The efficacy and safety of cyclosporine A plus androgen versus androgen alone for adult patients with non-severe aplastic anemia in China: a meta-analysis of randomized controlled trials

ABSTRACT

Background

Whether combined CsA with androgen therapy was superior to androgen therapy alone in NSAA remains controversial. This study aimed to assess the efficacy and safety of combined therapy versus androgen therapy for NSAA patients using a meta-analytic approach.

Methods

An electronic database of PubMed, EmBase, Cochrane library, CNKI, VIP, and Wanfang was systematically searched for randomized controlled trials (RCTs) from their inception to February 2020. The primary endpoint was effective rate, while the secondary endpoints included white blood cell (WBC), hemoglobin, platelet, and potential adverse events. The pooled results from included trials were calculated with the random-effects model.

Results

Forty-three RCTs recruited 2610 NSAA patients for the final quantitative meta-analysis. We noted that combined therapy was associated with an increased incidence of effective rate than androgen therapy alone (relative risk [RR]: 1.35; 95% confidence interval [CI]: 1.29–1.41; P < 0.001). Moreover, patients treated with combined therapy were associated with higher WBC (weighted mean difference [WMD]: 1.22; 95%CI: 0.94–1.49; P < 0.001), hemoglobin (WMD: 12.93; 95%CI: 8.86–17.01; P < 0.001), and platelet (WMD: 8.65; 95%CI: 7.05–10.24; P < 0.001). Finally, the pooled incidence of hirsutism, handshake, gingiva hyperplasia, liver function damage, and renal function damage were 0.35 (95%CI: 0.22–0.48), 0.24 (95%CI: 0.15–0.32), 0.22 (95%CI: 0.10–0.35), 0.19 (95%CI: 0.14–0.25), and 0.06 (95%CI: 0.01–0.11), respectively.

Conclusions

This study found that combined CsA with androgen therapy was superior to androgen therapy alone for Chinese patients with NSAA, and the most common adverse of combined therapy included hirsutism, handshake, gingiva hyperplasia, liver function damage, and renal function damage.

KEYWORDS:

• Cyclosporine A
• androgen
• non-severe aplastic anemia

Introduction

Aplastic anemia (AA) was characterized by hypocellular bone marrow and pancytopenia, which could be divided into non-severe AA (NSAA), severe AA, and very severe AA based on the degree of cytopenia. NSAA is an acquired condition with a diminished bone marrow activity, which could cause anemia, thrombocytopenia, and leukopenia. NSAA could be caused by a viral infection, medications, and other unknown causes [1–3]. Patients with NSAA may not receive any intervention, and the immunosuppressive therapy or bone marrow transplantation is just given for patients who progress to severe AA. There was little attention paid to the treatment of NSAA.

Cyclosporine A (CsA), as an immunoagent, plays an important effect on lymphocyte proliferation and has significant application value in the treatment of aplastic anemia. Moreover, androgens are commonly used in the treatment of AA in China, promoting the production of erythropoietin and increasing the hematopoietic function of the bone marrow. However, whether the combined CsA with androgen therapy was superior to androgen therapy alone in NSAA is limited and unclear. Therefore, the current meta-analysis was conducted to assess the efficacy and safety of combined therapy in the treatment of NSAA in Chinese people compared with androgen therapy alone.

Methods

Data sources, search strategy, and selection criteria

This review was performed following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Statement issued in 2009 [4]. The study designed as randomized controlled trials (RCTs) and compared the efficacy and safety of combined CsA with androgen therapy versus androgen therapy alone for Chinese patients with NSAA was appropriate. The electronic databases of PubMed, EmBase, Cochrane library, CNKI, VIP, and Wanfang were systematically searched from their inception up to February 2020 and using the following core terms: aplastic anemia [MeSH] OR aplastic anemia AND cyclosporine AND RCT.

