Reduced Lymphatic Reserve in Heart Failure With Preserved Ejection Fraction

Background Microvascular dysfunction plays an important role in the pathogenesis of heart failure with preserved ejection fraction (HFpEF). However, no mechanistic link between systemic microvasculature and congestion, a central feature of the syndrome, has yet been investigated. Objectives This study aimed to investigate capillary–interstitium fluid exchange in HFpEF, including lymphatic drainage and the potential osmotic forces exerted by any hypertonic tissue Na+ excess. Methods Patients with HFpEF and healthy control subjects of similar age and sex distributions (n = 16 per group) underwent: 1) a skin biopsy for vascular immunohistochemistry, gene expression, and chemical (water, Na+, and K+) analyses; and 2) venous occlusion plethysmography to assess peripheral microvascular filtration coefficient (measuring capillary fluid extravasation) and isovolumetric pressure (above which lymphatic drainage cannot compensate for fluid extravasation). Results Skin biopsies in patients with HFpEF showed rarefaction of small blood and lymphatic vessels (p = 0.003 and p = 0.012, respectively); residual skin lymphatics showed a larger diameter (p = 0.007) and lower expression of lymphatic differentiation and function markers (LYVE-1 [lymphatic vessel endothelial hyaluronan receptor 1]: p < 0.05; PROX-1 [prospero homeobox protein 1]: p < 0.001) compared with control subjects. In patients with HFpEF, microvascular filtration coefficient was lower (calf: 3.30 [interquartile range (IQR): 2.33 to 3.88] l × 100 ml of tissue–1 × min–1 × mm Hg–1 vs. 4.66 [IQR: 3.70 to 6.15] μl × 100 ml of tissue–1 × min–1 × mm Hg–1; p < 0.01; forearm: 5.16 [IQR: 3.86 to 5.43] l × 100 ml of tissue–1 × min–1 × mm Hg–1 vs. 5.66 [IQR: 4.69 to 8.38] μl × 100 ml of tissue–1 × min–1 × mm Hg–1; p > 0.05), in keeping with blood vascular rarefaction and the lack of any observed hypertonic skin Na+ excess, but the lymphatic drainage was impaired (isovolumetric pressure in patients with HFpEF vs. control subjects: calf 16 ± 4 mm Hg vs. 22 ± 4 mm Hg; p < 0.005; forearm 17 ± 4 mm Hg vs. 25 ± 5 mm Hg; p < 0.001). Conclusions Peripheral lymphatic vessels in patients with HFpEF exhibit structural and molecular alterations and cannot effectively compensate for fluid extravasation and interstitial accumulation by commensurate drainage. Reduced lymphatic reserve may represent a novel therapeutic target.

H eart failure (HF) is a leading cause of morbidity and mortality (1). Current trends of increased HF hospitalizations are mostly driven by HF with preserved ejection fraction (HFpEF), carrying a 1-year prognosis almost as poor as in patients with HF with reduced ejection fraction (2)(3)(4).
HFpEF is a clinical syndrome closely associated with multiple cardiovascular comorbidities and risk factors, such as hypertension, obesity, coronary artery disease, diabetes mellitus, atrial fibrillation, and chronic kidney disease (5). The notion that these comorbidities not only are associated with HFpEF, but also may be directly involved in its pathogenesis via comorbidityassociated inflammation and coronary microvascular dysfunction (6) has gained support from a large body of postmortem, noninvasive, and invasive evidence (7)(8)(9). In addition to the coronary vascular bed, peripheral vessels also appear dysfunctional (8). In particular, impaired systemic vasodilator reserve and low skeletal muscle capillary density, as well as low peripheral O 2 extraction paralleling microvascular rarefaction, were reported as determinants of exercise intolerance in patients with HFpEF (4, [10][11][12]. Such findings challenged the paradigm of a purely cardiac disorder in favor of a more systemic phenomenon (13). However, whether and how a dysfunctional microcirculation could directly impact congestion, "the core of the HF syndrome" (5), has not yet been investigated.
In recent years, understanding of body fluid homeostasis has evolved through reappraisal of tissue sodium (Na þ ) accumulation. This phenomenon was proposed to be water-independent (14) and to induce a hypertonicity-driven lymphatic network expansion in order to facilitate local Na þ drainage (15). Notably, tissue Na þ excess was found in most of the conditions and risk factors associated with the clinical HFpEF syndrome (i.e., older age, hypertension, diabetes, and chronic kidney disease) (16)(17)(18), but its hypertonic nature and functional relevance to fluid homeostasis in HFpEF lacks demonstration.
Therefore, the HAPPIFY (Heart fAilure with Preserved ejection fraction: Plethysmography for Interstitial Function and skin biopsY) study was designed to investigate capillary-interstitium fluid exchange and to test the hypothesis that an osmotic effect secondary to high interstitial Na þ levels could impact  (19). Exclusion criteria were a history of recent (<3 months) cerebrovascular event, myocardial infarction, or coronary revascularization;  and Clyde NHS Research and Development (ref.

