Increased Serum Pro-B-type Natriuretic Peptide in Hematopoietic Progenitor Cell Donors Stimulated with G-CSF

Abstract

Introduction

Hematopoietic stem cells (HPCs) donors mobilized by granulocyte-colony-stimulating factor (G-CSF) can develop various signs and symptoms. proBNP (pro-B-type natriuretic peptide) is a serum marker of heart failure. A donor who developed severe adverse reactions after G-CSF mobilization was found to have high serum proBNP levels. We followed additional donors who received identical mobilization regimen to investigate the prevalence of this phenomenon.

Donors and Methods

Eighteen healthy donors underwent a mobilization regimen of 10 μg/Kg G-CSF daily for 5 days prior to allogeneic HPC collection using Spectra Optia between January 2016 to February 2017 were included in this study. Serum proBNP levels were measured before and after G-CSF stimulation and immediately after apheresis. Apheresis collection parameters and other laboratory results were also reviewed.

Results

The majority of donors (86.7%) had post-G-CSF elevation of serum proBNP. Seven of those had elevated proBNP above upper normal range (124 pg/mL). The subgroup of donors with normal proBNP is younger (median age of 37 vs. 42 years), with majority being male (90.9% vs. 28.6%) and with smaller processed blood volume (2.2 vs. 3 x total blood volume).

Discussion

This case series demonstrates an increase of serum proBNP can be common in HPC donors stimulated with 5 days of 10 mcg/Kg G-CSF. This is an adverse reaction that has not been described before. The temporary elevation of proBNP in these donors is not associated with ventricular dysfunction of the heart. The risk factors for marked elevation of proBNP post G-CSF should be further investigated.

Introduction:

Hematopoietic progenitor cells (HPCs) have been used for restoring marrow function for both primary marrow failure (such as aplastic anemia) and therapy-related marrow failure (secondary to chemotherapy) for more than 60 years. The application of HPCs has evolved and expanded since the first allogeneic stem cell transplantation in the 1950s was used to treat hematological malignancies. Hematopoietic stem cell transplantation (HSCT) is now used for congenital diseases such as sickle cell disease and congenital immunodeficiencies^1,2. Beside the more traditional role of hematological reconstitution by transplantation, HPCs are being increasingly tested in other clinical conditions such as liver insufficiency³, liver regeneration⁴ and cardiac insufficiency⁵.

With the introduction of granulocyte-colony stimulating factor (G-CSF) as a means to mobilize CD34+ HPCs for both marrow and peripheral harvesting, clinical applications for these HPCs have further widened. G-CSF mobilizes HPCs by an indirect way. G-CSF facilitates the breaking of the bond between HPCs and the niche in the marrow. It also stimulates the peripheral nervous system to increase catecholamine secretions which suppress osteocytes and osteoblasts, with an end result of reducing production of stromal cell-derived factor 1 (SDF1), also called C-X-C motif chemokine-12 (CXCL12), and changes of the sphingosine-1-phosphate (S1P) gradient that facilitates the movement of HPCs into the circulation⁶. Not only are healthy donors being actively recruited for G-CSF-stimulated HPC harvesting, patients with hematological malignancy or marrow failure diseases as well as patients with myocardial infarction⁷ and non-ischemic dilated cardiomyopathy^5,8 also used G-CSF mobilization for autologous HPC collection.

B-type (or brain) natriuretic peptide (BNP) is a myocardial hormone that is secreted and released into plasma in response to stress of the left ventricular wall. It includes several molecules, including mature BNP (BNP_1–32) and its metabolites, as well as the pre-prohormone (pre-proBNP), which is cleaved into N-terminal proBNP and BNP_1–32⁹. Traditionally, it is a biomarker for cardiac failure; however, increases in BNP have also been described in non-cardiac diseases, such as periodontitis¹⁰ and Kawasaki disease¹¹.

There are several well-known adverse reactions associated with G-CSF mobilization among healthy donors^12–14, and certain donor characteristics appear to be associated with risks of experiencing certain toxicities¹³. We encountered a donor who experienced unusual toxicities post G-CSF mobilization that included a surprisingly high level of serum BNP, a phenomenon which has never being reported before. We initiated a preliminary observational study to evaluate whether this is a common presentation among our HPC donors.