The literature search and study selection were conducted following the standard flow by two reviewers independently, and any disagreement was resolved by group discussion. The inclusion criteria of this meta-analysis are listed as follows: (1) patients: adults patients from China and diagnosed with NSAA; (2) intervention: combined CsA with androgen therapy; (3) control: androgen therapy; (4) outcomes: effective rate, white blood cell (WBC), hemoglobin, platelet, and potential adverse events; and (5) study design: the study had to have an RCT design.

Data collection and quality assessment

Two reviewers independently conducted the data abstracted and quality assessment, and any inconsistency was settled by an additional author reviewing the original article. The abstracted information included the first authors’ surname, publication year, sample size, mean age, percentage of males, intervention, control, treatment duration, and reported outcomes. The quality of included studies was assessed by the Jadad scale, which was based on randomization, concealment of the treatment allocation, blinding, completeness of follow-up, and the use of intention-to-treat analysis [5].

Statistical analysis

The effective rate between combined therapy and androgen therapy was assigned as categories data, while the levels of WBC, hemoglobin, and platelet were assigned as continuous data. The relative risk (RR) and weighted mean difference (WMD) with a 95% confidence interval (CI) were applied to assess the pooled effect estimates for categories and continuous data, respectively. Moreover, the incidence of adverse events in patients treated with combined therapy was abstracted as the number of events and sample size. The pooled analyses in this study were conducted by the random-effects model [6,7]. Heterogeneity, among included trials, was assessed using I2 or Q statistic, and I2 > 50.0% or P < 0.10 was considered significant heterogeneity [8,9]. The robustness of pooled conclusions was assessed using sensitivity analysis [10]. Subgroup analysis was conducted for effective rate based on publication year, mean age, percentage of males, control, and study quality, and the difference between subgroups was assessed using the interaction P test [11]. Publication biases for effective rate, WBC, hemoglobin, and platelet were assessed by the funnel plot, Egger, and Begg test results [12,13]. The inspection level for all pooled results is 2-sided, and P < 0.05 was considered a significant difference. STATA software (version 10.0; Stata Corporation, College Station, TX, USA) was applied to analyze all data analyses in our meta-analysis.

Results

Literature search

The initial electronic searches yielded 2137 records, and 781 were excluded because of duplicate titles. Additional 1246 articles were excluded due to the study reported irrelevant titles. A total of 110 studies were retrieved for further assessment, and 67 studies were excluded due to no appropriate control (n = 29); did not being treated with combined therapy (n = 27), and not RCT (n = 11). No new eligible study was found by reviewing the reference of retrieved studies. Finally, 43 studies involving 2610 NSAA patients from China were selected for this meta-analysis (Figure 1) [14–56].

Figure 1. The details of the literature search and study selection process.

Study characteristics

The general characteristics of studies and patients are summarized in Table 1. The included studies published from 2001to 2020, and 30–120 patients were included in each included trial. The mean age of included patients ranged from 30.0 to 47.5 years, and the percentage of males ranged from 41.7 to 74.2 percent. A total of 36 studies compared combined therapy with stanozolol, while the remaining 7 studies compared combined therapy with other therapies, including testosterone undecanoate, and testosterone propionate, or did not define the type of androgen therapy.

Table 1. The baseline characteristics of included studies.

