The Medullary Circulation
Black (1965) provided an historical overview on the development of current concepts of medullary function: ‘Kuhn and Ryffel had suggested that the arrangement of tubes in the renal medulla qualified it to act as a single “countercurrent multiplier”; but it was some years until the patient advocacy of Wirz brought the suggestion to the effective notice of English-speaking nephrophiles.
It was the German nephrophiles, however, Kramer, Thurau and Deetjen who recognized the implications of the blood supply for the countercurrent mechanism, who developed the first methodology required to provide a quantitative assessment of medullary perfusion and who applied these methods to assessing medullary perfusion directly.
The low medullary blood flow, which is required for effective countercurrent function, is not due to poorly developed vasculature, but rather is accounted for by the very high resistance to flow created by the length of the vasa recta. A low flow is critical; the osmolar concentration at the tip of the medulla is an inverse function of the flow rate when all other factors remain equal (Thurau & Levine 1971).
The concept of countercurrent exchange-in which fluid leaves the descending vasa recta and re-enters the ascending vasa recta-has received continuing theoretical attention and substantial experimental verification (Wirz & Dirix 1973). The renal medullary blood supply must provide more than the usual circulatory task of delivering substrate and oxygen for metabolism and removing metabolic products.
The ascending vasa recta must perform an additional task, that of removal of fluid entering the interstitial region from the collecting ducts during antidiuresis. While direct evidence that vasa recta actually serve this function is lacking, the need to assign such a role comes from an examination of the balance sheet (Marsh & Segel 1971).
In the steady state the amounts of fluid entering and leaving the region must be equal. The urine is concentrated by transfer of water from the collecting ducts to the interstitium. The descending loops of Henle and vasa recta face an osmolal gradient and thus lose water. The ascending limb of the loop neither gains nor loses water (Jamison 1970). Thus the result of all tubular water operation is the net deposition of water into the medulla.
The entire burden for removing reabsorbate is left, therefore, to the ascending vasa recta, which exceed the number of descending vasa recta by a ratio which approaches 2:1 (Marsh & Segel 1971). The medullary circulation is unique, therefore, in that venous outflow must exceed considerably the inflow of blood for extended periods during antidiuresis, a factor which must have contributed to the difficulties in measuring medullary perfusion.
Recently Sanjana et al. (1975) have measured the hydraulic and oncotic forces acting in the inner medulla of the mammalian kidney to define directly the factors determining net fluid reabsorption in the vasa recta, and the magnitude of the transcapillary forces involved.
Their findings suggested that net fluid reabsorption by the ascending vasa recta is governed by the same combination of forces that influence transcapillary fluid flux in other areas. The net inward driving force for fluid reabsorption is explicable on the basis of the transcapillary oncotic pressure difference, which exceeds the opposing hydraulic pressure difference.
As pointed out above, the medullary vascular bed is unique in that in no other vascular bed does venous outflow consistently exceed the inflow for extended periods. Thus techniques for measuring medullary perfusion based on an assessment of inflow, such as the technique used by Lillienfield (1961), which monitors the accumulation of tagged intravascular tracer with time, measure a different flow than those based on transit of intravascular indicator, such as tagged proteins (Kramer et al. 1960) or red cells (Wolgast 1968) which will also be influenced by the rate of outflow.
Measurement is further complicated by the fact that tubular flow in the descending loop of Henle and collecting ducts rivals blood flow quantitatively, so that delivery of tracer and removal of diffusible indicators by these routes must be quantitatively significant. This factor must have complicated measurements made, for example with radiorubidium, which presumably enters the medulla not only by its blood supply but also by way of tubular flow.
Note in Table 2.1 (Read: The Glomerular Regional Intrarenal Circulations) that the medullary blood flow measurement made with rubidium (Steiner & King 1970) exceeds considerably those made by other methods. Similarly diffusible indicators such as heat, hydrogen, krypton and xenon must both enter and leave the system via How of urine and tubular fluid.
Finally the countercurrent arrangement will both blunt the delivery of diffusible tracers and, when saturation is achieved, delay their outflow, through countercurrent trapping. These methods will, therefore, underestinuitc medullary blood flow. To complicate the assessment further there are independent vascular beds in the outer and inner medulla (Fourman & Moffat l97l).
