Impact on acute myeloid leukemia relapse in granulocyte colony-stimulating factor application: a meta-analysis

ABSTRACT

Objectives: This meta-analysis evaluated the impact of granulocyte colony-stimulating factor (G-CSF) added to chemotherapy on treatment outcomes including survival and disease recurrence in patients with acute myeloid leukemia (AML).

Methods: Medline, Cochrane, EMBASE, and Google Scholar databases were searched until 19 September 2016 using search terms. Studies that investigated patients with AML who underwent stem-cell transplantation were included.

Results: The overall analysis revealed a significant improvement in overall survival (OS) (P = .019) and disease-free survival (DFS) (P = .002) for patients receiving G-CSF with chemotherapy. Among patients without prior AML treatment, there was a significant improvement in DFS (P = .014) and reduction in incidence of relapse (P = .015) for those who received G-CSF. However, subgroup analyses found no significant difference between G-CSF (+) and G-CSF (−) treatments in rates of OS (P = .104) and complete remission (CR) (P = .572) for patients without prior AML treatment. Among patients with relapsed/refractory AML, there was no significant difference found between G-CSF (+) and G-CSF (−) groups for OS (P = .225), DFS (P = .209), and CR (P = .208).

Discussion: Treatment with chemotherapy plus G-CSF appears to provide better survival and treatment responses compared with chemotherapy alone, particularly for patients with previously untreated AML.

Abbreviations: AML, acute myeloid leukemia; CI, confidence interval; CR, complete remission; DFS, disease-free survival; G-CSF, granulocyte colony-stimulating factor; GM-CSF, granulocyte macrophage colony-stimulating factor; HR, hazard ratio; MDS, myelodysplastic syndrome; OR, odds ratio; OS, overall survival; RCTs, randomized control trials; RR, relative risk

KEYWORDS:

• Acute myeloid leukemia
• disease-free survival
• granulocyte colony-stimulating factor (G-CSF)
• overall survival
• relapse
• survival
• remission

Introduction

The majority of younger adult patients with acute myeloid leukemia (AML) are treated with curative intent. However, relapse is a significant treatment issue due to residual leukemic cells escaping the cytotoxic effects of the chemotherapy [1]; consequently, AML treatment is still associated with considerable morbidity and mortality [2].

Several studies have evaluated the use of the myeloid growth factors, granulocyte colony-stimulating factor (G-CSF), or granulocyte macrophage colony-stimulating factor (GM-CSF), used in conjunction with chemotherapy with the idea that these molecules may sensitize the leukemic cells to the cytotoxic effects of the chemotherapy, and hence reduce the rate of relapse [3–5]. In support of this, studies in AML tissue culture cells have shown that priming with G-CSF or GM-CSF may alter the cell cycle of AML cells, making them more susceptible to the cytotoxicity of chemotherapy [6,7].

In addition, colony-stimulating factors (CSFs) can reduce complications resulting from chemotherapy suppression of bone marrow. Previous meta-analyses have found that the addition of CSFs to treatment in adults and children cancer patients reduced the risk of chemotherapy-associated complications and thus the need to reduce or delay dosing of chemotherapy [8–11]. CSFs are commonly used to accelerate neutrophil recovery following stem-cell transplantation to reduce the risk of infectious complications and the subsequent need of supportive care [12].

However, in AML, disease recurrence can result from leukemic clones surviving high-dose chemo- and/or radiotherapy or contaminating autologous stem cells used for transplantation [13,14]. Hence, the use of G-CSF in AML is controversial as surviving leukemic blasts may express G-CSF receptors and the interaction of the receptor with the cytokine may increase the risk of relapse and reduce disease-free survival (DSF) [15–17]. However, findings from prior studies evaluating the use of G-CSF as part of the treatment regimen for patients with AML receiving stem-cell transplantation have been inconsistent, and hence the effect of G-CSF on disease-related outcomes and survival in AML is not clear. This meta-analysis evaluated the impact of G-CSF on outcomes including survival and disease recurrence in patients with AML undergoing stem-cell transplantation.