GN17CA152).
TISSUE SAMPLES. All participants (except for 1 patient, owing to ineffective anesthesia) underwent a 4-mm skin punch biopsy on a gluteal external upper quadrant, after topical numbing with Na þ -free lidocaine cream (LMX4, Ferndale Pharma Group, Ferndale, Michigan) as described (20). One-half of the skin sample was fixed in paraformaldehyde 2% for histology, and the other one-half was immediately frozen and stored at -80 C until chemical analysis.
A skin portion of the larger biopsies obtained from participants eligible and consenting also to the substudy was immediately frozen for molecular biology (7 patients with HFpEF and 10 HC subjects); the remaining tissue was used to dissect small resistance arteries for ex vivo functional testing of endotheliumdependent and endothelium-independent relaxation by wire myography. Briefly, strain gauges were positioned on the nondominant forearm and on the ipsilateral calf and were maintained at the height of the right atrium.
Inflatable cuffs for venous occlusion were placed proximal to the strain gauges. P V was determined by gradually increasing the occlusion pressure, sustained until any limb volume change was detected. Arterial blood flow was determined from the rate of change in limb volume after consecutive cycles of sudden venous occlusion to 45 mm Hg (E20 Rapid Cuff Inflator, D.E. Hokanson Inc., Bellevue, Washington), as described previously (22). Arterial resistance to flow was calculated as: (mean BP À P V ) / blood flow (24).
Compared with HC subjects, they had higher BNP, albuminuria, and plasma urea, and lower eGFR and hematocrit, but similar plasma albumin.
Classical assessment of peripheral vascular function also revealed higher carotid-femoral pulsed wave velocity, as a measure of stiffness of large arteries, and lower brachial flow-mediated dilatation, which was accompanied by lower post-ischemic reactive hyperemia (Supplemental Table 2, Supplemental Values are n (%), mean AE SD, or median (interquartile range). Chronic kidney disease was defined as eGFR <60 ml/min/1.73 m 2 .
Rossitto et al. Lymphatics in HFpEF  Values are mean AE SD or median (interquartile range).
Abbreviations as in Table 1. Lymphatics in HFpEF rest P V was higher, although not to values suggestive of overt decompensation ( Table 2).    Table 2). Red lines (HFpEF) intersect the x-axis at lower pressures compared with blue lines (HC subjects; scale magnification below): the intersect indicates the threshold above which interstitial fluid accumulation starts to develop (isovolumetric pressure [P i ]). Data presented as mean AE SD; **p < 0.01, ***p < 0.001. P cuff ¼ pressure applied to the cuff; other abbreviations as in Figure 1.

Because of motion artifacts
Rossitto et al. Lymphatics in HFpEF CVP (as one can estimate by P V values in Table 2 In healthy subjects, the fluid filtering out of the capillary bed of the blood vasculature (BV) is evenly balanced by commensurate fluid drainage by lymphatic vasculature (LV); as a result, the physiological amount of interstitial fluid is homeostatically preserved. In heart failure with preserved ejection fraction (HFpEF), the net fluid extravasation tends to be lower because of the reduced vascular surface available for fluid exchange (i.e., capillary rarefaction) and the possible diversion of blood flow toward arteriovenous shunts; however, drainage by the impaired lymphatic system is inadequate to meet demands and facilitates accumulation of interstitial fluid at lower venous pressures. Both anatomical and functional defects (light green) could explain the reduced lymphatic reserve.