Study	Sample size	Age (years)	Percentage male (%)	CsA	Androgen	Duration of treatment	Study quality
Yu 2001 [14]	57 (31/26)	33.1	56.1	CsA 5–8 mg/kg/day, bid/tid	Stanozolol 2 mg, tid	6.0 months	2
Xu 2004 [15]	76 (42/34)	32.6	67.1	CsA 0.5–5 mg/kg/day, bid	Stanozolol 6 mg/day or TU 500 mg, twice/month	6.0 months	1
Dong 2006 [16]	80 (44/36)	32.6	65.0	CsA 0.5–5 mg/kg/day, bid	Stanozolol 6–8 mg/day or TU 250–500 mg/15 days	6.0 months	2
Zhao 2007 [17]	48 (28/20)	36.8	56.3	CsA 3–5 mg/kg/day	TU 120 mg/day, tid	NA	2
Tang 2007 [18]	31 (15/16)	32.0	54.8	CsA 3–5 mg/kg/day	Stanozolol 6–8 mg/day	6.0 months	2
Yang 2008 [19]	65 (35/30)	37.0	55.4	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	2
Shu 2008 [20]	34 (18/16)	30.0	52.9	CsA 3–6 mg/kg/day, bid/tid	Stanozolol 2 mg, tid	3.0–9.0 months	2
Zhu 2009 [21]	60 (30/30)	36.2	58.3	CsA 3–5 mg/kg/day, bid	TP 50 mg, 3 times/week	NA	2
Wang 2009 [22]	50 (26/24)	32.0	56.0	CsA 3–6 mg/kg/day, bid/tid	Stanozolol 2 mg, tid	3.0–9.0 months	1
Xia 2009 [23]	52 (29/23)	36.5	57.7	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	2
Zhou 2009 [24]	38 (24/14)	36.0	60.5	CsA 3–5 mg/kg/day, bid	TU 40 mg/day, tid	6.0 months	1
Li 2010 [25]	42 (22/20)	37.0	64.3	CsA 3–5 mg/kg/day, bid	TU 40 mg/day, tid	6.0 months	2
Dai 2010 [26]	56 (28/28)	33.0	53.6	CsA 3–6 mg/kg/day, bid/tid	Stanozolol 2 mg, tid	3.0–9.0 months	1
Zeng 2010 [27]	68 (34/34)	34.5	55.9	CsA 3–5 mg/kg/day, bid	Stanozolol 6–12 mg, tid	6.0 months	2
Zhao 2010 [28]	30 (16/14)	32.0	60.0	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	3.0–10.0 months	1
Yang 2011 [29]	62 (31/31)	34.1	56.5	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	2
Feng 2011 [30]	104 (52/52)	14.0-65.0	62.5	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	1
Yu 2011 [31]	52 (26/26)	34.8	55.8	CsA 3–6 mg/kg/day, bid/tid	Stanozolol 2 mg, tid	3.0–9.0 months	2
Liang 2012 [32]	42 (24/18)	31.2	61.9	CsA 4 mg/kg/day, bid	Stanozolol 6 mg/day	6.0 months	1
Huang 2012 [33]	60 (30/30)	13.0-60.0	41.7	CsA 5–12 mg/kg/day	Stanozolol 2 mg, tid	6.0 months	2
Tang 2012 [34]	50 (25/25)	47.5	54.0	CsA 12 mg/kg/day	Stanozolol 2 mg, tid	6.0 months	1
Liu 2013 [35]	37 (20/17)	33.0	54.1	CsA 4–6 mg/kg/day, bid	Stanozolol 2 mg, tid	NA	1
Yin 2013 [36]	40 (20/20)	NA	57.5	CsA 5–12 mg/kg/day	TU 40 mg/day, tid	6.0 months	1
Wei 2013 [37]	60 (30/30)	35.2	56.7	CsA 4 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	2
Yinlidaer 2013 [38]	48 (24/24)	37.5	58.3	CsA 3–4 mg/kg/day, bid	Androgen 40 mg, tid	NA	1
Hu 2013 [39]	88 (44/44)	45.3	61.4	CsA 4 mg/kg/day, bid	Stanozolol 6 mg/day	NA	1
Sun 2014 [40]	73 (39/34)	34.3	53.4	CsA 3 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	1
Liu 2014 [41]	52 (26/26)	35.3	53.8	CsA 3–5 mg/kg/day, bid	Stanozolol 6–12 mg/day, tid	NA	2
Zhuang 2015 [42]	90 (45/45)	44.7	NA	CsA 6 mg/kg/day, bid	Stanozolol 6 mg/day	NA	2
Du 2015 [43]	98 (49/49)	31.0	56.1	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	NA	2
Xiong 2015 [44]	70 (35/35)	34.0	51.4	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	1
Liang 2016 [45]	120 (60/60)	33.7	60.0	CsA 3–5 mg/kg/day, bid	TU 2–4 mg, tid	6.0 months	2
Pang 2016 [46]	50 (25/25)	31.4	56.0	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	2
Zhou 2016 [47]	38 (19/19)	NA	50.0	CsA 4 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	2
Tian 2017 [48]	76 (38/38)	43.4	53.9	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	1
Chen 2017 [49]	62 (31/31)	44.0	74.2	CsA 3–5 mg/kg/day, bid	Stanozolol 6–12 mg/day, tid	NA	1
Yang 2017 [50]	44 (22/22)	40.5	59.1	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	2
Ru 2018 [51]	68 (34/34)	32.4	70.6	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	2
Yang 2018 [52]	80 (40/40)	34.3	61.3	CsA 4 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	2
Hu 2019 [53]	64 (32/32)	36.2	48.4	CsA 4 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	1
Sun 2019 [54]	50 (25/25)	38.5	65.0	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	1
Zhang 2019 [55]	84 (42/42)	39.9	59.5	CsA 4 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	2
Yao 2020 [56]	61 (30/31)	46.8	55.7	CsA 3–5 mg/kg/day, bid	Stanozolol 2 mg, tid	6.0 months	1

