Article ID: CJ-17-0510
Background: The Therapeutic Angiogenesis by Cell Transplantation (TACT) trial demonstrated the efficacy and safety of autologous bone marrow-derived mononuclear cells (BM-MNCs) in patients with critical limb ischemia (CLI). The present study aimed to assess the long-term clinical outcomes of therapeutic angiogenesis using autologous BM-MNC implantation under advanced medical treatment in Japan.
Methods and Results: The study was retrospective, observational, and non-controlled. We assessed no-option CLI patients who had BM-MNC implantation performed in 10 institutes. Overall survival (OS), major amputation-free (MAF), and amputation-free survival (AFS) rates were primary endpoints of this study. The median follow-up duration was 31.7 months. The 10-year OS rate was 46.6% in patients with arteriosclerosis obliterans (ASO) (n=168), 90.5% in patients with thromboangiitis obliterans (TAO) (n=108), and 67.6% in patients with collagen disease-associated vasculitis (CDV) (n=69). The 10-year MAF rate was 70.1%, 87.9%, and 90.9%, respectively. The 10-year AFS rate was 37.8%, 80.9%, and 61.2%, respectively. Major adverse cardiovascular events occurred in 6.0% of patients with ASO, 1.9% of patients with TAO, and no patients with CDV.
Conclusions: Therapeutic angiogenesis using autologous BM-MNC implantation may be feasible and safe in patients with no-option CLI, particularly those with CLI caused by TAO or CDV.
Arteriosclerosis obliterans (ASO) caused by atherosclerosis is rapidly becoming a worldwide health concern because of the aging population. Critical limb ischemia (CLI) is the most advanced stage of ASO, thromboangiitis obliterans (TAO), and collagen disease-associated vasculitis (CDV). Many patients with CLI require limb amputation when conventional therapeutic options including bypass surgery or endovascular treatment (EVT) fail or are not indicated, particularly patients with TAO or CDV.1,2 Therefore, augmenting new blood vessel formation is an important way to salvage ischemic tissue.
Since Isner et al3 reported the efficacy of gene therapy in promoting angiogenesis using a vascular endothelial growth factor-A plasmid, therapeutic angiogenesis has shown promising potential for patients with CLI. After discovering that circulating endothelial progenitor cells (EPCs) mobilized from the bone marrow (BM) participate in postnatal neovascularization, we performed basic research related to therapeutic angiogenesis using EPCs or BM cells and described their angiogenic effects in ischemic tissues.4,5 In 2002, we published results from the Therapeutic Angiogenesis by Cell Transplantation (TACT) study, our first clinical pilot trial. Patients with CLI had significantly improved transcutaneous oxygen pressure values, resting pain, ankle-brachial index, and pain-free walking time following intramuscular implantation of autologous BM-derived mononuclear cells (BM-MNCs).6 We reported that this angiogenic cell therapy could induce long-term improvement in limb ischemia, extending the amputation-free interval and survival rates.7,8 The TACT trial demonstrated the efficacy and safety of autologous BM-MNC implantation in patients with CLI.
Subsequently, therapeutic angiogenesis using intramuscular BM-MNC implantation for CLI has been undertaken in clinical trials and as advanced medical care. Since the TACT study began over 10 years ago, numerous approaches to therapeutic angiogenesis using different treatment modalities, including BM-MNCs, granulocyte colony-stimulating factor-mobilized peripheral blood mononuclear cells, and G-CSF-mobilized cluster of differentiation CD34+ cells, have been attempted worldwide, with inconsistent results.9,10 In this study, we conducted a follow-up field survey involving patients with CLI who underwent BM-MNC implantation under advanced medical treatment in Japan to assess the long-term safety and clinical outcomes of therapeutic angiogenesis.
The study protocol was similar to that of the TACT study.6 CLI was classified by Fontaine stages III and IV. All patients with CLI who were intractable to any other treatment options (e.g., medical therapy, rehabilitation, sympathetic ganglion block, and revascularization by EVT or surgery), qualified for BM-MNC implantation. All patients underwent pre-assessment to determine whether they fulfilled any of the exclusion criteria, which included untreated coronary artery disease or cerebrovascular disease, clinical or laboratory signs of chronic or acute inflammation, a previous (5 years) or current history of neoplasms, diabetes with untreated retinopathy, age >80 years, the possibility of pregnancy, or the lack of informed consent.6 We obtained written informed consent from all patients, and the ethics committees of the participating hospitals approved the protocol.
Data CollectionWe developed a web-based database using FileMaker Pro (FileMaker, Inc., Santa Clara, CA, USA) and an electronic clinical record form (eCRF) to collect information about patient characteristics, previous medical history, adverse events, and outcomes on the last day on which the physician confirmed the patient’s status. All patient data were stored and analyzed independently at the Center for Advanced Medicine and Clinical Research, Nagoya University Hospital, Nagoya, Japan. In 2015, 14 Japanese hospitals participated in this part of the TACT trial. Data were collected with the permission of the Ministry of Health, Labour, and Welfare.
