In general, lymph from the lower part of the torso and viscera enters the bloodstream via the thoracic duct at the left subclavian-jugular venous junction (see Figs. 13-1 and 13-2). Lymphatics from the head and neck and from the upper extremities enter the central veins either independently or by a common supraclavicular cistern. Numerous interconnections exist within this rich vascular network, and subvariant anatomic pathways are plentiful. For example, the bulk of cardiac and pulmonary lymph, as well as intraperitoneal fluid, which drains through fenestrae of the diaphragm into substernal mediastinal collectors, unite as a common trunk to empty into the great veins in the right side of the neck. In contrast, intestinal lymph, which transports cholesterol, fat-soluble vitamins (vitamins A, D, K, and E), and long-chain triglycerides as chylomicra, courses retroperitoneally to the aortic hiatus, to form with other visceral and retroperitoneal lymphatics, the multichannel cisterna chyli and the thoracic duct, unlike intestinal blood, which flows directly into the liver. The bulk of the lymph formed in the liver flows retrograde or countercurrent to portal blood flow and joins intestinal lymph collectors just before the origin of the thoracic duct. Only a small amount of hepatic lymph drains anterograde along the major hepatic veins to the anterior mediastinum and right lymph duct. Although these topographic variants influence the development and progression of peripheral (lymph)edema only indirectly, they are nonetheless essential for a broad understanding of edema syndromes, including those accompanied by visceral lymphatic abnormalities, celomic effusions, and chylous reflux.

Although the retina and brain do not technically have lymphatic apparatuses, they possess analogous circulations, such as the aqueous humor canal of Schlemm (the anterior chamber of the eye) or the cerebrospinal fluid/subarachnoid villus (pacchionian bodies) connections (the brain). Glial elements and non–endothelial-lined intracerebral perivascular (Virchow-Robin) spaces probably also serve to transport interstitial fluid to nearby intracranial venous sinuses. Extensive interruption of the cervical lymphatic trunks (e.g., after bilateral radical neck dissection) therefore causes prominent facial suffusion and a transient neurologic syndrome resembling pseudotumor cerebri,³ whereas an infusion of crystalloid solution directly into the canine cisterna magna causes an elevation in intracranial pressure and increases lymph flow from draining neck lymphatics.^4,5 Although abundant lymphatic pathways thus exist for surplus tissue fluid to return to the bloodstream, homeostasis of the internal environment nonetheless still depends on an unimpeded, intact, interstitial-lymph fluid circulatory system (see Figs. 13-1 and 13-2).

Embryonic Development

Controversy has persisted since the early 1900s about the embryologic origin of lymphatics, that is, lymphovascu­logenesis (endothelial precursors or stem cells, such as lymphangioblasts, differentiate and proliferate into a primitive tubular network) and subsequent lymphangiogenesis (sprouting from existing vessels) (Box 13-1).⁶ According to Sabin, lymphatics and veins derive from a common primordium in the venous system (Fig. 13-3).⁷ After injection of dye into pig skin, gradually extending lymphatic plexuses were observed, starting first at the base of the neck, which were in direct communication with the venous system. Sabin concluded that the primary lymphatic plexuses derive from central veins and that their growth progresses centrifugally by sprouting toward the periphery and ultimately forming the superficial lymphatic system. In contrast, Kampmeier,⁸ after a review of serial tissue sections, including Sabin’s original human embryo preparations, and phylogenetic considerations, proposed the centripetal theory that the lymphatic system arises independently from tissue mesenchyme in peripheral tissues and the surrounding primary lymph sacs and only later joins the central venous system to complete the blood-lymph loop. Kampmeier reasoned further that as the developing blood capillaries begin to “leak clear fluid into tissues, lymph amasses in ever growing volume within the mesenchyme first collecting in separate pools or spaces, which then become confluent as the high pressure stream seeks an outlet…the current usurps the course of decadent and vanishing venules, expands and pours into the developing jugular lymph sac, which meanwhile has acquired continuity with the forming arch of the cardinal lymphatic. Thoracic and abdominal lymphatics form similarly: Discontinuous spaces spring up in the mesenchyme next to the intercardinal or periaortic venous plexus in the path of the potential lymphatic; and primordia quickly widen, lengthen and fuse into a continuous vessel.”⁸

Although recent studies of the growth regulatory gene PROX1⁹ and the receptor for lymphatic vascular endothelial growth factor (VEGFR-3)¹⁰ have tended to support the centrifugal theory (origin from venous spouting), other elegant work has demonstrated a substantial centripetal contribution from mesenchymal lymphangioblasts in the engrafted wing lymphatic system of chimeric quail chick embryos¹¹ and early avian embryos through PROX1 staining.¹² Both processes may contribute in various degrees to the ultimate link between the lymph and blood vasculature (e.g., the cervical thoracic duct/venous junction in humans). In addition, similar processes of lymphovasculogenesis and lymphangiogenesis are recapitulated after birth and probably involve the same complex expanding cascade of vascular growth factors (particularly the VEGF and angiopoietin families), endothelial receptors, and transcription factors already implicated in lymphatic growth and lymphedema-angiodysplasia syndromes (Table 13-1) (see the section on Lymphangiogenesis).