Methods

Search strategy

This study was performed in accordance with PRISMA. Medline, Cochrane, EMBASE, and Google Scholar databases were searched until 19 September 2016 using the following search terms: acute myeloid leukemia or AML, relapse, granulocyte colony-stimulating factors or G-CSF. Randomized control trials (RCTs), two-arm, and prospective studies were included. All studies had patients with AML who planned to undergo stem-cell transplantation, and had at least two treatment arms: one in which G-CSF was administered as part of the treatment regimen (either as priming or given prior to or at transplantation) and another in which G-CSF was not given. Studies not published in English, letters, comments, editorials, case reports, proceedings, and personal communications were excluded. All potential studies were searched by two independent reviewers, with a third reviewer acting as consultant to resolve any uncertainty regarding eligibility.

Data extraction

The following information/data were extracted from studies that met the inclusion criteria: the name of the first author, year of publication, study design, number of participants in each treatment group, patients’ age and gender, type of AML (relapsed or primary/secondary AML with no prior treatment), and outcomes of interest.

Quality assessment

Quality assessment was performed using the Cochrane Risk of Bias Tool (Cochrane Handbook for Systematic Reviews of Interventions) [18].

Outcome measures

The primary outcomes were overall survival (OS) and DFS between AML patients receiving G-CSF plus chemotherapy (G-CSF (+)) compared with patients receiving chemotherapy alone (G-CSF (−)). The secondary outcomes were complete remission and incidence of relapse between two groups.

Statistical analysis

Hazard ratios (HRs) with 95% confidence interval (CI) for survival outcomes were extracted for each individual study. If the available data were presented from the Kaplan–Meier curve, we extracted their survival rates at some specified times in order to reconstruct the HR estimate and its variance, with the assumption that the rate of patients censored was constant during the study follow-up [19]. Odds ratios (ORs) with 95% CIs were calculated for the dichotomous outcomes in the G-CSF (+) group compared with the G-CSF (−) group. For OS and DFS, an HR/OR <1 indicated that the patients treated with G-CSF plus chemotherapy were favored. For CR, an OR <1 indicated that the patients treated with G-CSF alone were favored. A χ²-based test of homogeneity was performed and the inconsistency index (I²) and Q statistics were determined. If the I² statistics were >50%, a random-effects model was used. Otherwise, fixed-effect models were employed. Pooled effects were calculated and a 2-sided P value <0.05 was considered to indicate statistical significance. A prospectively planned subgroup analysis was performed according to AML status (untreated AML or relapsed/refractory AML). Sensitivity analysis was carried out using the leave-one-out approach. Publication bias analysis was not performed because the number of studies was too few (<10) to detect an asymmetric funnel [20]. All analyses were performed using Comprehensive Meta-Analysis statistical software, version 2.0 (Biostat, Englewood, NJ, USA).

Results

Search results

The initial database search identified 233 studies that were screened (Figure 1). Of these, 182 were removed for not being relevant. Fifty-one studies were fully reviewed and 40 were excluded for being one-arm studies, not investigating patients with AML, not reporting outcomes of interest, not administering G-CSF prior to transplantation, being a letter, being an in vitro study, and not all patients underwent stem-cell transplantation.

Figure 1. Flow chart of study selection.

Eleven studies were included in the analysis with a total of 5076 patients (range 18–1029 patients): 2466 in the G-CSF (+) group and 2610 in the non-G-CSF (−) group (Table 1) [21–31]. The age of patients ranged from 10 to 60 years and the percentage of males and females was similar between treatment groups. Of the 11 included studies, 7 focused on patients with untreated AML (primary or secondary) and 4 were on patients with relapsed or refractory AML. The type of chemotherapy used and the dosage of G-CSF varied across studies (Supplemental Table 1).

Table 1. The characteristics of the studies included in this study.