This study was retrospective, observational, and non-controlled. The primary endpoints were the rates of overall survival (OS), major amputation-free (MAF), and amputation-free survival (AFS). AFS in this study included OS and major amputation. The safety of the therapy was assessed in the context of major adverse cardiovascular events (MACE; death, non-fatal myocardial infarctions, decompensated heart failure, and stroke) and all-cause adverse events during the 6-month follow-up period after BM-MNC implantation.
Statistical AnalysisProportions are expressed as percentages, and continuous variables as median and interquartile range. In addition, the follow-up periods are expressed as median and range from minimum to maximum. After confirming the homoscedasticity of the ASO, TAO and CDV groups, categorical variables and continuous variables were compared between these groups using a non-parametric test (Kruskal-Wallis test). Time-to-event primary endpoints analyses were performed using the Kaplan-Meier method. To examine the variables associated with OS, MAF, and AFS within each group, the log-rank test or proportional hazard regression analysis was performed. The 95% confidence interval (CI) was calculated for OS, MAF, and AFS. To identify the independent prognostic factors for OS and MAF, multivariate logistic regression analysis was performed with the variables of the baseline characteristics in the 3 groups. All statistical analyses were performed using IBM® SPSS® software, version 23 (IBM Corporation, Armonk, NY, USA). A value of P<0.05 was considered statistically significant.
We included 374 patients (389 cases) from 10 hospitals in the survey. Of them, 15 (15 cases) underwent cell implantations twice during the follow-up period; 29 (29 cases) were excluded because 18 patients (18 cases) did not have CLI and 11 patients (11 cases) had erroneous data entries. Finally, a total of 345 patients, comprising 168 patients with ASO, 108 patients with TAO, and 69 patients with CDV, were included in this study (Figure 1). We assessed the long-term clinical outcomes from the first cell implantation in the 15 patients who underwent cell implantation twice. The baseline characteristics of these patients are shown in Table 1. Large differences were evident in the patients’ baseline characteristics. The average age, number of risk factors, complications associated with ischemic heart disease, and rates of chronic renal disease, including dialysis and cerebrovascular disease, were significantly higher in the ASO group than in the TAO and CDV groups. In addition, the TAO and CDV groups had more cases of patients who simultaneously required BM-MNC implantation to the lower and upper limbs than in the ASO group. Both the total number of implanted BM-MNC and the number of BM-MNC implanted per limb were highest in the TAO group (Table 2).
Flowchart of patient selection. ASO, arteriosclerosis obliterans; CLI, critical limb ischemia; CDV, collagen disease-associated vasculitis; TAO, thromboangiitis obliterans.
| ASO (n=168) |
TAO (n=108) |
CDV (n=69) |
P value | |
|---|---|---|---|---|
| Age, years | 67.2±9.8 | 48.7±12.3 | 54.8±14.0 | <0.01 |
| >70 years, n (%) | 71 (42) | 4 (3.7) | 8 (12) | <0.01 |
| Male, n (%) | 109 (65) | 95 (88) | 9 (13) | <0.01 |
| Weight, kg | 55.3±11.2 | 66.1±15.2 | 52.0±11.8 | <0.01 |
| BMI, kg/m2 | 21.5±3.6 | 23.5±4.5 | 21.3±4.0 | <0.01 |
| Hypertension, n (%) | 129 (77) | 20 (19) | 24 (35) | <0.01 |
| Hyperlipidemia, n (%) | 51 (30) | 20 (19) | 10 (14) | 0.01 |
| DM, n (%) | 87 (52) | 16 (15) | 8 (12) | <0.01 |
| DM with insulin, n (%) | 52 (31) | 3 (2.8) | 0 (0) | <0.01 |
| CKD, n (%) | 64 (38) | 3 (2.8) | 3 (4.3) | <0.01 |
| Hemodialysis, n (%) | 50 (30) | 1 (0.9) | 1 (1.4) | <0.01 |
| Smoking, n (%) | 102 (61) | 94 (87) | 19 (28) | <0.01 |
| Medical therapy, n (%) | ||||
| Peripheral vasodilator agents | 159 (95) | 99 (92) | 64 (93) | 0.61 |
| Antihypertensive agents | 138 (82) | 23 (21) | 31 (45) | <0.01 |
| Statins | 62 (37) | 11 (10) | 11 (16) | <0.01 |
| Oral hypoglycemic agents | 42 (25) | 8 (7.4) | 6 (8.7) | <0.01 |
| Complications, n (%) | ||||
| IHD, n (%) | 86 (51) | 1 (0.9) | 2 (2.9) | <0.01 |
| CVD, n (%) | 25 (15) | 3 (2.8) | 1 (1.4) | <0.01 |
| Aortic disease, n (%) | 7 (4.2) | 2 (1.9) | 1 (1.4) | 0.39 |
| Malignancy, n (%) | 5 (3.0) | 0 (0) | 0 (0) | 0.07 |
| Fontaine classification, n | ||||
| III/IV | 49/119 | 30/78 | 10/59 | 0.06 |
| Target limb, n (%) | ||||
| Lower limb | 162 (96) | 98 (91) | 48 (70) | <0.01 |
| Upper limb | 7 (4.2) | 25 (23) | 34 (49) | <0.01 |
| Lower and upper limbs | 1 (0.6) | 16 (15) | 13 (19) | <0.01 |
| Total no. of BM-MNC (×109) | 20.3±10.9 | 29.8±16.1 | 21.7±11.7 | <0.01 |
| No. of BM-MNC per weight (×108) | 3.81±2.11 | 4.70±2.73 | 4.23±2.28 | <0.01 |
| No. of BM-MNC per limb (×109) | 18.6±10.9 | 24.8±15.0 | 15.2±11.4 | <0.01 |
Data are presented as n (%) or median (interquartile range). The P values indicate the differences among the PAD, TAO, and CDV groups. ASO, arteriosclerosis obliterans; BM-MNC, bone marrow-derived mononuclear cells; BMI, body mass index; CDV, collagen disease-associated vasculitis; CKD, chronic kidney disease; CLI, critical limb ischemia; CVD, cerebrovascular disease; DM, diabetes mellitus; IHD, ischemic heart disease; PAD, peripheral artery disease; TAO, thromboangiitis obliterans.
