CHAPTER IX
Pathology of the Kidney in TraumaticShock
The kidney lesion associated with the clinicalsyndrome of renal insufficiency following resuscitation from shock is constantbut not pathognomonic. Lucké,1 in a study of materialfrom the Army Institute of Pathology, has shown that a similar lesion,which he has called "lower nephron nephrosis," develops in a variety ofconditions such as mismatched blood transfusion, infusion of human hemoglobin,the crush syndrome, heat stroke, blackwater fever, thermal burns, carbontetrachloride poisoning, mushroom poisoning (Amanita phalloides)and sulfonamide sensitivity. The author, from his own experience, can addto the list of etiologic agents chemical burns, anaphylactic shock followingtherapeutic use of antipneumococcus rabbit serum, and hemolytic reactionsfollowing transurethral resections in which the operative field is irrigatedwith tap water. Many years ago F. B. Mallory2 noted the lesionin an occasional patient with hemolytic streptococcus infection. Therecan be little doubt that with increasing breadth of experience many othercausal agents will be brought to light.
Descriptions of the lesion in various stages maybe found in innumerable case reports under such titles as transfusion kidney,3hemoglobinuric nephrosis,4 interstitial nephritis,5hepatorenal syndrome,6 and other terms, but surveys of
1LUCKÉ, B.: Lower nephron nephrosis (renal lesions of crush syndrome, of burns, transfusions, and other conditions affecting lower segments of nephrons). Mil. Surgeon 99: 371-396, November 1946.
2MALLORY, F. B.: Personal communication.
3DENAVASQUEZ, S.: Excretion of haemoglobin, with special reference to "transfusion" kidney. J. Path. & Bact. 51: 413-425, November 1940.
4MALLORY, T. B.: Hemoglobinuric nephrosis in traumatic shock. Am. J. Clin. Path. 17: 427-443, June 1947.
5KIMMELSTIEL, P.: Acute hematogenous interstitial nephritis. Am. J. Path. 14: 737-761, November 1938.
6HELWIG, F. C., and ORR, T. G.: Traumatic necrosis of liver with extensive retention of creatinine and high grade nephrosis. Arch. Surg. 24: 136-144, January 1932.
225
any considerable mass of material are few. Themost notable contributions are those of Lucké7 and ofBywaters and Dible8 in their studies of the crush syndrome.The present study is based primarily on a review of material from 60 outof 63 necropsies on patients studied by the Board, but is fortified byexperience gained from review of more than 150 similar nephropathies atthe 15th Medical General Laboratory. The records and material from the60 cases, correlated with the clinical and biochemic data, provided themeans for surveying the lesion in all stages of development as manifestedin wounded soldiers resuscitated from shock by human plasma or whole bloodtransfusions.
Microscopic Pathology
The characteristic features of the lesion emphasizedto a greater or lesser degree by different authors are seven: pigment casts;internal hydronephrosis; degeneration of epithelial elements, particularlyin the lower segments of the nephrons; rupture of tubules with extrusionof their content into the stroma; interstitial inflammation; granulomatousinflammation, and thrombophlebitis of small intrarenal veins. Let us examinethem successively.
Pigment Casts
Pigment casts, as in other forms of lower nephronnephrosis, are the most conspicuous feature of the "shock kidney." Theywere present in all cases of proved renal insufficiency in the presentseries in which the survival period was 3 days or longer. Their absencein many of the more acute cases is an interesting feature which will beconsidered later.
The pigment casts were found in the distal convolutedand collecting tubules and in the ducts of Bellini, their relative frequencyin these respective locations varying considerably from case to case. Thenumber of nephrons involved was difficult to estimate in routine sections,since only portions of each nephron were visualized. In most cases withclinically demonstrated renal insufficiency, the proportion was obviouslyhigh but the range was wide, probably from 20 to 80 percent.
7See footnote 1.
8BYWATERS, E. G. L., and DIBLE, J. H.: Renal lesion in traumatic anuria. J. Path. & Bact. 54: 111-120, January 1942.
