Impaired microvascular reactivity in patients treated with 5-fluorouracil chemotherapy regimens: Potential role of endothelial dysfunction

Abstract

Background: 5-fluorouracil (5-FU) is the second most common cancer chemotherapy associated with short- and long-term cardiotoxicity. Although the mechanisms mediating these toxicities are not well understood, patients often present with symptoms suggestive of microvascular dysfunction. We tested the hypotheses that patients undergoing cancer treatment with 5-FU based chemotherapy regimens would present with impaired microvascular reactivity and that these findings would be substantiated by decrements in endothelial nitric oxide synthase (eNOS) gene expression in 5-FU treated human coronary artery endothelial cells (HCAEC).

Methods: We first performed a cross-sectional analysis of 30 patients undergoing 5-FU based chemotherapy treatment for cancer (5-FU) and 32 controls (CON) matched for age, sex, body mass index, and prior health history (excluding cancer). Cutaneous microvascular reactivity was evaluated by laser Doppler flowmetry in response to endothelium-dependent (local skin heating; acetylcholine iontophoresis, ACh) and -independent (sodium nitroprusside iontophoresis, SNP) stimuli. In vitro experiments in HCAEC were completed to assess the effects of 5-FU on eNOS gene expression.

Results: 5-FU presented with diminished microvascular reactivity following eNOS-dependent local heating compared to CON (P = 0.001). Iontophoresis of the eNOS inhibitor L-NAME failed to alter the heating response in 5-FU (P = 0.95), despite significant reductions in CON (P = 0.03). These findings were corroborated by lower eNOS gene expression in 5-FU treated HCAEC (P < 0.01) compared to control. Peak vasodilation to ACh (P = 0.58) nor SNP (P = 0.39) were different between groups.

Conclusions: The present findings suggest diminished microvascular function along the eNOS-NO vasodilatory pathway in patients with cancer undergoing treatment with 5-FU-based chemotherapy regimens and thus, may provide insight into the underlying mechanisms of 5-FU cardiotoxicity.

Keywords: 5-fluorouracil; Cardio-oncology; Cardiotoxicity; Endothelium; Microcirculation.

Graphical abstract

Fig. 1

Impaired cutaneous microvascular reactivity following local heating in cancer patients treated with 5-FU. A) A representative tracing of the rise in CVC from baseline during a 35-minute local heating protocol for a 5-FU treated patient (5-FU, Male 64 years old) and matched control (CON, Male, 63 years old). B) 5-FU patients (n = 17, 7 women) presented with a significantly lower %ΔCVC than control subjects (n = 20, 9 women) (CON: 566 ± 305 %; 5-FU: 282 ± 172 %; P = 0.001, ES = 0.93) following 35-minutes of local heating suggesting the presence of impairment along the eNOS-NO vasodilator pathway following 5-FU chemotherapy. Data were compared using the Welch’s t-test and presented as mean ± SD.

Fig. 2

L-NAME fails to reduce the local heating response in 5-FU-treated cancer patients. A) L-NAME iontophoresis significantly reduced the %ΔCVC response to local heating in control participants (n = 7, 3 women) (SH: 564 ± 371 %; L + SH: 175 ± 122 %; P = 0.03, ES = 1.05) but not in patients undergoing cancer treatment with 5-FU (n = 7, 3 women) (SH: 327 ± 118 %; L + SH: 333 ± 227 %; P = 0.95, ES = 0.03) suggesting a lesser contribution of NO to the heated plateau following treatment with 5-FU chemotherapy. B) The absolute difference in the percent change between the SH and L treated sites was significantly greater in control participants compared to 5-FU treated cancer patients (CON: 389 ± 388 %; 5-FU: −6 ± 247 %; P = 0.04, ES = 1.21). SH = standard heating; L + SH = L-NAME + standard heating. Data were compared using Welch’s t-test (A, control cohort) and independent-samples t-test (A, 5-FU cohort; B) and presented as mean ± SD.

Fig. 3

Cutaneous microvascular reactivity in response to ACh iontophoresis is maintained in 5-FU-treated cancer patients. A) ACh iontophoresis resulted in a significant rise in %ΔCVC with successive ACh impulses (P = < 0.001); however, this rise was not different between patients treated with 5-FU (n = 17, 4 women) and control subjects (n = 15, 7 women) (P = 0.44). B) This finding was corroborated by the similarity in the peak response to ACh between groups (CON: 920 ± 416 %; 5-FU: 835 ± 448 %; P = 0.58), suggesting some endothelium-dependent vasodilatory pathways may remain intact following 5-FU-based chemotherapy. Data were compared using two-way repeated measures ANOVA (A) and independent- samples t-test (B). All data are presented as mean ± SD.

Fig. 4

eNOS gene expression is reduced in 5-FU treated HCAEC. There was a significant main effect for differences in eNOS gene expression between the groups following the 24-hr. incubation period (P = 0.002, ES = 0.72). CON cells (1.114 ± 0.06 a.u) had significantly higher levels of eNOS gene expression compared to cells treated with both 0.7 mM (0.711 ± 0.12 a.u; P = 0.008, ES = 0.52) and 7.0 mM (0.578 ± 0.13 a.u; P = 0.002, ES = 0.69) of 5-FU, while no differences were detected between the two 5-FU concentrations (P = 0.34). n = 3 cultures per group. All samples were run in duplicate. Data were analyzed using One-way ANOVA and are presented as mean ± SD. a.u. = arbitrary units. eNOS = endothelial nitric oxide synthase.