| All (n=345) | ASO (n=168) | TAO (n=108) | CDV (n=69) | P value | |
|---|---|---|---|---|---|
| All-cause death, n (%) | 49 (14) | 38 (23) | 6 (5.6) | 5 (7.2) | <0.01 |
| All-cause adverse events, n (%) | 38 (11) | 28 (17) | 4 (3.7) | 6 (8.7) | <0.01 |
| MACE, n (%) | 12 (3.5) | 10 (6.0) | 2 (1.9) | 0 (0) | 0.04 |
| Death | 5 (1.4) | 4 (2.4) | 1 (0.9) | 0 (0) | 0.33 |
| AMI | 0 (0) | 0 (0) | 0 (0) | 0 (0) | – |
| Heart failure | 5 (1.4) | 5 (3.0) | 0 (0) | 0 (0) | 0.07 |
| Stroke | 2 (0.6) | 1 (0.6) | 1 (0.9) | 0 (0) | 0.73 |
| Other, n (%) | 26 (7.5) | 18 (11) | 2 (1.9) | 6 (8.7) | 0.02 |
| Severe bleeding, n (%) | 8 (2.3) | 5 (3.0) | 0 (0) | 3 (4.3) | 0.13 |
| Pelvic visceral disability, n (%) | 1 (0.3) | 1 (0.6) | 0 (0) | 0 (0) | 0.59 |
| Pelvic or sampling site pain, n (%) | 7 (2.0) | 5 (3.0) | 2 (1.9) | 0 (0) | 0.33 |
| Emergence of new tumor, n (%) | 1 (0.3) | 1 (0.6) | 0 (0) | 0 (0) | 0.59 |
| Severe infection | 2 (0.6) | 2 (1.2) | 0 (0) | 0 (0) | 0.35 |
| Worsening liver function | 11 (3.2) | 8 (4.8) | 1 (0.9) | 2 (2.9) | 0.21 |
| Worsening kidney function | 0 (0) | 0 (0) | 0 (0) | 0 (0) | – |
| Worsening retinopathy | 11 (3.2) | 8 (4.8) | 1 (0.9) | 2 (2.9) | 0.21 |
Data are presented as n (%). The P values indicate the differences among the PAD, TAO, and CDV groups. AMI, acute myocardial infarction; MACE, major adverse cardiovascular events. Other abbreviations as in Table 1.
The median follow-up duration of all patients was 31.7 months (range: 0–164.8 months, mean: 47.4 months). Figure 2 shows the Kaplan-Meier analysis of OS, MAF, and AFS rates following BM-MNC implantation. The primary endpoints were analyzed separately in patients with CLI caused by ASO, TAO, or CDV. The 1-, 5-, and 10-year primary endpoints in patients with ASO, TAO, and CDV were determined. In all patients, the 1-year OS, MAF, and AFS rates were 95.6%, 86.8%, and 82.7%, respectively. The 5-year OS, MAF, and AFS rates were 87.3%, 82.0%, and 71.8%, respectively. The 10-year OS, MAF, and AFS rates were 68.6%, 80.7%, and 59.1%, respectively (Figure 2A–C). The 1-, 5-, and 10-year OS rates were 92.0%, 75.4%, and 46.6% in patients with ASO, 98.9%, 97.6%, and 90.5% in patients with TAO, and 98.2%, 94.8%, and 67.6% in patients with CDV, respectively (Figure 2D). The 1-, 5-, and 10-year MAF rates were 78.7%, 74.0%, and 70.1% in patients with ASO, 93.2%, 87.9%, and 87.9% in patients with TAO, and 95.1%, 90.9%, and 90.9% in patients with CDV, respectively (Figure 2E). The 1-, 5-, and 10-year AFS rates were 71.6%, 55.0%, and 37.8% in patients with ASO, 92.1%, 85.6%, and 80.9% in patients with TAO, and 93.3%, 85.9%, and 61.2% in patients with CDV, respectively (Figure 2F). The OS, MAF, and AFS rates were significantly worse in patients with CLI caused by ASO than in patients with CLI caused by TAO or CDV. Major amputation and death occurred among patients with CLI caused by ASO within 6 months of BM-MNC implantation, and their frequencies had increased in this study.