226
The pigment, when viewed in unstained sectionsor sections only lightly stained with hematoxylin, shaded from orange throughorange-green to a muddy greenish-brown. It might precipitate as separatecoarse globules 4 to 8 microns in diameter (which are frequently mistakenon casual examination for red blood cells); as ropes of globules, ratherlike strings of sausages; or as granular debris usually embedded in a cast-likematrix. The staining reactions of the pigment were not consistent. A portionof it almost always showed the usual staining reactions of hemoglobin;namely, an affinity for eosin and for the fuchsin in the Masson trichrome,a dense blue color with phosphotungstic acid hematoxylin, and a positivebenzidine reaction. Another portion of the pigment, indistinguishable inthe unstained or eosin-stained preparations, failed to accept the fuchsinin the Masson or the hematoxylin in the phosphotungstic stain, and wasbenzidine-negative. Characteristic staining reactions with hematoxylinand eosin, phosphotungstic acid hematoxylin, and Masson trichrome are illustratedin Figures 1, 2, and 3.
Although there were some exceptions, two generalizationscould be made regarding the staining reactions of the pigment: First, theshorter the time interval between injury and death, the greater was theproportion of pigment which gave the staining reactions of hemoglobin.Second, the farther down the nephron the pigment was located, the lesswas this tendency. Figures 4 and 5 illustrate this difference in the pigmentcasts of the cortex and medulla. The obvious conclusion is that with thepassage of time and with the chemical changes in the glomerular filtrateas it passed along the tubule, the original hemoglobin-like material wasaltered, presumably by a process of degradation in the direction of acidhematin. This degradation never reached the point where iron could be demonstratedwith the ferrocyanide reaction. One other differential point has been noted:The hemoglobin-like pigment rarely excited a significant inflammatory reaction,whereas the altered pigment sometimes attracted leukocytes in considerablenumber (Fig. 6).
Internal Hydronephrosis
Many authors have described marked dilatationof portions of the nephron above the obstructing pigment casts. Their illustrationsoften show great dilatation of proximal tubules and of Bowman`s capsule.Our experience suggests that this apparent dilatation is largely a shrinkageartefact. If formalin-fixed kidney tissue is rapidly dehydrated and embeddedin paraffin, shrinkage is
227
difficult to avoid. The glomeruli appear as tinyspheres occupying less than half of the capsular space, and the diameterof the lumen of the proximal tubules is from 50 to 70 percent of the externaldiameter of the tubule. If the tissue is Zenkerized before embedding, theshrinkage is reduced and primary embedding in Zenker-formol or in Zenkerwith acetic acid lessens it still further. In such sections Bowman`s capsuleis rarely dilated, although some dilatation of tubules may be apparent.A still better check upon this debatable issue is provided by the examinationof frozen sections in which the factor of dehydration is excluded.
The following statements are based primarily uponthe examination of frozen sections which were prepared from almost everycase in this series. No dilatation of Bowman`s capsule was observed, buttubular dilatation of moderate but unmistakable degree was usual, withthe proximal convoluted tubule the segment most commonly involved. Dilatationwas not observed before the third day after injury, was almost constantbetween the third and fifth days, and became less frequent in the caseswith longer periods of survival. Dilatation of distal convoluted tubulesand upper collecting tubules, usually those situated in the cortical rays,was less common and was ordinarily seen in the more chronic stages fromthe sixth day onward. In some cases, however, dilatation was marked inthese segments and minimal or absent in the proximal tubules (Figs. 7 and8).
Dilatation of proximal tubules, to judge fromfrozen sections, correlated well with two characteristic phenomena notedon gross examination: enlargement of the kidney, particularly wideningof the cortex, and a wet surface wherever freshly cut.
Degeneration and Regeneration ofTubular Epithelium
Degenerative changes in the epithelial cells wererare in the proximal tubules, absent in the descending limb of Henle`sloop, usual in the ascending limb and the distal convoluted tubule, andinconstant in the collecting tubules. The proximal tubule was usually normalexcept for slight to moderate dilatation. The brush border was well maintained,the cells were not swollen, and the cytoplasmic granules were normal insize and distribution. The tubular lumens frequently contained granularnonpigmented precipitate, but this is not surprising in view of the constantalbuminuria demonstrable in patients with severe trauma and constitutesno proof of tubular injury.