Note: CsA: cyclosporine A; TU: Testosterone Undecanoate; TP: testosterone propionate.

Primary endpoint

For the effect of combined therapy versus androgen therapy on the incidence of effective rate, data were reported in 42 trials. The pooled RR indicated combined therapy was associated with an increased incidence of effective rate (RR: 1.35; 95%CI: 1.29–1.41; P < 0.001; Figure 2), and no evidence of heterogeneity was seen across included trials (I2 = 0.0%; P = 1.000). The pooled conclusion was stable and not altered by sequential excluding individual trials (Supplemental 1). This significant difference remained in all subgroups (Table 2).

Figure 2. Effect of combined therapy versus androgen therapy on the incidence of effective rate.

Table 2. Subgroup analysis for the effective rate.

Variables	Subgroup	Number of studies	RR and 95%CI	P-value	Heterogeneity (%)	P-value for Q statistic	P-value between subgroups
Publication year	Before 2010	11	1.47 (1.31–1.66)	<0.001	0.0	0.998	0.094
	2010 or after	31	1.33 (1.26–1.40)	<0.001	0.0	0.999
Mean age (years)	≥35.0	19	1.31 (1.23–1.40)	<0.001	0.0	0.890	0.513
	<35.0	19	1.40 (1.29–1.51)	<0.001	0.0	1.000
Percentage male (%)	≥60.0	12	1.35 (1.24–1.47)	<0.001	0.0	0.956	0.155
	<60.0	29	1.37 (1.29–1.46)	<0.001	0.0	1.000
Control	Stanozolol	35	1.34 (1.27–1.40)	<0.001	0.0	1.000	0.266
	Other	7	1.44 (1.26–1.65)	<0.001	0.0	0.941
Study quality	2	23	1.35 (1.27–1.44)	<0.001	0.0	0.997	0.923
	1	19	1.34 (1.25–1.44)	<0.001	0.0	0.980

White blood cell

Data for the effect of combined therapy versus androgen therapy on WBC were reported in 10 trials. We noted that combined therapy was associated with high WBC levels compared with androgen therapy alone (WMD: 1.22; 95%CI: 0.94–1.49; P < 0.001; Figure 3), and significant heterogeneity was detected among included trials (I2 = 85.4%; P < 0.001). Sensitivity analysis indicated the pooled conclusion was not changed by sequential excluding individual trials (Supplemental 1).

Figure 3. Effect of combined therapy versus androgen therapy on the level of the WBC.

Hemoglobin

Data for the effect of combined therapy versus androgen therapy on hemoglobin were reported in 12 trials. The pooled result found patients treated with combined therapy were associated with a high level of hemoglobin (WMD: 12.93; 95%CI: 8.86–17.01; P < 0.001; Figure 4), and significant heterogeneity was seen across included trials (I2 = 79.8%; P < 0.001). This conclusion was robust and not altered by sequential excluding included trials (Supplemental 1).