(A–F) Kaplan-Meier analysis of overall survival, major amputation-free, and amputation-free survival following BM-MNC implantation in all patients (A–C) and in patients with ASO, TAO, and CDV (D–F). BM-MNC, bone marrow-derived mononuclear cells. Other abbreviations as in Figure 1.
We analyzed the primary endpoints in each group according to CLI severity, which was defined using the Fontaine classification. OS in all groups is shown in Figure 3A–C. OS was the worst in patients with ASO and CLI classified as Fontaine IV. However, it did not differ significantly compared with that in patients with ASO and CLI classified as Fontaine III (Figure 3A). The MAF and AFS rates were the worst in patients with ASO and CLI classified as Fontaine IV and were significantly worse those in patients with ASO and CLI classified as Fontaine III (Figure 3D,G) (P<0.001 for each). The MAF and AFS rates in patients with TAO or CDV and CLI classified as Fontaine IV did not differ significantly compared with those in patients with TAO or CDV and CLI classified as Fontaine III (Figure 3E,F,H,I). Thus, the Fontaine classification before BM-MNC implantation was associated with the post-implantation MAF rate in patients with CLI caused by ASO (P<0.01) but not in patients with CLI caused by CDV (P=0.54). Fontaine classification IV was associated with an increased MAF rate in patients with CLI caused by TAO, but there was no significant difference (P=0.16).
Kaplan-Meier analysis of overall survival (OS), major amputation-free, and amputation-free survival (AFS) following BM-MNC implantation based on the severity of critical limb ischemia. OS rate in (A) patients with ASO classified as Fontaine III or IV, (B) patients with TAO classified as Fontaine III or IV and (C) patients with CDV classified as Fontaine III or IV. Major amputation-free rate in patients with (D) ASO classified as Fontaine III or IV, (E) TAO classified as Fontaine III or IV and (F) CDV classified as Fontaine III or IV. AFS rate in patients with (G) ASO classified as Fontaine III or IV, (H) TAO classified as Fontaine III or IV and (I) CDV classified as Fontaine III or IV. Abbreviations as in Figures 1,2.
MACE and all-cause adverse events occurred in 3.5% and 11.0% of all patients, respectively. The occurrence of MACE was significantly higher in patients with ASO (6.0%) (P=0.04) than in patients with TAO (1.9%) or CDV (0%). The occurrence of all-cause adverse events was higher in patients with ASO (16.7%) (P<0.01) than in patients with TAO (3.7%) or CDV (8.7%). The most frequent adverse events in all patients were worsening liver dysfunction (3.2%) and retinopathy (3.2%), followed by severe bleeding/anemia (2.3%), and severe pain at the BM aspiration site (2.0%). No patients died, and no severe adverse events associated with BM-MNC implantation were noted.
Prognostic FactorsWe evaluated the prognostic factors affecting OS and major amputation in patients with CLI by multivariate logistic regression analysis using the following baseline variables: age, sex, Fontaine classification, hemodialysis (HD) and history of hypertension, hyperlipidemia, diabetes mellitus (DM), smoking, ischemic heart disease and cerebrovascular disease (Table 3). In all patients with CLI, age and male sex were independently associated with the OS rate (hazard ratio (HR): 0.93, 95% CI: 0.89–0.97; P<0.01 and HR: 0.36, 95% CI: 0.15–0.85; P=0.02, respectively). In addition, Fontaine classification IV was independently associated with the MAF rate (HR: 4.19, 95% CI: 1.55–11.3; P<0.01). Table 4 shows the prognostic factors affecting OS and major amputation for patients in each category. In patients with ASO, age was independently associated with the OS rate (HR: 0.94, 95% CI: 0.89–0.99; P=0.01). Moreover, Fontaine classification IV was independently associated with the MAF rate (HR: 5.78, 95% CI: 1.60–21.1; P<0.01). In patients with TAO and CDV, we could not analyze using all the variables because events such as death and major amputation were rare. Thus, we evaluated the prognostic factors only by variables that could withstand analysis in patients with TAO and CDV. In patients with TAO, age was independently associated with the OS rate (HR: 0.89, 95% CI: 0.79–0.99 and P=0.04). Fontaine classification IV tended to increase the MAF rate (HR: 11.8, 95% CI: 0.88–159 and P=0.06). In patients with CDV, no independent prognostic factors affecting OS and major amputation were found.