First author (year)	Study design	AML status	Interventions	Age (yrs)	Male (%)	No of patients
Krug (2016)	RCT	Untreated primary or secondary	G-CSF (+)	60*	56	1029
			G-CSF (−)	60*	54	1023
Gao (2014)	RCT	Relapsed/refractory	G-CSF (+)	18–40:60%	45	89
			G-CSF (−)	8–40:63%	51	89
Pabst (2012)	RCT	Untreated primary or secondary	G-CSF (+)	49*	52	456
			G-CSF (−)	49*	52	461
Beksac (2010)	RCT	Untreated primary	G-CSF (+)	39	60	123
			G-CSF (−)	38	54	137
Milligan (2006)	RCT	Untreated primary or secondary	G-CSF (+)	NA	54	178
			G-CSF (−)		49	178
Lowenberg (2003)	RCT	Untreated primary or secondary	G-CSF (+)	44*	51	321
			G-CSF (−)	45*	51	319
Kern (1998)	P	Relapsed/refractory	G-CSF (+)	48*	54	68
			G-CSF (-−)	50*	54	91
Thomas (1999)	RCT	Relapsed/refractory	G-CSF (+)	47*	52	95
			G-CSF (−)	46*	70	97
Laver (1997)	P	Relapsed/refractory	G-CSF (+)	10.1	56	10
			G-CSF (−)			8
Heil (1995)	RCT	Untreated primary or secondary	G-CSF (+)	50*	NA	41
			G-CSF (−)	50*		39
Estey (1992)	P	Untreated primary or secondary	G-CSF (+)	48*	NA	56
			G-CSF (−)	50*		66
			G-CSF (−)	47^		110

Note: P, prospective study; G-CSF, granulocyte colony-stimulating factor; RCT, randomized controlled trials; and NA: not available.

* Median (Range).

^ Meduan (IQR).

The HR for OS ranged from 0.513 to 1.1 and for DFS from 0.63 to 0.94 in the G-CSF (+) group relative to the G-CSF (−) group across studies (Table 2). The incidence of relapse ranged from 29.8% to 59% in the G-CSF (+) group and from 36.6% to 60.7% in the G-CSF (−) group.

Table 2. Summarized reported outcomes of the included studies.

First author (year)	No of patients	Interventions	CR rate	OS (in HR with 95% CI)	DFS (in HR with 95% CI)	Incidences of relapse
Krug (2016)	1029	G-CSF (+)	59%	0.96 (0.87, 1.06)	0.94 (0.86, 1.03)	na
	1023	G-CSF (−)	59%	1.00	1.00
Gao (2014)	89	G-CSF (+)	na	0.57 (0.4, 0.8)	0.63 (0.46, 0.88)	38.2%
	89	G-CSF (−)		1.00	1.00	60.7%
Pabst (2012)	456	G-CSF (+)	81%	0.99 (0.83–1.17)	0.92 (0.78-1.08)	34%
	461	G-CSF (−)	77%	1.00	1.00	36%
Beksac (2010)	123	G-CSF (+)	62.5%	0.67 (0.48, 1)	na	29.8%
	137	G-CSF (−)	64.6%	1.00		36.6%
Milligan (2006)	178	G-CSF (+)	58%	1.03 (0.81–1.31)	0.89 (0.65-1.22)	58%
	178	G-CSF (−)	61%	1.00	1.00	66%
Lowenberg (2003)	321	G-CSF (+)	79%	0.87 (0.72–1.06)	0.77 (0.62–0.96)	37%
	319	G-CSF (−)	83%	1.00	1.00	44%
Kern (1998)	68	G-CSF (+)	56%	na	na	na
	91	G-CSF (−)	47%
Thomas (1999)	95	G-CSF (+)	65%	na	0.92 (0.75, 1.12)	na
	97	G-CSF (−)	59%		1.00
Laver (1997)	10	G-CSF (+)	40%	na	na	na
	8	G-CSF (−)	50%
Heil (1995)	41	G-CSF (+)	81%	1.1 (0.58, 2.09)	na	na
	39	G-CSF (−)	79%	1.00
Estey (1992)	56	G-CSF (+)	48%	0.513	na	na
	66	G-CSF (−)	74%	1.00
	110	G-CSF (−)	65%

Note: G-CSF, granulocyte colony-stimulating factor; na, not available/not reported; HR, hazard ratios; and CI, confidence interval.

Meta-analysis

Overall survival

Three studies, Kern et al. [27], Laver et al. [29], and Estey et al. [31], did not provide enough information to calculate HRs for OS and were excluded from the meta-analysis. Heterogeneity of the data was not significant among the seven included studies, and therefore a fixed-effect model was used (Q statistic = 11.473, I² = 38.99%). The overall analysis revealed a significant survival advantage in the G-CSF (+) group compared with the G-CSF (−) (pooled HR = 0.93; 95%CI, 0.87–0.99; P = .019) (Figure 2(A)). However, subgroup analysis found no significant difference in OS between patients without prior AML treatment (pooled HR = 0.94; 95% CI, 0.88–1.01; P = .104), or in those with relapsed/refractory AML (pooled HR = 0.74; 95% CI, 0.46–1.20; P = .225).