As an afferent vascular system, the lymphatics originate within the interstitium as specialized capillaries, although in certain organs, such as the liver, they seem to emanate from nonendothelialized precapillary channels (e.g., the spaces of Disse).¹³ Lymphatic capillaries are remarkably porous and readily permit the entry of even large macromolecules (molecular weight >1000 kD). In this respect, they resemble the uniquely “leaky” fenestrated sinusoidal blood capillaries of the liver, but are in distinct contrast to most other blood capillaries, which are relatively impervious to macromolecules even the size of albumin (molecular weight, 69 kD).¹⁴

Under light microscopy without pre–paraffin-embedded tracer or intravascular latex injection, it is difficult to distinguish between blood and lymph vessels, although the latter are usually thin walled and tortuous, have a wider, more irregular lumen, and are largely devoid of red blood cells. Many staining features have been advocated to differentiate between blood and lymph microvasculature, such as the endothelial marker factor VIII–related antigen : von Willebrand factor (vWF). Although staining characteristics vary in both normal and pathologic states and at different sites (perhaps related to endothelial cell de-differentiation), in general, lymphatic staining resembles but is less intense than its blood vessel counterpart. In other words, the staining differences have been more quantitative than qualitative.^15–17 Most recently, several new markers have been proposed that appear to be more lymphatic specific, particularly antibodies to lymphatic endothelial hyaluronan receptor-1 (LYVE-1; a homologue of CD44 glycoprotein),¹⁸ podoplanin,¹⁹ PROX1,⁹ and to a lesser degree, VEGFR-3 (the endothelial receptor for VEGF-C, the so-called lymphatic growth factor),^20,21 as well as 5′-nucleotidase (see under the next section, Lymphatic-Specific Markers).²²

Lymphatic-Specific Markers

Lymphatic vessels can increasingly be distinguished from blood vessels in tissue sections by whole-mount staining with specific markers. Only 10 years ago, little histochemical specificity existed to distinguish lymphatic vessels from blood capillaries and veins, and identification was based primarily on morphology, distinctive ultrastructure, or both. One of the most commonly used and most specific markers in use today is LYVE-1.^19,23 It has been applied to tissues ranging from early mouse embryo to adult human and highlights collecting vessels and lymphatic capillaries (but not larger caliber vessels). Another strong marker localizing to the nucleus of lymphatic endothelial cells and adaptable to multiple tissues is the transcription factor PROX1.²³ Podoplanin recognizes a transmembrane glycoprotein²⁴ in lymphatics, but not blood vessel endothelial cells in the mouse, whereas its human analogue, D2-40,²⁵ also sharply distinguishes lymphatic from blood vessel endothelium, but stains other distinguishable cells. This feature has been particularly useful in identifying preexisting and new lymphatics in tumor specimens and in generating quantitative differentials from blood vessels and indices of lymphatic invasiveness and tumor dissemination.²⁶ 5′-Nucleotidase staining is used by some research laboratories for its lymphatic specificity,^22,27 and other common markers that show some cross reactivity to veins include VEGFR-3¹⁰ and neuropilin-2²⁸ (reviewed elsewhere^29,30). Thus, an array of lymphatic markers are now available to distinguish lymphatic from blood vessels, although there is some overlap of cell types in normal conditions and even more so in pathologic states.

Ultrastructure

Ultrastructurally, lymph capillaries display both “open” and “closed (tight)” endothelial junctions, often with prominent convolutions.³¹ Depending on the extent of tissue “activity,” these capillaries can dramatically adjust their shape and lumen size. Unlike blood capillaries, a basal lamina (basement membrane) is tenuous or lacking altogether in lymph capillaries.^31,32 Moreover, complex elastic fibrils, termed anchoring filaments, tether the outer portions of the endothelium to a fibrous gel matrix in the interstitium.^33–35 These filaments allow the lymph microvessels to open wide, which causes a sudden increase in tissue fluid load and pressure, in contrast to the simultaneous collapse of adjacent blood capillaries (Fig. 13-4). Just beyond the lymph capillaries are the terminal lymphatics. In contrast to more proximal and larger lymph collectors and trunks, the terminal lymphatics are devoid of smooth muscle, although the endothelial lining is rich in the contractile protein actin.¹⁵ Intraluminal bicuspid valves are also prominent features and serve to partition the lymphatic vessels into discrete contractile segments termed lymphangions.³⁶ These specialized microscopic features of the lymphatic network support this delicate apparatus’ function of absorbing and transporting lymph node elements and the large protein moieties, cells, and foreign agents of the bloodstream (e.g., viruses, bacteria) that gain access to the interstitial space (Figs. 13-5 and 13-6).

