Balance of transforming growth factor-β1 and platelet-derived growth factor-BB is associated with kidney allograft rejection

Abstract

Abstract

Background

About 50% of kidney-transplant patients undergo organ rejection within 10 years. Chronic allograft nephropathy (CAN) represents the dominant cause of kidney transplant failure and accounts for 50–80% of graft loss in long-term surviving patients. CAN pathogenesis is multifactorial and not-completely elucidated; several reports indicate TGF-β1 and platelet-derived growth factor (PDGF)-BB expression in CAN suggesting a possible role of these factors in the allograft arteriosclerosis and graft failure.

Methods

We investigated the plasma expression concentrations of human growth factors with enzyme-linked immunosorbent assays and appropriate statistical analysis.

Results

We present evidences showing statistically significant association of CAN with a specific balance between TGF Beat1 and PDGF-BB plasma concentrations, in 129 kidney-transplant patients and 15 healthy controls. Odds ratios were computed to correlate expression-levels with CAN occurrence.

Conclusion

We believe these data may suggest a novel non-invasive method to identify early molecular markers of graft deterioration.

Introduction

About 50% of kidney-transplant patients undergo organ rejection within 10 years. Chronic allograft nephropathy (CAN) represents the main cause of kidney-transplant failure and accounts for 50–80% of graft loss in long-term surviving patients. ¹ CAN pathogenesis is multifactorial and not completely understood; several reports indicate transforming growth factor (TGF)-β1 and platelet-derived growth factor (PDGF)-BB expression in CAN, ^2–5 suggesting a role of these factors in the allograft arteriosclerosis and graft failure. We present evidence showing significant association of CAN with plasma TGF-β1/PDGF-BB ratio, in 129 kidney-transplant patients. We believe these data might indicate novel mechanisms underlying CAN occurrence and may suggest a novel non-invasive method to identify early molecular markers of graft deterioration.

Methods

All patients were undergoing regular haemodialysis prior to transplantation and showed continued allograft function for at least one year post-transplant. The mean patient age was 37 ± 11 years. For each patient, human leukocyte antigens (HLA) and panel reactive lymphocytotoxic antibodies (PRA) were measured before transplantation. Pretransplant blood transfusions occurrence, PRA status, HLA A-B-DR mismatches showed no difference between CAN and non-CAN transplanted patients. Baseline immuno-suppression was achieved by cyclosporine treatment (30% of patients), tacrolimus (60% of patients) and sirolimus + calcineurin inhibitors (10% of patients). All patients received steroids for induction; prednisolone dosage was tapered to reach a maintenance dose of 5 mg/day by the end of the third postoperative month. Over 90% of patients underwent continued steroid maintenance therapy. A suspicion of CAN was confirmed by graft biopsy, performed under ultrasound guidance, and classified according to the Banff criteria for chronic nephropathy.

Peripheral venous blood was collected using EDTA as anticoagulant and centrifuged at 1000 rpm within 30 min. Platelets were removed with centrifugation at 10.000 g (10 min, 2–8°C). Samples were stored at −70°C until each assay was performed. All measurements were performed according to manufacturer's instructions. TGF-β1 was activated by acidification and subsequent neutralization. Samples were then diluted 1:12 with Calibrator Diluent RD6M. Expression levels of TGF-β1 and PDGF-BB were then determined using a solid-phase specific sandwich enzyme-linked immunosorbent assay (R&D Systems).

The receiver operating characteristic (ROC curve) plots of sensitivity versus 1-specificity for all TGF-β1 and PDGF-BB values were obtained. ⁶ The best cut-off values were identified, as indicated by the point closest to the upper left corner of the diagram. According to this analysis, the cut-off values identified were 17,500 pg/mL for TGF-β1 and 1000 pg/mL for PDGF-BB and were used for the following statistical analyses. The nonparametric Mann-Whitney U-test for independent samples was used to compare controls to transplanted patients. Odds ratios (ORs ) adjusted for sex, age and time from transplant were also calculated. All analyses were performed using the software SPSS for Windows, version 9.0 (SPSS Inc, USA).