Figure 2. Meta-analysis of (A) overall survival, (B) disease-free survival, (C) complete remission rate, and (D) incidence of relapse.

Disease-free survival

Six studies quantitatively reported DFS and were included in the meta-analysis. A fixed-effect model was used because no heterogeneity was seen among the studies (Q statistic = 7.671, I² = 34.82%) [21–23,25,26,28]. The overall analysis revealed there was a benefit of DFS in favor of patients treated with G-CSF plus chemotherapy (pooled HR = 0.90; 95% CI, 0.84–0.96; P = .002). In the untreated AML subgroup, patients treated with G-CSF (+) were associated with significant improvement in DFS compared with those who did not receive G-CSF (−) [pooled HR = 0.91; 95% CI, 0.85–0.98; P = .014]. No significant difference between G-CSF (+) and G-CSF (−) groups was observed in the patients with relapsed or refractory AML (pooled HR = 0.78; 95% CI, 0.53–1.15; P = .209) (Figure 2(B)).

Complete remission

One study, Gao et al. [22], did not provide data for CR and was omitted from the analysis. No heterogeneity was observed among the nine included studies; therefore, a fixed-effect model was used (Q statistic = 13.566, I² = 33.66%). The overall analysis revealed there was no difference in CR between G-CSF (+) and G-CSF (−) groups (pooled OR = 0.99; 95%CI, 0.87–1.12; P = .870) (Figure 2(C)). Subgroup analysis found that the addition of G-CSF to chemotherapy compared to chemotherapy alone did not alter the rate of CR in both untreated AML (pooled OR = 0.96; 95% CI, 0.85–1.10, P = .572) and relapsed/refractory AML subgroups (pooled OR = 1.31; 95% CI, 0.86–1.99, P = .208).

Incidence of relapse

For evaluation of the incidence of relapse, we only assessed the subgroup of previously untreated AML patients due to lack of reported data in the studies. Four studies, Pabst et al. [23], Beksac et al. [24], Milligan et al. [25], and Lowenberg et al. [26], provided the frequency of relapse rates for the two intervention groups’ subjects without prior AML treatment. No heterogeneity was observed among these four studies, so a fixed-effect model was used (Q statistic = 1.512, I² = 0%). The overall analysis revealed patients receiving G-CSF plus chemotherapy had a significantly lower incidence of relapse than those receiving chemotherapy alone (pooled OR = 0.81; 95%CI, 0.68–0.96; P = .015) (Figure 2(D)).

Sensitivity analysis

Sensitivity analyses were performed using the leave-one-out approach for each outcome (Table 3). For OS, when we removed each study in turn, the pooled estimates for OS generally remained in the same direction (all pooled HR <1), except the removal of two studies, Gao et al. [22] and Beksac et al. [24], caused the findings to become non-significant (P = .076 and .062, respectively). For DFS and CR, the direction of the pooled estimates did not vary markedly with the removal of each study individually, indicating the findings are robust. The pooled estimates for relapse rate remained <1 after each study was removed in turn; though removal of Lowenberg et al. [26] caused the difference between treatments to become non-significant. Therefore, sensitivity analysis found that individual studies may have overly influenced the findings for OS, and relapse rate, but not for DFS and CR.

Table 3. Results of sensitivity analysis.