| Variable | All | ||
|---|---|---|---|
| HR | 95% CI | P value | |
| Overall survival | |||
| Age | 0.93 | 0.89–0.97 | <0.01 |
| Male | 0.36 | 0.15–0.85 | 0.02 |
| Hypertension | 2.28 | 0.99–5.26 | 0.06 |
| Hyperlipidemia | 1.18 | 0.55–2.52 | 0.67 |
| DM | 1.50 | 0.75–3.00 | 0.25 |
| Hemodialysis | 1.67 | 0.74–3.78 | 0.22 |
| Smoking | 0.67 | 0.30–1.48 | 0.32 |
| History of IHD | 0.79 | 0.37–1.65 | 0.52 |
| History of CVD | 1.89 | 0.74–4.81 | 0.18 |
| Fontaine IV | 1.53 | 0.70–3.35 | 0.29 |
| Major amputation-free | |||
| Age | 0.98 | 0.95–1.01 | 0.13 |
| Male | 0.69 | 0.31–1.51 | 0.35 |
| Hypertension | 2.03 | 0.98–4.21 | 0.06 |
| Hyperlipidemia | 0.91 | 0.42–1.97 | 0.81 |
| DM | 1.63 | 0.82–3.24 | 0.16 |
| Hemodialysis | 1.81 | 0.83–3.94 | 0.13 |
| Smoking | 0.65 | 0.31–1.39 | 0.27 |
| History of IHD | 0.56 | 0.26–1.22 | 0.15 |
| History of CVD | 0.44 | 0.12–1.62 | 0.22 |
| Fontaine IV | 4.19 | 1.55–11.3 | <0.01 |
Data are presented as hazard ratio (HR) and 95% confidence interval (95% CI). Abbreviations as in Table 1.
| Variable | HR | 95% CI | P value |
|---|---|---|---|
| (A) CLI caused by ASO | |||
| Overall survival | |||
| Age | 0.94 | 0.89–0.99 | 0.01 |
| Male | 0.47 | 0.17–1.27 | 0.14 |
| Hypertension | 2.00 | 0.67–5.97 | 0.22 |
| Hyperlipidemia | 1.20 | 0.50–2.89 | 0.68 |
| DM | 1.61 | 0.72–3.57 | 0.25 |
| Hemodialysis | 1.50 | 0.61–3.68 | 0.38 |
| Smoking | 1.08 | 0.42–2.77 | 0.88 |
| History of IHD | 0.90 | 0.40–2.01 | 0.79 |
| History of CVD | 1.41 | 0.50–4.00 | 0.52 |
| Fontaine classification IV | 2.08 | 0.82–5.30 | 0.12 |
| Major amputation-free | |||
| Age | 1.01 | 0.97–1.05 | 0.73 |
| Male | 0.93 | 0.33–2.59 | 0.89 |
| Hypertension | 1.08 | 0.41–2.85 | 0.87 |
| Hyperlipidemia | 0.90 | 0.33–2.40 | 0.83 |
| DM | 1.32 | 0.58–3.02 | 0.51 |
| Hemodialysis | 1.15 | 0.47–2.78 | 0.76 |
| Smoking | 0.72 | 0.27–1.94 | 0.51 |
| History of IHD | 0.47 | 0.20–1.10 | 0.08 |
| History of CVD | 0.12 | 0.01–0.93 | 0.04 |
| Fontaine classification IV | 5.78 | 1.60–21.1 | <0.01 |
| (B) CLI caused by TAO | |||
| Overall survival | |||
| Age | 0.89 | 0.79–0.99 | 0.04 |
| Hypertension | 0.57 | 0.05–6.00 | 0.64 |
| Hyperlipidemia | 1.84 | 0.20–17.1 | 0.59 |
| DM | 0.65 | 0.05–8.04 | 0.74 |
| Fontaine classification IV | 0.27 | 0.04–1.84 | 0.18 |
| Major amputation-free | |||
| Age | 0.95 | 0.89–1.02 | 0.16 |
| Male | 1.01 | 0.10–10.9 | 0.96 |
| Hypertension | 3.14 | 0.63–15.7 | 0.16 |
| Hyperlipidemia | 1.19 | 0.19–7.44 | 0.85 |
| DM | 1.67 | 0.25–11.1 | 0.60 |
| Smoking | 0.26 | 0.04–1.64 | 0.15 |
| Fontaine classification IV | 11.8 | 0.88–159 | 0.06 |
| (C) CLI caused by CDV | |||
| Overall survival | |||
| Age | 0.91 | 0.80–1.04 | 0.17 |
| Male | 1.70 | 0.04–66.7 | 0.78 |
| Hypertension | 6.27 | 0.54–72.6 | 0.14 |
| Hyperlipidemia | 0.59 | 0.03–11.6 | 0.73 |
| DM | 4.27 | 0.29–63.8 | 0.29 |
| Smoking | 1.45 | 0.14–15.4 | 0.76 |
| Major amputation-free | |||
| Age | 0.98 | 0.91–1.06 | 0.58 |
| Hypertension | 2.66 | 0.38–18.6 | 0.32 |
| Smoking | 0.58 | 0.06–5.82 | 0.64 |
| Fontaine classification IV | 0.61 | 0.05–6.87 | 0.69 |
Data are presented as HR and 95% CI. Abbreviations as in Tables 1,3.