Figure 4. Effect of combined therapy versus androgen therapy on the level of hemoglobin.

Platelet

Data for the effect of combined therapy versus androgen therapy on platelet were reported in 14 trials. We noted that combined therapy was associated with high platelet levels compared with androgen therapy alone (WMD: 8.65; 95%CI: 7.05–10.24; P < 0.001; Figure 5), and significant heterogeneity was found among included trials (I2 = 51.1%; P = 0.014). The conclusion was stable and not altered by excluding any particular trial (Supplemental 1).

Figure 5. Effect of combined therapy versus androgen therapy on the level of the platelet.

Adverse events

The distribution of adverse events of hirsutism, handshake, gingiva hyperplasia, liver function damage, and renal function damage in patients treated with combined therapy is listed in Table 3. The pooled incidence of hirsutism, handshake, gingiva hyperplasia, liver function damage, and renal function damage for patients treated with combined therapy was 0.35 (95%CI: 0.22–0.48), 0.24 (95%CI: 0.15–0.32), 0.22 (95%CI: 0.10–0.35), 0.19 (95%CI: 0.14–0.25), and 0.06 (95%CI: 0.01–0.11), respectively.

Table 3. The results of adverse events in patients treated with combined therapy.

Study	Sample size	Hirsutism	Handshake	Gingiva hyperplasia	Liver function damage	Renal function damage
Yu 2001	31	11	9	3	5	1
Xu 2004	42	38	24	30	20	–
Dong 2006	44	39	25	30	20	–
Zhao 2007	28	25	–	27	–	8
Tang 2007	15	3	4	2	1	–
Yang 2008	35	16	19	17	14	–
Shu 2008	18	12	2	2	4	–
Wang 2009	26	14	1	3	4	–
Xia 2009	29	–	1	2	–	–
Zhou 2009	24	3	–	2	5	1
Dai 2010	28	8	9	5	10	–
Zeng 2010	34	8	12	5	10	–
Zhao 2010	16	8	1	2	3	–
Yang 2011	31	–	2	2	–	–
Feng 2011	52	12	18	8	15	–
Yu 2011	26	12	13	10	11	–
Liang 2012	24	19	–	8	3	–
Liu 2013	20	10	2	2	4	–
Yin 2013	20	1	–	–	–	–
Wei 2013	30	6	–	–	2	–
Yinlidaer 2013	24	3	–	–	5	–
Xiong 2015	35	6	–	–	3	1
Pang 2016	25	2	1	1	–	–
Yang 2017	22	–	–	1	1	–
Yang 2018	40	2	–	–	3	–
Hu 2019	32	–	2	1	1	–
Sun 2019	25	4	6	2	5	–
Zhang 2019	42	1	–	–	2	–
Yao 2020	30	1	–	1	6	2
Pooled	848	264	151	166	157	13
Incidence and 95%CI	–	0.35 (0.22–0.48)	0.24 (0.15–0.32)	0.22 (0.10–0.35)	0.19 (0.14–0.25)	0.06 (0.01–0.11)

Publication bias

Publication biases for effective rate, WBC, hemoglobin, and platelet were assessed and displayed in Supplemental 2. Although no significant publication biases for WBC (PEgger: 0.204; PBegg: 0.721), hemoglobin (PEgger: 0.154; PBegg: 0.837), and platelet (PEgger: 0.074; PBegg: 0.101) were detected, we noted significant publication bias for effective rate (PEgger: < 0.001; PBegg: < 0.001). However, the conclusion was not altered by adjusting the trim and fill method [57].