Case Report:

A healthy 41-year-old female (BMI 27.7) allogeneic peripheral blood HPC donor was mobilized with 10 μg/Kg G-CSF daily for 5 days. During the apheresis procedure, she complained of intractable headache, photophobia and nausea, however her vitals remained stable, with pre-procedure blood pressures of 113/74 mmHg vs. post-procedure 113/76 mmHg. She was counseled and discharged with a dose of oxycodone. However, her symptoms worsened and she presented to the emergency department (ED) 5 hours after completion of her donation. She reported “the worst headache in my life” at the ED, and she also appeared to be dyspneic. Her blood pressure at ED was elevated to 131/78 mmHg, heart rate was 90 beats per minute. Head CT was negative for hemorrhage or mass effect. Laboratory findings included several signs associated with G-CSF mobilization and apheresis, such as leukocytosis, decreased platelets, increased alkaline phosphatase (257 U/L [reference range 35–105 U/L]) and transaminases (AST 53 U/L (range <31 U/L) and ALT 50 U/L (range < 33 U/L), in addition to markedly elevated serum proBNP at 483 pg/mL (normal < 125 pg/mL for < 75 year-old), despite no other clinical evidence of cardiac insufficiency or pulmonary abnormalities. Her chest X-ray and echocardiogram were within normal limits. Her other laboratory results showed normal kidney function (BUN/Creatinine ratio of 9, normal 5–20), troponin T < 0.01 ng/mL.

This is the first known report of elevated serum proBNP associated with G-CSF mobilization in the literature. To investigate whether this was an isolated case or a common side effect among donors receiving G-CSF mobilization prior to HPC collection, we observed the subsequent 17 donors presenting to our institution who underwent the same mobilization regimen over the course of 12 months.

Donors and Methods:

Eighteen consecutive healthy donors who underwent a mobilization regimen of 10 μg/Kg G-CSF daily for 5 days prior to allogeneic HPC collection between January 2016 to February 2017 were included in this study. Serum proBNP levels (Roche COBAS 6000 Analyzer) were measured before G-CSF stimulation, after 5-days of G-CSF stimulation, but prior to the apheresis procedure, and immediately after apheresis. For four healthy donors who returned to have post-collection follow-up more than 7 days after apheresis collection, serum proBNP levels were also measured. Apheresis collection parameters, including processed volume and time of collection, as well as other laboratory values, including complete blood counts (CBC) and hepatic panel, were also reviewed.

Donors were evaluated after receiving G-CSF for HPC mobilization. The symptoms were graded as: mild – no medication needed, moderate – pain responded to medication (non-steroid anti-inflammatory drugs (NSAIDs) or oxycodone), and severe – when pain did not respond to medication (NSAID or oxycodone).

Apheresis for HPC collection was conducted on Spectra Optia Apheresis System (Terumo BCT, Inc, Lakewood, CO). ACD-A (anticoagulant citrate dextrose A) was used as the anticoagulant for the procedure. Blood volumes processed were predetermined based on the mobilization results and the weights of recipient/donor pair, ranging from 1.5 to 6 x total blood volumes (TBV) (median = 2.55 x TBV).

Statistical analysis

Statistical analyses were performed using GraphPad Prism (Version 8.0.2), including Mixed effects analysis, one-way ANOVA and Welch’s t-test. Also, simple linear regression by EXCEL (Microsoft) was used to analyze the relationship between post-stimulation proBNP levels and processed apheresis volume.

Results:

Donor characteristics

Donors included in this observational study were between 21 and 58 years old at the time of collection with the median age of 38 years. There were 12 males and 6 females. Ten donors were white, 6 were Hispanic and 2 were black (Table 1). The donors weighed from 47.7 Kg to 118 Kg at the time of collection with a median weight of 88 Kg. The pre-collection CD34 counts (post stimulation) ranged from 52 cells/mcL to 292 cells/mcL, with the median count of 87.5 cells/mcL. The final HPC product doses ranged from 5.34 × 10⁶/kg (of recipient weight) to 28.4 × 10⁶/kg with the median of 9.07 × 10⁶/kg (Data not shown). The processing volumes for HPCs apheresis in comparison to the donor’s total blood volume range from 1.5 to 5.9 fold with the median of 2.6 fold. (Table 1)

Table 1.