The study was approved by the ethics committee and written informed consent was obtained.

Results and discussion

Median (interquartile range) of TGF-β1 and PDGF-BB concentrations were markedly and significantly different in 15 healthy controls as compared to transplanted patients (TGF-β1: controls, 5040 [4059–6218] pg/mL; CAN patients, 9682 [5317–16,116] pg/mL; non-CAN patients, 13,674 [7841–21,786] pg/mL; PDGF-BB: controls 44 [2–230] pg/mL; CAN patients, 1069 [373–1835] pg/mL; non-CAN patients, 947 [394–1488] mg/L). Differences between healthy controls and patients were always statistically significant (P ≤ 0.001 in all cases).

CAN occurrence was found to be strongly associated with the time from transplant, with nine months from transplant being the limit beyond which the occurrence increased up to six times, as compared with shorter times (Table 1). CAN occurrence was not associated with the concentrations of TGF-β1 or PDGF-BB, categorized as TGF-β1 concentration below or above 17,500 pg/mL and PDGF-BB concentration below or above 1000 pg/mL (Table 1), according to cut-off concentrations identified by ROC analysis. Since TGF-β1 and PDGF-BB are both suggested to have a role in CAN pathogenesis, we hypothesized that the balance between the two cytokines may be more informative than the absolute values; we thus expressed such balance as the ratio of their plasma concentrations. Interestingly, when CAN occurrence was estimated as a function of the ratio TGF-β1/PDGF-BB, patients with a ratio below 10 showed a 4.9-fold higher association with CAN compared with patients with a ratio above 10 (P < 0.001) (Table 1).

Table 1

Association (odds ratio [OR]) with chronic allograft nephropathy (CAN)

		Overall (n)	CAN		OR	95% CI	P value
			n	%
Overall	–	144	49	33.6	–	–
Gender	Male	92	31	33.3	1	–
	Female	52	18	34	1.0*	0.5–2.3	0.905
Age (years)	<40	49	15	31.3	1	–
	40–49	45	16	35.6	1.4*	0.5–3.4	0.509
	≥50	50	16	32	1.0*	0.4–2.5	0.995
Time from transplant (months)	<9	37	5	13.2	1	–
	10–35	20	8	40	6.0*	1.5–24.4	0.013
	≥36	87	36	40.9	6.2*	2.0–19.5	0.002
TGF-β1 (pg/mL)	≥17,500	45	11	23.9	1	–
	<17,500	99	38	38	2.1	0.9–4.9	0.084
PDGF-BB (pg/mL)	<1000	74	23	30.7	1	–
	≥1000	70	26	36.6	1.3	0.6–2.7	0.054
TGF-β1/PDGF-BB ratio	≥10	112	30	26.5	1	–
	<10	32	19	57.6	4.9	1.9–12.2	<0.001

*Estimates for sex, age and time from transplant are from the model with TGF-β1.

OR adjusted for sex, age and time from transplant were computed with SPSS for Windows, version 9.0 (SPSS Inc, USA).

TGF, transforming growth factor; PGDF, platelet-derived growth factor

Conclusion

These findings suggest that a specific balance ratio of TGF-β1 and PDGF-BB concentrations may be required to allow the allograft to maintain its proper function and to prevent rejection. Any change in this ratio may then promote kidney graft failure. According to Savikko et al. ³ and to Sihvola, ⁵ inhibiting PDGF signalling strongly reduces CAN occurrence. We may therefore speculate that a TGF-β1/PDGF-BB ratio >10 may inhibit PDGF signalling because the PDGF concentration is reduced or because increased concentrations of TGF-β1 may interfere with PDGF signalling. Such data also suggest that a serum marker associated with CAN occurrence may be identified in the blood stream.