In only five cases were the proximal tubules clearlyabnormal. In two there
228
STAINING REACTIONSOFPIGMENT CASTS
FIGURE 1.Stain: Hematoxylin and eosin. The glomerulus is normal. The distalconvoluted tubules contain orange
colored castsof precipitated hemoglobin. X 200
FIGURE 2.Stain: Masson trichrome. The casts stain deep red as do red cellswith this stain.
FIGURE 3.Stain: Phosphotungstic acid hematoxylin. The deep blue stainingof the pigment casts is characteristic of the
reactionsof hemoglobin with this stain. The proximal convoluted tubules are dilatedand contain precipitated albumin,
but the epithelialcells show no degenerative changes.
229
STAININGREACTIONS OF PIGMENTCASTS
230
was extensive fat vacuolation; in the two othersthe cells were swollen, the brush border was obscured, and the cytoplasmfilled with hyaline, presumably albuminous droplets. Nothing in the clinicalpicture served to distinguish these cases from others of the series. Theinfrequency of albuminous droplets in the cells of the proximal tubulesunder conditions of marked albuminuria might be interpreted as evidenceof failure of resorptive function in this segment of the nephron. Suchfunctional paralysis would also explain the surprising fact that only oncein 200 hemoglobinuric cases was absorption of hemoglobin by the proximaltubules observed. In one instance (Case 86) there were numerous foci ofcomplete coagulative necrosis of proximal tubular epithelium. The possibilityof direct traumatic injury of the kidneys could not be excluded here.
The descending limbs of Henle`s loops frequentlycontained considerable amounts of orange pigment, which was not iron-positivewith ferrocyanide but accepted the fuchsin counterstain as does so-calledhemofuchsin. It was sometimes weakly sudanophilic, and because of its orange-yellowcolor was difficult to distinguish from fat in Sudan preparations. Sincethis pigment was also frequently found in cases of sudden death used ascontrols, we believe it to be a normal phenomenon.
The earliest and most frequent evidence of tubulardegeneration appeared in the form of fat vacuolation in the ascending limbsof Henle`s loops. The vacuoles were extremely small, mostly from 1 to 2microns; were hard to identify in paraffin sections; were intensely redin frozen sections stained with Sudan IV; and were not doubly refractileunder polarized light. In some cases they were found only in a limitedsegment of the ascending limb, always the portion closest to the corticomedullaryjunction; in others the process extended almost to the apex of the pyramid,involving the entire ascending limb (Figs. 9 and 10). A rough quantitationof the severity of this change was attempted by visual estimation, usinga 1- to 4-plus scale.
Well-marked (2- to 4-plus) fat vacuolation waspresent in 6 of 8 cases in this study in which death followed injury byless than 48 hours, including Case 77 with a survival period of only 20hours. It was present, usually in 3- or 4-plus grade, in 14 of the 15 casesin which death occurred between the second and fourth days. From the fifthday onward it decreased in frequency and intensity; it was found in only14 of the 27 cases and in only 1 was it above 2-plus grade. A more extensivediscussion of this process will be found in Chapter XII in connection withsimilar observations on other organs.
231
INTERNAL HYDRONEPHROSIS
FIGURE 7.(Left) Moderate dilatation of both proximal and distal convolutedtubules is shown in a case of 6 days` duration.
A raresulfonamide crystal is present. Stain:Hematoxylin and eosin.
FIGURE 8.(Right) Marked dilatation of the distal convoluted tubules withcomparatively slight widening of the proximal
ones isapparent. Stain: Hematoxylin and eosin.
STAINING REACTIONS OF FAT
FIGURE10. (Right) Stain:Sudan IV. Diffuse fat vacuolation of entire ascendinglimb of Henle`s loop.