Donors Characteristics

		N (%) (Total N=18)
Sex	Male	12 (66.7%)
	Female	6 (33.3%)
Age (years)		Median = 38 (Range: 21–58)
BMI		Median 29 (Range: 20.9–38.1)
Race	White	10 (55.6%)
	Black	2 (11.1%)
	Hispanic	6 (33.3%)
Processed Volume (fold of TBV)		Median = 2.6 (Range: 1.5–5.9)
Baseline proBNP (pg/mL) (n=15)		Median = 24 (Range: 5–64)
Post-G-CSF symptoms	Severe	3 (16.7%)
	Moderate	2 (11.1%)
	Mild	13 (72.2%)

Serum proBNP Levels Post-G-CSF mobilization

Baseline (pre-G-CSF) serum proBNP levels were available for 15 donors, ranging from 5 pg/mL to 64 pg/mL with a median of 24 pg/mL. (Table 1) Among these 15 individuals, all but two donors (N=13) developed increased serum proBNP levels after G-CSF stimulation, ranging from 17 pg/mL to 237 pg/mL with a median of 78 pg/mL. Most donors experienced only mild increase of serum proBNP levels, which were well within the normal range (<125 pg/mL). Among those 13 who had elevated serum proBNP post G-CSF stimulation, seven demonstrated serum proBNP elevation above 125 pg/mL cut off value. Excluding the index donor who did not have pre-apheresis proBNP information, the remaining 6 donors had median post-stimulation proBNP level of 153 pg/mL (range: 133–237 pg/mL) immediately before apheresis, and a median of 181 pg/ml (range 124–254 pg/mL) post apheresis. The changes of serum proBNP levels at baseline, pre-apheresis, post-apheresis and 7 days after apheresis are depicted in Figure 1 (A and B).

Figure 1.

(A) Serum BNP levels over time. (B) Individual donor’s serum BNP levels over time.

Eight donors had hepatic panel results for pre-G-CSF stimulation and post-stimulation. All 8 donors demonstrated elevated alkaline phosphatase from baseline, and two of them also demonstrated elevated transaminases. (Table 2)

Table 2.

Hepatic panel (fold changes post G-CSF) in 8 donors

	High proBNP (n=4)	Normal proBNP (n=4)
ALK	2.5	2.8
ALT	2.3	0.8
AST	1.8	1.3

ALK: alkaline phosphatase; ALT: alanine aminotransferase; AST: aspartate aminotransferase

Donors with abnormal serum proBNP post stimulation vs. those with normal proBNP

Donors were stratified by their post-stimulation serum proBNP levels, using the cutoff of 125 pg/mL. The subgroup with normal proBNP (N=11) was younger (median age of 37 vs. 42 years, p < 0.0001), with majority being male (90.9% vs. 28.6%; p = 0.0004), and with slightly smaller processed blood volume (fold of total blood volume (TBV)) (median of 2.2 vs. 3) (Table 3). However, there was no relationship between processed volumes and proBNP elevation (Figure 2) (R² = 0.0095).

Table 3.

Donor characteristics stratified by normal vs. high proBNP post G-CSF mobilized HPCs apheresis.

		With high proBNP (Total N=7)	With normal proBNP (Total N=11)
Sex	Male	2 (28.6%)	10 (90.9%)
	Female	5 (71.4%)	1 (9.1%)
Age (years)		Median = 42 (Range: 27–58)	Median = 37 (Range: 21–50)
Race	White	3 (42.9%)	7 (63.6%)
	Black	1 (14.3%)	1 (9.1%)
	Hispanic	3 (42.9%)	3 (27.3%)
Processed Volume (fold of TBV)		Median = 3.0 (Range: 1.5–5.9)	Median = 2.2 (Range: 1.6–4.7)
proBNP (pg/mL) after G-CSF x 5 days		Median = 223 (Range: 124–483)	Median = 72.5 (Range: 12–109)
Post G-CSF BP (mmHg)		Median = 121/67 (Range 113–131/51–82)	Median = 126/72 (Range 112–152/56–80)
Post-apheresis BP (mmHg)		Median = 132/64 (Range: 113–146/61–94)	Median = 136/71 (Range: 117–154/55–81)
Post-G-CSF symptoms	Severe	3 (42.9%)	0 (0%)
	Moderate	1 (14.3%)	1 (9.1%)
	Mild	3 (42.9%)	10 (90.9%)

Figure 2.

Post-apheresis serum BNP levels in relationship to fold of processed volume to total blood volume.

Three donors who reported severe symptoms post G-CSF mobilization, including bone pain and headache not relieved by medication, were all Hispanic females with elevated serum proBNP. These comparisons are summarized in Table 3.

None of the donors with elevated serum proBNP post G-CSF stimulation demonstrated any clinical evidence of heart failure. Four donors with transient elevation of serum proBNP returned for follow-up and their serum proBNP levels returned to normal levels in 2–4 weeks (range 11–108 pg/mL).