Both derive their blood supply from the efferent arterioles of juxtamedullary glomeruli, but are anatomically and functionally independent. For all of these reasons there is continuing debate concerning the normal rate of medullary perfusion and how it changes in physiological and pathophysiological states.
In a recent review, Morel and de Rouffignac (1973) drew the pessimistic conclusion that the exact role of medullary blood 110W in relationship to and as a determinant of medullary function remained unclear despite considerable investigation. In 1960 Thurau et al. reported a striking reduction in the transit time of intravascular indicators through the medulla during a water diuresis, which was rapidly reversed with the administration of antidiuretic hormone.
A number of studies based on a wide variety of techniques have confirmed the original observation, demonstrating a reduction in medullary blood flow during hydropenia (Fourman & Moffat 1963, Fourman & Kennedy 1966, Harsing et al. 1966, Moffat 1968, Grangsjo 1968, Nissen 1968, Pinter 1969, Bencsath & Takacs 1971).
Both Fourman and Kennedy (1966) and Fisher et al. (1970) also reported an increase in medullary flow in animals with diabetes insipidus, which was reversed by antidiuretic hormone. Similarly, Cross et al. (1974) demonstrated a preferential reduction in papillary flow induced by antidiuretic hormone.
Thurau and Deetjen (1962) attributed the increase in urinary flow which accompanied a rise in arterial pressure to a measured increase in medullary blood flow, which reflected the failure of this portion of the renal vasculature to autoregulate blood flow effectively. Taken in all these observations are consistent, and have been widely reproduced.
Aukland (1968b), conversely, was unable to demonstrate with hydrogen electrodes either an increase in medullary flow during diuresis, or a decrease induced by antidiuretic hormone. Wolgast (1968) also failed to confirm an increase in medullary flow during diuresis with measurements of the intravascular transit of red cells in the medulla.
More recently, Persson at al. (1974) were able to demonstrate with the same method small, but statistically significant, increases in outer and inner medullary flow during a water diuresis and a decrease in outer medullary but not inner medullary llow during the administration of antidiuretic hormone. They concluded that the changes in medullary flow, while demonstrable, were too small to be physiologically significant.
Tanner and Selkurt (1970) demonstrated isosthenuria in the squirrel monkey after correction of haemorrhagic hypotension. They found a marked reduction in papillary osmolality which they attributed to washout of the medullary gradient. Medullary blood flow was restored but there was a persistent reduction in juxtamedullary nephron filtration following correction of the hypotension.
There has been controversy concerning the influence of haemorrhagic hypotension on blood How to the medulla. Kramer (1962) demonstrated a large reduction in medullary flow during haemorrhagic hypotension in the dog, the reduction in medullary flow equalling that in the cortex. This was supported by Aukland and Wolgast (1968) with measurements of local hydrogen gas clearance. Carriere et al. (1966) showed a striking reduction in cortical perfusion with krypton 85 washout curves and radioautography in dogs, but believed that medullary blood flow was not altered.
It seems likely that the latter methodology is suspect in a situation in which reduced cortical perfusion in some areas will result in flow components which make compartmental analysis especially unreliable as an index of flow in the medulla. Whereas medullary flow is reduced, it is better sustained than juxtamedullary nephron filtration rate, so that washout of the medullary gradient occurs during haemorrhagic hypotension as proposed by Tanner and Selkurt (1970).
Taken in all, these data indicate that medullary flow is considerably lower than that in the cortex and is modified in states of diuresis and antidiuresis. Whereas medullary perfusion is unlikely to be a determinant of medullary function in the normal state, it seems likely to play an important role in abnormal states, as exemplified by the effects of hypotension.
The characteristics of medullary perfusion create a number of consequences for local metabolism. First, a high vascular resistance results in perfusion rates rather lower per unit tissue mass than in most metabolically active tissues. Secondly. oxygen is lipid-soluble, thus tending to diti‘use directly from vessels entering the renal medulla to vessels leaving it.
Thirdly, there is sufficient plasma skimming that red cell delivery to the medulla is even lower than that of plasma (Ullrich et al. l96l, Wolgast 1973), further reducing oxygen delivery. The result is a precarious oxygen tension, an important element of anaerobiosis to support local medullary function, and a number of clinical syndromes characterized by papillary destruction-apparently due to inadequate medullary perfusion. These factors probably account for papillary necrosis in diabetes, sickle cell disease and phenacetin nephropathy.