First author (year)	Statistics with study removed
	Points	Lower limit	Upper limit	Z-value	P value
OS
Krug (2016)	0.90	0.83	0.98	−2.40	.017
Gao (2014)	0.94	0.88	1.01	−1.77	.076
Pabst (2012)	0.92	0.85	0.98	−2.48	.013
Beksac (2010)	0.94	0.88	1.00	−1.87	.062
Milligan (2006)	0.92	0.86	0.98	−2.50	.013
Lowenberg (2003)	0.93	0.87	1.00	−1.98	.048
Thomas (1999)	0.92	0.86	0.99	−2.23	.026
Heil (1995)	0.92	0.87	0.99	−2.38	.017
DSF
Krug (2016)	0.85	0.77	0.94	−3.07	.002
Gao (2014)	0.92	0.85	0.98	−2.50	.013
Pabst (2012)	0.90	0.83	0.97	−2.87	.004
Milligan (2006)	0.90	0.84	0.97	−2.94	.003
Lowenberg (2003)	0.92	0.85	0.98	−2.42	.016
Thomas (1999)	0.90	0.84	0.96	−3.01	.003
CR
Krug (2016)	0.98	0.82	1.17	−0.23	.817
Pabst (2012)	0.95	0.83	1.08	−0.81	.419
Beksac (2010)	0.99	0.87	1.12	−0.14	.890
Milligan (2006)	1.00	0.88	1.14	0.01	.995
Lowenberg (2003)	1.02	0.89	1.16	0.25	.800
Kern (1999)	0.97	0.86	1.11	−0.39	.694
Thomas (1999)	0.98	0.86	1.11	−0.35	.723
Laver (1997)	0.99	0.87	1.12	−0.14	.892
Heil (1995)	0.99	0.87	1.12	−0.19	.849
Estey (1992)	1.03	0.90	1.16	0.39	.697
Relapse
Pabst (2012)	0.73	0.58	0.92	−2.64	.008
Beksac (2010)	0.81	0.68	0.98	−2.16	.030
Milligan (2006)	0.83	0.68	1.00	−1.97	.049
Lowenberg (2003)	0.83	0.67	1.03	−1.72	.086

Note: OS, overall survival; DFS, disease-free survival; and CR, complete remission.

Quality analysis

Quality analysis indicated that three of the studies, Kern et al. [27], Laver et al. [29], and Estey et al. [31], had potential selection bias and performance bias and did not describe an intent-to-treat analysis (Figure 3(A,B)). Across the studies, it was unclear if there was detection bias due to blinding of outcome assessments. There was no reporting bias among the studies.

Figure 3. Quality assessment results of (A) all included studies and (B) individual studies using the Cochrane Risk of Bias Tool.

Discussion

The prevention of relapse remains a significant challenge in the treatment of AML patients because of residual leukemic cells escaping the cytotoxic effect of chemotherapy. To optimize results of standard chemotherapy, the use of G-CSF or GM-CSF concurrently with induction chemotherapy has been studied in several randomized trials. In this meta-analysis, we evaluated the efficacy of treatment of AML with chemotherapy plus G-CSF. We found that for the overall analysis, significant advantages in OS and DFS were observed in patients receiving G-CSF plus chemotherapy compared with chemotherapy alone. Subgroup analysis found no significant difference between treatments in patients without prior AML treatment in OS or CR. However, patients with previously untreated primary/secondary AML who received G-CSF had significantly longer DFS and reduced incidence of relapse than those who did not receive G-CSF therapy. In the subgroup of patients with relapsed/refractory AML, no difference was found between treatment groups for OS, DFS, or CR. Sensitivity analysis suggests that the data for DFS are robust, but the findings for OS, CR, and relapse rate may have been overly impacted by certain studies. The findings of our meta-analysis suggest that treatment with G-CSF plus chemotherapy may provide better survival and treatment outcomes than with chemotherapy alone, particularly in patients with previously untreated AML.

Several prior meta-analyses have assessed the effect of G-CSF containing regimens in the treatment of AML [32–35]. Gurion et al. [34] evaluated the impact of CSFs (e.g. GM-CSF, G-CSF, etc.) added to chemotherapy in the treatment of patients with AML [34]. Their analysis included 19 studies comprising a total of 5256 patients. In contrast to our study, the addition of CSF to chemotherapy gave no additional benefit in all-cause mortality at 30 days and at the end of follow-up (relative risk (RR), 0.97; 95% CI 0.80–1.18 and RR 1.01; 95% CI 0.98–1.05, respectively) or in OS (HR 1.00; 95% CI 0.93–1.08). There was no difference between treatment with or without CSF in complete remission (RR 1.03; 95% CI 0.99–1.07), relapse rates (RR 0.97; 95% CI 0.89–1.05), or disease-free survival (HR 1.00; 95% CI 0.90–1.13). The presence of CSF also did not decrease the risk of bacteremia (RR 0.96; 95% CI 0.82–1.12) or invasive fungal infections (RR 1.40; 95% CI 0.90–2.19), and slightly increased adverse events (RR 1.33; 95% CI 1.00–1.56). Their findings suggest that CSF did not add benefit for treating patients with AML. However, their analysis pooled findings from studies including different CSFs, while our study focused only on G-CSF.