Discussion:

The adverse effects of G-CSF in healthy HPC donors have been previously described^2,3,13–18. Despite the signs and symptoms of bone pain, headache, and transiently elevated alanine aminotransferase, alkaline phosphatase and lactate dehydrogenase, a 5-day course of G-CSF is generally considered safe for HPC donors. However, the differences in donor responses to G-CSF mobilization can also lead to qualitative differences of these HPC products. These variations are partially attributed to wide variations of pro- and anti-inflammatory cytokines, soluble adhesion molecules and other immunoregulatory molecules detected in healthy stimulated HPC donors¹⁹. One mechanism is for G-CSF to act through IL-6/gp130 signaling pathway to increase acute phase reactants, but the heterogeneous immunomodulation effects on donors add complexity to this widely used HPC-mobilizing regimen.

The index donor who suffered from debilitating headache requiring a post-apheresis ED visit was an unusual case. She demonstrated the expected laboratory findings post-G-CSF stimulation, such as transient elevated alanine aminotransferase, alkaline phosphatase and lactate dehydrogenase. However, the incidental finding of increased serum proBNP levels to the extent that suggesting potential cardiac toxicity was quite alarming. The donor received further cardiac evaluations but no cardiac pathology could be identified. Other causes of the aberrant serum proBNP results, including the apheresis procedure alone or in combination with G-CSF administration, are worth investigating.

The 17 consecutive HPC donors following this index donor in 12 months did demonstrate a common presentation of a transient increase of serum proBNP level from baseline from pre-G-CSF mobilization in the majority of donors (N=16, 88.9%), although most of the elevations were mild and the serum proBNP levels never reached above the upper reference range.

Our laboratory measures N-terminal proBNP (NT-proBNP) using COBAS system, which can react to both NT-proBNP and proBNP⁹. Various measurements of different BNP-related molecules, including deglycosylated NT-proBNP are suggested to better detect decompensating acute heart failure patients^9,20. We did not evaluate the rest of donors with elevated serum proBNP with echocardiogram or CT imaging, but clinically, none experienced exertional intolerance or other signs of cardiac dysfunction.

There is currently no literature linking G-CSF administration to elevated serum BNP levels, although certain apheresis procedures have been demonstrated to have positive effect on reducing serum BNP levels in selected patient populations. For example, there was documented reduction of BNP levels immediately after low-density lipoprotein (LDL) apheresis in patients with familial hypercholesterolemia²¹. Similarly, therapeutic plasma exchange can also decrease NT-proBNP and troponin in cardiac patients²². Several studies using G-CSF-stimulated endogenous HPC to treat patients with various cardiac conditions such as acute myocardial infarction (MI) and nonischemic dilated cardiomyopathy (DCM)^5,7,23 also demonstrated an improvement of left ventricular function with lowered serum BNP levels. For patients with acute MI, there might be a direct benefit of G-CSF in preventing cardiac remodeling by way of G-CSF acting via JAK-STAT3 pathway²³ in addition to the benefit of endogenous HPCs repair of damaged myocardium. Another similar approach was mobilizing endogenous CD34+ cells by G-CSF to treat patients with nonischemic DCM patients. In this DCM study, the endogenous CD34+ stem cells mobilized by G-CSF were shown to be beneficial for patients without diabetes mellitus (DM). These patients had decreased serum NT-proBNP levels and increased left ventricular ejection fraction (LVEF) 6 months after autologous CD34+ cells infusion⁵. One possible linkage of G-CSF mobilization and serum NT-proBNT might lie in the phosphoprotein osteopontin (OPN), which is increased in healthy G-CSF mobilized donors’ plasma²⁴. Kwee LC et al. identified several miRNAs that link OPN and NT-proBNT, two prognostic biomarkers in patients with non-ST elevated acute coronary syndrome (NSTE-ACS).²⁵ Although none of our donors demonstrated any cardiac signs and symptoms, the molecular pathways linking both biomarkers via miRNAs certainly point to possible interactions.

The findings in our pilot case series are certainly surprising given the seemingly benign and even beneficial effects of G-CSF in cardiac patients^5,7. Although three donors with the most significant elevation of serum proBNP levels were all female Hispanic donors, we refrain from hypothesizing donor’s risk factors based on this case series. The mechanisms of the transient elevation of BNP in these G-CSF mobilized HPC donors should be investigated further.