It was the German nephrophiles, however, Kramer, Thurau and Deetjen who recognized the implications of the blood supply for the countercurrent mechanism, who developed the first methodology required to provide a quantitative assessment of medullary perfusion and who applied these methods to assessing medullary perfusion directly.
The low medullary blood flow, which is required for effective countercurrent function, is not due to poorly developed vasculature, but rather is accounted for by the very high resistance to flow created by the length of the vasa recta. A low flow is critical; the osmolar concentration at the tip of the medulla is an inverse function of the flow rate when all other factors remain equal (Thurau & Levine 1971).
The concept of countercurrent exchange-in which fluid leaves the descending vasa recta and re-enters the ascending vasa recta-has received continuing theoretical attention and substantial experimental verification (Wirz & Dirix 1973). The renal medullary blood supply must provide more than the usual circulatory task of delivering substrate and oxygen for metabolism and removing metabolic products.
The ascending vasa recta must perform an additional task, that of removal of fluid entering the interstitial region from the collecting ducts during antidiuresis. While direct evidence that vasa recta actually serve this function is lacking, the need to assign such a role comes from an examination of the balance sheet (Marsh & Segel 1971).
In the steady state the amounts of fluid entering and leaving the region must be equal. The urine is concentrated by transfer of water from the collecting ducts to the interstitium. The descending loops of Henle and vasa recta face an osmolal gradient and thus lose water. The ascending limb of the loop neither gains nor loses water (Jamison 1970). Thus the result of all tubular water operation is the net deposition of water into the medulla.
The entire burden for removing reabsorbate is left, therefore, to the ascending vasa recta, which exceed the number of descending vasa recta by a ratio which approaches 2:1 (Marsh & Segel 1971). The medullary circulation is unique, therefore, in that venous outflow must exceed considerably the inflow of blood for extended periods during antidiuresis, a factor which must have contributed to the difficulties in measuring medullary perfusion.
Recently Sanjana et al. (1975) have measured the hydraulic and oncotic forces acting in the inner medulla of the mammalian kidney to define directly the factors determining net fluid reabsorption in the vasa recta, and the magnitude of the transcapillary forces involved.
Their findings suggested that net fluid reabsorption by the ascending vasa recta is governed by the same combination of forces that influence transcapillary fluid flux in other areas. The net inward driving force for fluid reabsorption is explicable on the basis of the transcapillary oncotic pressure difference, which exceeds the opposing hydraulic pressure difference.
As pointed out above, the medullary vascular bed is unique in that in no other vascular bed does venous outflow consistently exceed the inflow for extended periods. Thus techniques for measuring medullary perfusion based on an assessment of inflow, such as the technique used by Lillienfield (1961), which monitors the accumulation of tagged intravascular tracer with time, measure a different flow than those based on transit of intravascular indicator, such as tagged proteins (Kramer et al. 1960) or red cells (Wolgast 1968) which will also be influenced by the rate of outflow.
Measurement is further complicated by the fact that tubular flow in the descending loop of Henle and collecting ducts rivals blood flow quantitatively, so that delivery of tracer and removal of diffusible indicators by these routes must be quantitatively significant. This factor must have complicated measurements made, for example with radiorubidium, which presumably enters the medulla not only by its blood supply but also by way of tubular flow.
Note in Table 2.1 (Read: The Glomerular Regional Intrarenal Circulations) that the medullary blood flow measurement made with rubidium (Steiner & King 1970) exceeds considerably those made by other methods. Similarly diffusible indicators such as heat, hydrogen, krypton and xenon must both enter and leave the system via How of urine and tubular fluid.
Finally the countercurrent arrangement will both blunt the delivery of diffusible tracers and, when saturation is achieved, delay their outflow, through countercurrent trapping. These methods will, therefore, underestinuitc medullary blood flow. To complicate the assessment further there are independent vascular beds in the outer and inner medulla (Fourman & Moffat l97l).
Both derive their blood supply from the efferent arterioles of juxtamedullary glomeruli, but are anatomically and functionally independent. For all of these reasons there is continuing debate concerning the normal rate of medullary perfusion and how it changes in physiological and pathophysiological states.