Furthermore, we performed Kaplan-Meier analysis of the OS and MAF rates based on Fontaine classification in patients with ASO separately for age >70 years, DM, and HD (Figure 4A,B). In patients with ASO and Fontaine classification III, the OS and MAF rates were not significant with or without age >70 years (P=0.35 and P=0.42, respectively). In patients with ASO and Fontaine classification IV, age >70 years affected OS (P=0.02) but not limb salvage (P=0.32) (Figure 4A,B). The Kaplan-Meier analysis of the OS and MAF rates in patients with ASO based on Fontaine classification and presence of DM are shown in Figure 4C and Figure 4D. In patients with ASO and Fontaine classification III, the presence of DM affected OS (P=0.02) but not limb salvage (P=0.22). In patients with ASO and Fontaine classification IV, OS and MAF rates were not significant with or without DM (P=0.80 and P=0.67, respectively) (Figure 4C,D). Lastly, Kaplan-Meier analysis of the OS and MAF rates in patients with ASO based on Fontaine classification and presence of HD is shown in Figure 4E and Figure 4F. In patients with ASO and Fontaine classification III, the presence of HD affected OS (P=0.04) but not limb salvage (P=0.50). In patients with ASO and Fontaine classification IV, the OS and MAF rates were not significant with or without HD (P=0.61 and P=0.59, respectively) (Figure 4E,F).
Kaplan-Meier analysis of overall survival (OS) and major amputation-free in patients with arteriosclerosis obliterans (ASO) based on Fontaine classification and age >70 years, diabetes mellitus (DM), and hemodialysis (HD). (A) OS rate and (B) major amputation-free rate in patients with ASO based on Fontaine classification and age >70 years. (C) OS rate and (D) major amputation-free rate in patients with ASO based on Fontaine classification and presence of DM. (E) OS rate and (F) major amputation-free rate in patients with ASO based on Fontaine classification and presence of HD.
In patients with ASO and Fontaine classification III, the 1- and 5-year MAF rates were >90% and >80%, respectively, with or without age >70 years, DM, or HD. There was no marked decrease of even the chronic phase MAF rate after BM-MNC implantation. In patients with ASO and Fontaine classification IV, the 10-year OS rate was low even without age >70 years, DM, or HD. However, the 1- and 10-year MAF rates after the BM-MNC implantation were approximately 70% and 60%, respectively, with or without age >70 years, DM, or HD.
This study showed the long-term clinical outcomes and safety of BM-MNC implantation in patients with CLI in Japan. Our results indicated that BM-MNC implantation may be feasible and safe in patients with CLI, particularly if caused by TAO or CDV. In this study, patients with ASO had more risk factors that promoted arteriosclerosis, such as DM and HD, and were older than those with TAO or CDV.11,12 In addition, it has been reported that HD and DM correlated with major amputation among the control and BM-MNC implantation groups of patients with CLI caused by ASO or TAO.8 However, the present study showed that HD and DM did not correlate with major amputation in patients with CLI caused by ASO. In addition, in the TACT study, which evaluated the 3-year outcomes after BM-MNC implantation and was the closest to this study, HD and DM did not correlate with major amputation.7 Therefore, we speculate that HD and DM are not independent factors for major amputation in CLI patients who undergo BM-MNC implantation because Fontaine classification IV is a strong factor predicting major amputation.
The average 1-year limb amputation rate is 30% among patients with CLI who are treated using conventional therapy strategies, which include surgical options (e.g., bypass surgery or EVT).1 Moreover, the 1-year mortality risk reported in 2015 for patients with CLI was still high at 16–36%,2 which is similar to the 1-year mortality risk of 25% for patients with CLI reported by the Transatlantic Intersociety Consensus (TASC) II over 10 years ago.1 Revascularization therapy, including bypass surgery and EVT, is the primary way to improve CLI symptoms. However, approximately one-third of patients with CLI cannot undergo surgery for various reasons, including complications (e.g., cardiac/respiratory dysfunction), unfavorable general condition (e.g., severe dementia, being bedridden, extensive necrosis, or infection), and technical issues (e.g., absence of an artery that can be grafted).13 A major amputation rate of 24% and a minor amputation rate of 7% were reported for patients who did not undergo revascularization surgery. Significant differences in the 5-year limb salvage rate (83.5% vs. 55.8%) and the 5-year AFS rate (57.7% vs. 36.0%) were reported for patients who did or did not undergo revascularization.13 According to those reports, the OS, MAF, and AFS rates were still low in no-option CLI patients. BM-MNC implantation was developed almost 20 years ago, following the success of gene therapy in promoting angiogenesis. It advanced the development of therapeutic angiogenesis that is effective for patients without available options.5,6 Therapeutic angiogenesis aims to prevent limb amputation and to improve quality of life and survival in no-option CLI patients. The first report from the TACT study described significant improvement of symptoms in patients with CLI who underwent BM-MNC implantation.6 Of the 115 patients who participated in the TACT study, the 3-year MAF rates were 60% in patients with ASO and 91% in patients with TAO. Improved transcutaneous oxygen pressure (TcPO2) values tended to increase at 6 months after BM-MNC implantation and continued for 3 years in patients with ASO or TAO. The visual analog scale (VAS), ulcer size, and claudication distance were significantly improved 6 months after BM-MNC implantation. These improvements continued for 3 years in patients with ASO or TAO. The aforementioned findings suggest that BM-MNC implantation is safe and partially effective in patients with CLI, particularly in patients with TAO.7 The very long-term outcomes after cell therapy in patients with CLI have not been reported. Thus, we further evaluated the long-term outcomes after BM-MNC implantation in patients with CLI.