232
Occasionally on the third or fourth day, but alwaysby the fifth, other forms of epithelial reaction were obvious. The cellsbecame shrunken, their nuclei pyknotic, and they frequently lost theirattachment to the basement membrane and desquamated. At almost the sameperiod, mitoses became apparent and sometimes so numerous that two or threewere seen in a single cross-section of a tubule. These degenerative andregenerative changes were frequently limited to, and always most severein a comparatively limited zone close to the corticomedullary junction.
Distal Convoluted Tubules.-The distal convolutedtubules in apparently normal individuals frequently contain small granulesof lipid. In some cases of shock it was evident that the sudanophilic materialwas increased in amount, but variation in the normal range makes the significanceof this observation very difficult to assess. By the fourth or fifth dayafter injury, degenerative and regenerative changes sometimes became apparentin the distal tubule. They were never as marked or as widespread as inthe ascending limbs, but the difference was quantitative, not qualitative.
Collecting Tubules.-Alterations in thecharacter of the epithelium of the collecting tubules are difficult torecognize with certainty, since these tubules are particularly susceptibleto postmortem degeneration, including desquamation. Fat stains do not help,since lipoid degeneration rarely occurs. In well-fixed material, however,it was often evident that the cells were hyperchromatic. Frank necrosiswas rarely seen, yet mitoses were not infrequent. In other cases reduplicationof the epithelial layer was evident and sometimes, where no desquamationwas apparent, sheets of epithelial cells filled the lumina of the tubulesand encircled or even invaded the pigment casts. Sometimes mitoses werenumerous in these detached or semidetached cells. In the collecting tubules,in contrast to other segments of the nephron, the epithelial changes closelyparalelled the presence of pigment casts in the affected segments.
Interstitial Inflammation
Interstitial inflammation was not noted beforethe third day; from the third to the fourth day it was found in half thecases, and from the fifth day onward it was a constant phenomenon. It wasfirst manifested as focal infiltration limited to a fairly narrow zoneat the corticomedullary junction. The component cells were predominantlylymphocytes and plasma cells, but variable proportions of neutrophils andeosinophils were usually mingled with them (Fig. 11). The
233
INTERSTITIALINFLAMMATION
234
TUBULARRUPTURE
infiltrate tended to be most marked in the neighborhoodof the blood vessels and also to run in strands between the straight tubules.Edema, though usually present, was rarely a conspicuous feature (Fig. 12).As days passed, the zone of inflammatory infiltration widened and new fociappeared both in the cortex and the medulla. In the former, they tendedto be concentrated about the distal convoluted tubules. In a few of themost severe lesions, a clearly-defined second zone of inflammatory infiltrationwas noted in the outer millimeter of the cortex beneath the capsule (Fig.13).
Tubular Rupture
The distinctive phenomenon of tubular rupturewas generally apparent in any well-developed lesion from the fifth dayonward. It was usually associated with the presence of hyaline, non-pigmentedcasts in the affected segment, which was most commonly the ascending limbbut sometimes the distal convoluted tubule. These casts were chromophobicwith hematoxylin and eosin or phosphotungstic acid hematoxylin, but staineda fairly bright green in the Masson trichrome. In the affected segmentscontaining the casts, portions of the epithelial lining and of the underlyingbasement membrane disappeared
235
without trace as if by solution. The casts atthis point were frequently seen half within the tubular remnant and halfextruded into the stroma (Fig. 14).
Interstitial Granulomas
In the process of tubular rupture and cast extrusion,the advancing margin of the cast abutting on the stromal cells was quicklysurrounded by histiocytes. With further extrusion or with complete destructionof the tubular segment, the cast became completely surrounded by histiocyteswhich occasionally but rarely fused to form giant cells (Figs. 15 and 16).The histiocytes soon invaded the cast and appeared to break it up intocomparatively small fragments. It was possible to observe all stages fromthe original encapsulation to complete disintegration, leaving a focalaggregate of histiocytes with no trace of the original cast material (Fig.17).