In a recent review, Morel and de Rouffignac (1973) drew the pessimistic conclusion that the exact role of medullary blood 110W in relationship to and as a determinant of medullary function remained unclear despite considerable investigation. In 1960 Thurau et al. reported a striking reduction in the transit time of intravascular indicators through the medulla during a water diuresis, which was rapidly reversed with the administration of antidiuretic hormone.
A number of studies based on a wide variety of techniques have confirmed the original observation, demonstrating a reduction in medullary blood flow during hydropenia (Fourman & Moffat 1963, Fourman & Kennedy 1966, Harsing et al. 1966, Moffat 1968, Grangsjo 1968, Nissen 1968, Pinter 1969, Bencsath & Takacs 1971).
Both Fourman and Kennedy (1966) and Fisher et al. (1970) also reported an increase in medullary flow in animals with diabetes insipidus, which was reversed by antidiuretic hormone. Similarly, Cross et al. (1974) demonstrated a preferential reduction in papillary flow induced by antidiuretic hormone.
Thurau and Deetjen (1962) attributed the increase in urinary flow which accompanied a rise in arterial pressure to a measured increase in medullary blood flow, which reflected the failure of this portion of the renal vasculature to autoregulate blood flow effectively. Taken in all these observations are consistent, and have been widely reproduced.
Aukland (1968b), conversely, was unable to demonstrate with hydrogen electrodes either an increase in medullary flow during diuresis, or a decrease induced by antidiuretic hormone. Wolgast (1968) also failed to confirm an increase in medullary flow during diuresis with measurements of the intravascular transit of red cells in the medulla.
More recently, Persson at al. (1974) were able to demonstrate with the same method small, but statistically significant, increases in outer and inner medullary flow during a water diuresis and a decrease in outer medullary but not inner medullary llow during the administration of antidiuretic hormone. They concluded that the changes in medullary flow, while demonstrable, were too small to be physiologically significant.
Tanner and Selkurt (1970) demonstrated isosthenuria in the squirrel monkey after correction of haemorrhagic hypotension. They found a marked reduction in papillary osmolality which they attributed to washout of the medullary gradient. Medullary blood flow was restored but there was a persistent reduction in juxtamedullary nephron filtration following correction of the hypotension.
There has been controversy concerning the influence of haemorrhagic hypotension on blood How to the medulla. Kramer (1962) demonstrated a large reduction in medullary flow during haemorrhagic hypotension in the dog, the reduction in medullary flow equalling that in the cortex. This was supported by Aukland and Wolgast (1968) with measurements of local hydrogen gas clearance. Carriere et al. (1966) showed a striking reduction in cortical perfusion with krypton 85 washout curves and radioautography in dogs, but believed that medullary blood flow was not altered.
It seems likely that the latter methodology is suspect in a situation in which reduced cortical perfusion in some areas will result in flow components which make compartmental analysis especially unreliable as an index of flow in the medulla. Whereas medullary flow is reduced, it is better sustained than juxtamedullary nephron filtration rate, so that washout of the medullary gradient occurs during haemorrhagic hypotension as proposed by Tanner and Selkurt (1970).
Taken in all, these data indicate that medullary flow is considerably lower than that in the cortex and is modified in states of diuresis and antidiuresis. Whereas medullary perfusion is unlikely to be a determinant of medullary function in the normal state, it seems likely to play an important role in abnormal states, as exemplified by the effects of hypotension.
The characteristics of medullary perfusion create a number of consequences for local metabolism. First, a high vascular resistance results in perfusion rates rather lower per unit tissue mass than in most metabolically active tissues. Secondly. oxygen is lipid-soluble, thus tending to diti‘use directly from vessels entering the renal medulla to vessels leaving it.
Thirdly, there is sufficient plasma skimming that red cell delivery to the medulla is even lower than that of plasma (Ullrich et al. l96l, Wolgast 1973), further reducing oxygen delivery. The result is a precarious oxygen tension, an important element of anaerobiosis to support local medullary function, and a number of clinical syndromes characterized by papillary destruction-apparently due to inadequate medullary perfusion. These factors probably account for papillary necrosis in diabetes, sickle cell disease and phenacetin nephropathy.