The present study showed the very long-term outcomes after BM-MNC implantation in patients with CLI caused by ASO, TAO, and CDV. In patients with TAO, the 10-year OS rate was >90%, and the 10-year MAF rate was >85%. In addition, in patients with CDV, the 10-year OS rate was 67.6%, and the 10-year MAF rate was >90%. In particular, the MAF rate in patients with CLI caused by TAO or CDV followed a very good course despite Fontaine classification. Ohta et al14 reported that 42.7% of 110 patients underwent either major or minor limb amputations, and major amputations were performed on 13 patients (11.8%) who had been treated conventionally (e.g., bypass surgery). Those investigators found that bypass surgery (n=46) results were unsatisfactory. The primary graft patency rates were 41% at 1 year, 32% at 5 years, and 30% at 10 years after the operation, and the secondary graft patency rates were 54% at 1 year, 47% at 5 years, and 39% at 10 years.14 The MAF were 91.4% at 1 year, 88.6% at 5 years, and 85.4% at 10 years after bypass surgery.14 The results from the current survey showed that among the 108 patients with TAO, the 5- and 10-year AFS rates were 87.9% and 87.9%, respectively, after BM-MNC implantation. These rates appeared to reflect those achieved using a conventional clinical strategy that includes bypass surgery. The target vessels connecting bypass are very thin in patients with CDV-related CLI (e.g., systemic scleroderma). Therefore, the graft patency and limb salvage rates are lower in these patients than in those with CLI caused by ASO.13,15 The 3-year primary graft patency rate was 38.9%, and the major amputation rate was 19% in patients with CDV-related CLI who underwent revascularization surgery. Moreover, the 3-year MAF was 67.2% in patients with CDV-related CLI, regardless of whether they underwent revascularization surgery.13 To the best of our knowledge, no results from large-scale, long-term epidemiological studies of patients with CDV-related CLI who have undergone the BM-MNC implantation have been reported. The results from the current study demonstrated that both the 5-year and 10-year AFS rates in patients with CDV-related CLI were 90.9% (95% CI: 79.2–96.1). According to these results, despite the patients’ backgrounds differing somewhat, the BM-MNC implantation outcomes were comparable to those of bypass surgery in patients with CLI caused by TAO or CDV. Additionally, in many cases, patients had a recurrence because of the low rate of patency bypass grafting in clinical practice. Therefore, we recommend BM-MNC implantation for no-option CLI patients with TAO or CDV. In particular, in patients with CLI caused by TAO or CDV, we speculate that BM-MNC implantation may be more beneficial than conventional treatment, including bypass surgery. To compare the efficacy of BM-MNC implantation with that of conventional treatment, we recommend conducting randomized controlled trials (RCTs) in the future.