Wei et al. [33] performed a meta-analysis to assess the overall treatment efficacy and the adverse events of a treatment regimen that contained the topoisomerase inhibitors cytarabine and aclarubicin with G-CSF (CAG) in the treatment of patients with AML or myelodysplastic syndrome (MDS) [33]. The CAG treatment is widely used in China and Japan for these two diseases [33]. Their analysis included 814 AML and 215 MDS patients. They found that the CR rate was greater in AML patients (57.9%) than MDS patients (45.7%) (P < .01). They saw no difference in CR between previously untreated (new) (56.7%) and relapsed/refractory (60.1%) AML with CAG treatment (P > .05). Interestingly, CAG-containing regimens were associated with a significantly higher CR rate than non-CAG regimes (OR, 2.43). Four of our studies included cytarabine and another the topoisomerase inhibitor idarubicin [24,26,28,30]. However, a sufficient number of studies were not included to perform subgroup analysis to evaluate the efficacy of different chemotherapies with or without G-CSF in treating AML.

Lyman et al. performed a systematic review and meta-analysis that included 25 RCTs in which patients with AML or MDS received chemotherapy with (n = 6058) or without (n = 6746) G-CSF [35]. After a mean follow-up across studies of 60 months, AML/MDS was reported in 22 control patients and 43 G-CSF-treated patients, with an estimated relative risk (RR) of 1.92 (P = .009). Deaths were reported in 1845 patients treated with G-CSF and 2099 controls (RR 0.897 and absolute risk of 3.40% (P values < .001). The findings of the study indicate that the RR for AML/MDS relapse is increased but all-cause mortality is decreased in patients receiving chemotherapy with G-CSF support. Importantly the authors note that it is not possible to distinguish any leukemia risk associated with G-CSF from the known dose-dependent leukemia risk associated with myelosuppressive chemotherapeutic agents. In contrast to the findings of Lyman et al., we found that in the subgroup of patients with previously untreated AML, the priming of G-CSF reduced the relapse rate. The difference between studies may be due to the significantly different patient populations evaluated, and suggests that different types of patients might have responded to G-CSF therapy differently. Future studies are necessary to address this question.

A chemotherapy regimen that includes fludarabine, cytarabine, and G-CSF (FLAG) has been shown to be effective in treating relapsed and relapsed/refractory acute lymphoblastic leukemia (ALL) and AML [36–38]. Other commonly used related chemotherapies include FLAG plus mitoxantrone (FLANG), FLAG plus idarubicin (FLAG-IDA), and liposomal daunorubicin, fludarabine, cytarabine, G-CSF (L-DNR/FLAG) [36–38]. These combinations maximize favorable cytotoxic interactions between cytarabine and G-CSF, and between cytarabine and fludarabine [36]. FLAG-based regimens take advantage of the fact that fludarabine with cytarabine increases the intracellular retention of cytarabine’s active metabolite, cystosine arabinoside 5′ triphosphate, resulting in a synergistic antitumor effect. In addition, increasing cell cycling with G-CSF treatment response is thought to improve treatment response by causing dormant leukemic cells to be more sensitive to cytotoxic drugs [36]. Also, G-CSF increases the effect of cytarabine by increasing its incorporation into DNA [36].

There are several limitations to this study that should be considered when evaluating the findings. The number of studies was small and the type of chemotherapy, the different dosages of chemotherapy and G-CSF, and treatment cycles differed across studies. Moreover, patients across the studies differed with regard to AML-specific risk factors. These differences in treatment regimens and patient populations may have affected the response to treatments, which may have confounded our results. In addition, sensitivity analysis suggested that OS, CR, and relapse rate may have been overly impacted by certain studies, making these findings less robust.

In conclusion, this study examined the responses to G-CSF priming in patients with previously untreated or recurrent/relapsed AML who planned to undergo stem-cell transplantation. We found that receiving G-CSF together with chemotherapy might have provided better survival and treatment responses than with chemotherapy alone, particularly in patients with previously untreated AML. This and prior meta-analysis suggest that addition of G-CSF may improve survival.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This manuscript is supported by the National Natural Science Foundation of China, and the project number is U1401221.