosis, with 38.8% were at mayo stage III. 59 patients(81.9%) were alive at the last follow-up, with the median follow-up duration being 22 months. Median overall survival by Kaplan-Meier analysis for the whole cohort was not reached.1-year OS rate was 87.5%, and 79.3% of patients did not progress to MOD at 1 year (shown in Figure 1).
Comparison between prognosis in Bortezomib and Thalidomide based regimen
Baseline characteristics were not statistically different between the 2 groups (detailed in Table 1). The median follow-up time was 18 months in bortezomib and 28 months in thalidomide. Patients under bortezomib-based regimen received median 4 cycles of treatment. Among those who received thalidomide-based therapy, 60% of patients under CTD treatment, others received TD regimen for at least 6 months. 3 patients treated with TD regimen then switched to VD or VCD as a second-line regimen due to no hematologic response or organ dysfunction progression. There were 6 patients in VCD regimen who received second-line therapy,of which 2 patients switched to CTD regimen, 2 patients received Lenalidomide, dexamethasone (RD) or Pomalidomide, dexamethasone (PD) regimen, and 2 patients under autologous stem cell transplantation after achieving VGPR.
By intention-to-treat analysis,the overall hematologic response rate was 67.4% in bortezomib-based therapy, which was not statistically different compared to thalidomide-based regimen but complete responserate (27.9%) and overall responses at 3 months (51.2%) or 6 months (65.1%) were higher in patients treated with bortezomib-based regimen. (shown in Figure 2 and Table 2)
Median OSwas not reached in the bortezomib-based regimen and 1 year survival rate was 90.4%, with no difference compared to thalidomide-based regimen. However, MOD-PFS was significantly improved in the bortezomib-based regimen (1-year MOD-PFS was 86.2% and 64.2% respectively, P=0.026). The survival curve was shown in Figure 3a-b.
Patients receiving bortezomib therapy tended to achieve higher renal responses at 6 months and 12 months (15.4% and 28.8% respectively), but no statistical difference was found compared to thalidomide therapy (Figure 4). There was no significant difference between the two groups regarding the incidence of ESRD progression. By the end of follow-up, a total of 7 patients progressed to dialysis-dependent,and 1-yearrenal survival was 93.7% and 89.5% respectively in the bortezomib group and thalidomide group(P>0.05).The survival curve for renal survival was shown in Figure 5.
Comparison between prognosis in CVD and VD regimen
VD regimen was given in 36 patients and VCD in 16 patients. Patients were not matched for baseline characters but there were no statistical differences in baseline cardiac and renal function. No difference was noticed regarding to the numberof treatment cycles or bortezomib doses between VD and VCD group. Overall response rate was 83.3% and 61.3% respectively and higher VGPR rate was revealed in VCD regimen(83.3% vs 38.7%, P=0.022). Theaddition of pulsed cyclophosphamide to VD regimen was not associated with longer OS or MOD-PFS(shown in Supplement Table 1).
Prognosis in Mayo stage III patients
Among 28 patients with Mayo stage III, 19patients were treated with VD regimen and the other 9 patients were treated with TD or CTD regimen. No difference was noticed in overall response rates, but renal and cardiac response rates were higher in patients treated with bortezomib-based regimen(Supplement Table 2). Median OS for all Mayo stage III patients was not reached with 1 year OS rate of 68.8%. 3 patients died within the first 3 months from initiation of therapy. Landmark analysis at 3 months revealed that Mayo stage III patientswho received VD regimen had improved overall survival, with 1-year survival rate being 86.7% compared with 62.5% inthe thalidomide-based regimen(P=0.048). Kaplan Meier survival curve for Mayo stage III patients who survived more than 3 months was shown in Figure 6.
Adverse event
Incidence of all-grade adverse events for the whole cohort was 45.8%, of which 29.2% were >3 grade. Common adverse events include fluid retention,infection,and weakness. The incidence of hematological toxicity was low in both treatments.And higher proportion(30.0%) of patients in the Thalidomide-based regimen endured fluid retention compared to bortezomib(P<0.05), of which 5 patients needed CRRT to reduce volume overload(shown in Supplement Table 3).
Risk factors for overall survival
ROC curves were used to determine the cut-off values for the continuous variables, and the best predicted values were 1100ng/L for NT-proBNP (area under the curve: 0.807, P=0.001), 0.0625ng/ml for serum troponin T (area under the curve: 0.854, P<0.001), and 50mg/L for serum dFLC (area under the curve: 0.612, P=0.021),ROC curves for different variables are shown in Supplement figure 1.
13(18.1%) patients died during follow up. Risk factors for inferior OS by uni-variate analysis included involved organs ≥3,NT-proBNP>1100ng/L, troponin T>0.0625ng/ml, IVS≥13mm, cardiac dysfunction progression, eGFR<50ml/min/1.73m2, dFLC>50mg/l and Mayo stage III(compared to Mayo stage I),while achievement of VGPR was associated with reduced risk of death. Detailed results are shown in Table 3.
Risk factors for renal survival
7 patients(9.7%) progressed to dialysis during follow-up. Progression to renal replacement therapy was used as the endpoint to analyze the risk factors for renal prognosis. By uni-variate analysis, baseline eGFR<50ml/min/1.73m2 ,eGFR decrease>30% ,and renal stage 2 or 3(compared to renal stage 1) were associated with risk of renal failure.However, there was no difference in the risk of renal failure between bortezomib and thalidomide treatment (HR=0.33, 95%CI:0.06-1.65,P=0.175). Detailed results were shown in Table 4.