The death and major amputation rates were significantly greater in patients with ASO than in patients with TAO or CDV. Death and major amputation occurred within 6 months of BM-MNC implantation, and the OS rate gradually worsened in the chronic phase. Recently, the OS and MAF in no-option CLI patients has improved with revascularization and wound care.16–19 Benoit et al showed that the 1-year MAF rate was 62–90%, and the 1-year AFS rate was 48–81% from 2006 to 2010 for no-option CLI patients.20 Furthermore, Miyahara et al showed that the 5-year MAF rate was 55.8%, and the 5-year AFS rate was 36.0% in patients with CLI who could not undergo revascularization.13 In the present study, the 1- and 5-year MAF and AFS rates in all patients with CLI caused by ASO were 78.7% and 74.0%, respectively. The 1- and 5-year AFS rates were 71.6% and 55.0%, respectively. In addition, in patients with CLI caused by ASO with Fontaine classification III, the 1- and 5-year MAF rates were >90% and >80%, respectively, regardless of age, DM, or HD in this study. It was difficult to directly compare the clinical outcomes of conventional treatment and BM-MNC implantation in patients with CLI caused by ASO because the clinical outcomes after conventional treatment were variable. However, clinical outcomes after BM-MNC implantation in this study were not inferior to those after conventional treatment in previous studies. Additionally, the TACT study revealed that ulcer size, VAS, and claudication distance were significantly improved after BM-MNC implantation.7 Thus, we suggest that BM-MNC implantation may effectively improve ischemic symptoms and quality of life in no-option CLI patients with ASO, particularly those with Fontaine classification III. Recently, it was reported that performing sufficient conventional treatment, including revascularization, in patients with CLI caused by ASO was very important.21 However, many patients with CLI resist conventional treatment and cannot undergo revascularization.13 Moreover, in many patients with CLI caused by ASO, above-the-knee vascular flow is also affected; in most cases, restoration of vascular flow to the level required for wound healing cannot be achieved merely with peripheral vessel regeneration. There are also concerns about the patency rate in revascularization procedures, such as EVT and bypass surgery, in patients with serious, refractory conditions. However, by maintaining the above-the-knee vascular flow, a bypass from the superficial femoral artery or the deep femoral artery can be developed, and vascularization can be further augmented with BM-MNC implantation. In CLI cases in which the required TcPO2 or skin perfusion pressure for wound healing cannot be obtained because of inadequate peripheral vascular flow, even for EVT or bypass surgery, a combination of endovascular revascularization and BM-MNC implantation must be considered. Therefore, we speculate that BM-MNC implantation may become a treatment option for no-option CLI patients with ASO and may be a combination treatment for patients who achieve insufficient efficacy with conventional treatment including endovascular revascularization.
During this study, BM-MNC implantation for patients with CLI was performed under advanced medical treatment. This study revealed that BM-MNC implantation may be feasible and safe in no-option CLI patients, particularly those with TAO and CDV. Because there are expected to be more no-option CLI patients in the future, standardization of BM-MNC implantation is important to improve prognosis and the limb salvage rate. In addition, it is important to accurately evaluate some factors predicting major amputation and death after BM-MNC implantation. Thus, if necessary, we would like to consider clinical studies, such as well-designed, multicenter, RCTs, to evaluate further efficacy and specific BM-derived cell therapeutic strategies.22,23
Study LimitationsWe used an eCRF to survey 345 patients with CLI who were intractable to any other treatment options such as medical therapy, rehabilitation, sympathetic ganglion block, and revascularization by EVT or surgery. These CLI patients were implanted with BM-MNCs intramuscularly. The first limitation of this study is that the field survey was a non-randomized, single-arm, open-labeled, and non-controlled retrospective study. Investigation of all cases has not yet been completed. The possibility of study bias must be considered. However, we believe that the results of this study are robust because, to our knowledge, this is the first report with a long follow-up period surveying a large number of patients with CLI who had undergone BM-MNC implantation therapy. Well-designed, multicenter, double-blind RCTs that involve a considerable number of patients are necessary to validate the effects of BM-MNC implantation. The second limitation is that several patients underwent cell implantation twice during the follow-up period. Repeated cell injections are reported to better improve ischemic symptoms. In this study, repeated cell implantation may have affected the long-term clinical outcomes, such as MAF and AFS, compared with a single cell implantation.24
Therapeutic angiogenesis using autologous BM-MNC implantation may be feasible and safe in no-option CLI patients, particularly in those with CLI caused by TAO and CDV. In the future, well-designed, multicenter RCTs that involve a considerable number of patients are necessary to validate the effects of BM-MNC implantation.
The authors appreciate the help and support of all committee members for this project and members of the institutes that participated in the TACT Follow-up Study (Appendix).
The authors declare no competing financial interests.
None.
Study Investigators
Department of Cardiology, Nagoya University Graduate School of Medicine; Kazuhisa Kondo, Ryo Hayashida, Satoshi Shintani, Rei Shibata, and Toyoaki Murohara, Department of Cardiovascular Medicine, Kyoto Prefectural University of Medicine; Kenji Yanishi and Satoaki Matoba, Center for Advanced Medicine and Clinical Research, Nagoya University Hospital, Nagoya; Kenta Murotani, Masahiko Ando, Masaaki Mizuno, and Tadami Fujiwara, Department of Cardiovascular Regeneration and Medicine, Research Institute for Radiation Biology and Medicine, Hiroshima University; Yukihito Higashi, First Department of Internal Medicine, Nara Medical University; Yoshihiko Saito, Department of Internal Medicine, Division of Cardiovascular Medicine, Kurume University School of Medicine; Yoshihiro Fukumoto, Department of Cardiovascular Medicine, Shinshu University School of Medicine; Uichi Ikeda, Department of Cardiology, Yokohama City University Hospital; Tomoaki Ishigami, Department of Internal Medicine and Clinical Immunology, Yokohama City University Graduate School of Medicine; Ryusuke Yoshimi, Metabolism, Endocrinology, Department of Premier Preventive Medicine, Osaka City University Graduate School of Medicine; Shinya Fukumoto, and the National Hospital Organization Kumamoto Medical Center; Kazuteru Fujimoto.