Although this process unquestionably accountsfor many, perhaps for most, of the granulomatous lesions found in thesekidneys, the evidence is inadequate to prove that it was the only mechanismfor their formation. Granulomas were frequently found in the walls of veinsand occasionally in the capsule of the kidney, separated from the nearesttubule by apparently intact fibrous tissue of considerable density. Someof these lesions were quite similar in appearance to the Aschoff bodiesof rheumatic fever.
Hyaline masses, indistinguishable from the intratubularcasts by any staining method tried, were also occasionally found in thewalls of veins, even jutting into the lumen to become the basis for thrombusformation (Fig. 18). The question whether these represent extruded castswhich have wandered surprisingly from their original location or whichhave developed in situ in some entirely different manner cannot be answeredat present. It is of interest that on only one occasion in over 200 casesstudied did the author note extrusion of a pigment cast and consequentgranuloma formation.
Thrombophlebitis
In a considerable proportion of cases of 5 ormore days` duration, thrombi were readily found in the large but thin-walledveins of the corticomedullary junction. These thrombi were attached toa portion of the vein wall and they jutted into and narrowed but almostnever occluded the lumen. Beneath some points of attachment a mass of hyalinematerial as described in the preceding
236
INTERSTITIALGRANULOMAS
237
paragraph was found; beneath others were well-developedgranulomas with or without traces of hyaline material. The vein walls werefrequently invaded by lymphocytes, polymorphonuclear leukocytes, or eosinophils,apparently extending from focal concentrations of the interstitial infiltrate,such concentrations being characteristically clustered about these veins.
Summary of Histologic Findings inChronologic Sequence
The earliest change noted in the "shock kidney"was lipoid degeneration of the ascending limbs of the loops of Henle, whichusually made its appearance about 18 hours after injury, progressed inseverity until the third day, then tended to wane. Precipitation of thehemoglobin-like pigment in the distal convoluted and collecting tubuleswas infrequently observed short of 24 hours and was inconstant before 48hours. The pigment at first stained like hemoglobin but later lost thischaracteristic. Moderate dilatation of the proximal convoluted tubuleswas evident following the development of pigment casts.
By the third and fourth days, an interstitialinflammatory infiltrate and frank necrobiotic changes in the cells of theascending limbs were occasionally seen, and from the fifth day onward bothof these findings were constant. Rarely on the third or fourth day, butusually by the fifth, tubular rupture with extrusion of nonpigmented castswas apparent, and coincidentally stromal granulomas were found in moderatenumbers. From the fifth day onward, thrombophlebitis of the small veinswas found in an increasing proportion of the cases.
Gross Pathology
The kidney might or might not show gross abnormalities(Figs. 19, 20, 21, and 22). Enlargement was noted as early as the secondday but was not usual before the fourth. It tended to become more frequentand more extensive with duration of the lesion; the largest pair of kidneysobserved, 625 grams, was found in Case 93 with the unusual survival periodof 13 days. In patients with from 7- to 10-day survival periods, kidneysweighing up to 500 grams were usual (Figs. 20 and 21).
Enlarged kidneys were almost without exceptionpale, whereas kidneys of normal size, regardless of duration of the lesion,were usually congested. On section it became apparent that the enlargementwas chiefly due to swelling
238
GROSSAPPEARANCE OF THE CRUSHKIDNEY
239
of the cortex, which would measure from 9 to 12microns in thickness. The corticomedullary junction was unusually sharp,since the cortex was pale and the pyramids either unchanged or darker thannormal. A zone of intense congestion sometimes further intensified thedistinction. A constant finding was the unusual wetness of the freshlycut cortical surface.
Petechial hemorrhages beneath the pelvic mucosawere the only other gross manifestation which was at all frequent (Fig.20). They were commonest in cases of long duration with high grades ofazotemia and are reminiscent of similar hemorrhagic changes in advanced"decompensated" nephrosclerosis with renal insufficiency. They may be merelya phenomenon of the uremic state.