Early physicians were able to visualize the lymphatic system only by observing chyle-filled mesenteric lymphatics. Asellius’ initial publication³⁷ included what has been reported as the first color anatomic plate in history (Fig. 13-7).¹ This was followed by intralymphatic injection of mercury into cadavers by Nuck in 1692, which depicted fine channels,³⁸ then the detailed and elegant work of Mascagni in 1787,³⁹ and subsequently, the classic images of both subcutaneous and deep vessels by Sappey in 1874.⁴⁰ von Recklinghausen used silver nitrate, which allowed imaging to take place without removal of surrounding tissue and facilitated visualization of lymphatic vessels as distinct from blood capillaries.⁴¹ Gerota developed a technique in 1896 of injecting a mixture of Prussian blue and turpentine to highlight the vessels,⁴² and this was followed in 1933 by the intracutaneous injection of vital dyes that bind to tissue proteins by Hudack and McMaster,⁴³ which is a technique still used today for investigation and in the clinic (Fig. 13-8).

Modern imaging techniques also include direct (intralymphatic) injection of oily contrast agents, termed lymphangiography, as first described by Kinmonth in 1954,⁴⁴ and whole-body lymphangioscintigraphy after subcutaneous or intradermal injection of protein-bound radiotracers (see Witte et al⁴⁵ for an overview) (see Fig 13-8). Other agents used for indirect lymphography include a variety of fluorescent or magnetic particles,^46,47 infrared particles and dyes,^48–50 immunoglobulin conjugates,⁵¹ and microbubbles⁵² for detection with fluorescent microscopes, optical imaging systems, computed tomography (CT), magnetic resonance imaging (MRI) (with and without contrast),⁵³ and ultrasound,^54,55 with an expanding potential for highly specific molecular imaging. New contrast agents and techniques are continuing to be developed.^56–58

Flow-Pressure Dynamics

Although lymphatic vessels, like veins, are thin-walled flexible conduits that return liquid to the heart, the flow-pressure relationships in the venous system and the lymphatic system are different. The energy to drive blood in the venous system derives primarily from the thrust of the heart. The cardiac propulsive boost maintains a pressure head through the arteries and blood capillaries into the veins. There is extensive neural regulation of venous tone that is essential for many activities, postural changes, and the negative pressure in the chest during inspiration. In addition, external compressive forces such as muscle can substantially alter venous pressure and gradients. Thus, muscular contractions such as walking and running, in the presence of competent venous valves, supplement cardiac action in facilitating return of blood to the heart.

In contrast, lymph vessels in tissues are not directly contiguous with the blood vasculature, and the chief source of energy for propulsion of lymph emanates from the intrinsic lymphatic truncal wall contractions (propulsor lymphaticum).^61–77,80 Like amphibian lymph hearts (cor lymphaticum), mammalian lymphatic smooth muscle beats rhythmically, and in the presence of a well-developed intraluminal valve system, facilitates transport of lymph.⁸¹ In a sense, the lymphatic structures function as micropumps that respond to fluid challenges with increases in both rate and stroke volume.^36,68 Ordinarily, resistance to flow in the lymphatic vessels is relatively high in comparison to the low resistance in the venous system,⁸² but the pumping capacity of the lymphatics is able to overcome this impedance by generating intraluminal pressures of 30 to 50 mm Hg (see Fig. 13-11) and sometimes even equaling or exceeding arterial pressure.^61,82,83 This formidable lymphatic ejection force is modulated not only by filling pressure but also by temperature, sympathomimetics, neurogenic stimuli, circulating hormones, and locally released paracrine and autocrine cytokine secretions.⁸⁴

It is often mistakenly thought that lymph return, like venous return, is directly enhanced by truncal compression from skeletal muscle and other adjacent structures. Although muscular contraction and external massage clearly accelerate lymph return in the presence of edema,⁸³ under normal conditions, peripheral lymph flow is regulated primarily by spontaneous contraction of the lymphatics themselves.^83,84 In peripheral lymphatics, unlike peripheral veins, the column of liquid is incomplete. Accordingly, with normal intralymphatic pressure, external compression is ineffective in propelling lymph onward (Fig. 13-13), although it may increase the frequency and amplitude of lymphatic contractions. In other words, lymphatics, in contrast to veins, are not sufficiently “primed.”⁸⁵ During lymphatic obstruction and persistent lymph stasis, hydrostatic pressure in the draining tissue watersheds and lymphatics rise as intrinsic truncal contractions fail to expel lymph completely. In this circumstance, in contrast to the normal situation, the fluid column in the lymphatics becomes continuous, and skeletal muscle or forceful external compression then becomes an effective pumping mechanism that aids lymph transport.^83,85

Studies of the effect of gravity on peripheral lymphatic and venous pressure confirm these findings. Although assuming an erect position sharply raises distal venous pressure, peripheral intralymphatic pressure is unaffected, although lymphatic truncal pulsation increases in both frequency and amplitude.⁸⁵ This arrangement favors removal of tissue fluid by the lymphatics during dependency, because unlike veins, the lymphatics operate at much lower hydrostatic pressure. Once the lymphatics become obstructed, truncal contractions quicken at first, but then the intraluminal valves gradually give way, and as the lymph fluid column becomes continuous, this mechanical advantage is lost, and chronic lymphedema supervenes (Fig. 13-14).