Discussion

This meta-analysis based on published RCTs assesses the efficacy and safety of combined therapy versus androgen therapy for Chinese patients with NSAA. This is the first meta-analysis focused on this topic and given the accurate effect estimates regarding the outcomes of effective rate, WBC, hemoglobin, platelet, and potential adverse events. This study found that combined therapy versus androgen therapy could significantly improve effective rate, WBC, hemoglobin, and platelet for Chinese patients with NSAA. Moreover, the most common adverse events in patients treated with combined therapy were hirsutism, handshake, gingiva hyperplasia, liver function damage, and renal function damage. Finally, the treatment effect of combined therapy versus androgen therapy on effective rate does not differ when stratified by publication year, mean age, percentage of males, control, and study quality.

The causes of NSAA are complex and varied, and are closely related to the injury of hematopoietic stem cells, immune-mediated response, and abnormal blood microcirculation function. T cell immune-mediated dysfunction could exert an inhibitory effect on the function of hematopoietic stem cells and cause the normal hematopoietic function of the body, which is associated with various anemia symptoms [58]. Although traditional treatment strategies could improve the hematological parameters, they are prone to adverse reactions such as abnormal hair, skin rash, abnormal liver function, and gingival hyperplasia, which seriously affect the improvement effect of anemia symptoms.

The pooled results of this study found the effective rate, WBC, hemoglobin, and platelet could significantly be improved in NSAA patients treated with combined therapy. The potential reason for this could be that CsA as an immunosuppressive agent could significantly relieve the immune response by inhibiting the of negative factors (interleukin-interferon and tumor necrosis factor) by the effective regulation of the proportion of T cell subsets and reducing the influence of negative factors on hematopoietic function [59,60]. Moreover, androgen therapy could significantly increase the erythropoietin, enhance the differentiation and proliferation of stem cells, and regulate immune function [61]. Therefore, the combination of hematopoietic and immunosuppressants could enhance hematopoietic function.

Subgroup analyses found a significant difference between combined therapy and androgen therapy in effective rate in all subsets. However, we noted the treatment effect is more evident when trials published before 2010, mean age <35.0 years, a percentage of males <60.0%, used other androgen agents as background therapies, or studies reported the type of randomization with relatively high quality. These results could be explained by the treatment effect of combined therapy that was more suitable for younger male patients, and used testosterone undecanoate and testosterone propionate or did not define the type of androgen therapy as background therapies. Moreover, the physical examination was more standardized, and NSAA patients could help with early diagnosis.

The most common adverse events for NSAA patients treated with combined therapy were hirsutism, handshake, gingiva hyperplasia, liver function damage, and renal function damage. The timely treatment strategies were given for patients with adverse events which could be effectively alleviated. Therefore, we should pay more attention to the physical health and mental health of our NSAA patients during the treatment process. Moreover, communications with patients should enhance their attitude and confidence when suffering from the disease and result in the recovery of the disease. Finally, the family members of patients should be involved in the treatment process, which could play a supervisory role in the drugs and quantity to ensure the treatment effect and improve the prognosis of NSAA.

Several strengths and limitations in this meta-analysis should be highlighted: (1) the analysis of this study was based on RCTs, and the evidence level was relatively high; (2) the pooled results were based on a large sample size, and the results of this study are more robust than any individual trial; (3) subgroup analysis for effective rate was conducted to assess the treatment effect of combined therapy versus androgen therapy in specific subpopulations; (4) substantial heterogeneity for WBC, hemoglobin, and platelet were detected; (5) the study quality of included trials was low, and the blindness, concealment of the treatment allocation, and the use of intention-to-treat analysis were not reported in nearly all included trials; and (6) the analysis based on published articles, and publication bias was an inevitable problem.

Conclusion

This study found that combined therapy was significantly associated with the improvement in effective rate, WBC, hemoglobin, and platelet than androgen therapy alone for Chinese patients with NSAA. Moreover, the most common adverse events in patients treated with combined therapy were hirsutism, handshake, gingiva hyperplasia, liver function damage, and renal function damage. Further large-scale RCT should be conducted to assess other effective treatment strategies for patients with NSAA.

Ethics approval and consent to participate

The need for ethics approval by an institutional board review was waived as this study does not directly involve human subjects.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The data sets used and/or analyzed during the present study are available from the corresponding author on reasonable request.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [ grant number 82070127].