240
Pathogenesis
Shock
Two factors which are evidently of outstandingimportance in the development of the renal lesion may appear in severelytraumatized persons. One is pigment excretion in the urine and the otheris the physiologic state known as shock. A survey of the first hundredcases of pigment nephropathy studied at the 15th Medical General Laboratoryprovided interesting figures regarding the relative frequency of the lesionin different types of Army hospitals. In base hospitals (station and generalhospitals), nephropathy was found in only 6 percent of the necropsies,in evacuation hospitals the percentage rose to 18, and in the field hospitals(to which only the most severely wounded, nontransportable cases were admitted)to 30 percent. Clearly pigment nephropathy occurred in direct relationto severity of injury and, by implication, to the profundity of shock.
The cases studied by the Board offer more directevidence. Whenever possible, a clinician who was a member of the Boardmade the examination and recorded his impression regarding the presenceand severity of shock. When shock was recognized, it was classified asslight, moderate, or severe. In 183 cases so classified pigment nephrosiswas proved at necropsy in thirty-seven. (There were actually 38 provedcases in the Board`s entire series, but 1 had not been classified as todegree of shock.) Their distribution in the various shock categories isshown in Table 92.
The decreasing frequency of renal involvementfrom those with marked shock to those with lesser degrees of shock is apparent,but the six patients
TABLE 92.-RELATIONSHIPOF PROVED PIGMENT NEPHROPATHYTO DEGREE OF SHOCK
241
who were not considered to be in shock confusethe picture. This group, which consists of Cases 9, 22, 69, 70, 120, and132, is therefore considered in detail.
"No Shock" Group.-Case 9 is an exampleof a frank transfusion accident. Between 100 and 125 cc. of group A bloodhad been given in error to a group O recipient before a characteristicreaction developed and the transfusion was interrupted. Hemoglobinemiaand hemoglobinuria were demonstrated. This case belongs in a differentcategory and requires no further consideration.
The patient in Case 22 did not present evidenceof shock at the time of entry to the hospital 3 hours after wounding, orat any subsequent period of observation. He belonged to blood group A andreceived 2 units of blood, presumably group A, not from the theater bloodbank and not checked by any member of the Board. There was no clinicalevidence of reaction to these transfusions which were given on the fourthday of disease, only 2 days before his death. On the following day thenonprotein nitrogen level in the plasma was 294 mg. per 100 cc., so therenal lesion must have antedated these transfusions. The three early transfusionson the day of wounding were of the usual theater bank blood and were givenwithout incident. The Rh reaction of the patient was not determined. Noetiology for a hemolytic reaction was discovered, yet large amounts offree hemoglobin, 193 mg. per 100 cc., were found in the bladder urine atthe time of necropsy.
Cases 69, 70, and 132 are examples of the crushsyndrome with severe grades of myoglobinuria. Judged by the usual clinicalcriteria--blood pressure, pulse, pallor, coldness of extremities, anxiety--thesepatients did not impress the examiner as being in shock. Each of them,however, clearly exhibited one phenomenon which is emphasized by many studentsof shock: hemoconcentration, as shown by hematocrit values of 71, 80, and57, respectively. If shock itself was not present, there is no doubt thata closely allied physiologic abnormality was.
In Case 120 the patient was wounded in the rightthigh by a shell fragment which completely transected the femoral artery,vein, and nerve. Hemorrhage was severe and he was given 1,000 cc. of plasmaen route to the hospital. On arrival he seemed in good condition and wasoperated upon without further preparation, but 1,500 cc. of whole bloodwere given during the operation. Immediately after operation the bloodpressure was 110 mm. Hg systolic and 62 diastolic, but 7 hours later itdropped to 70/40 and throughout the day the systolic level ranged between60 and 70 millimeters of mercury. Nevertheless the patient`s color wasgood and his skin remained warm. The pathologist has
242
the temerity to suggest that in considerationof this case also shock can scarcely be excluded.
Pigment Excretion
In Chapter IV it was shown that all patientswith clinical evidence of renal involvement following traumatic injuryexcreted benzidine-positive pigment in the urine, although the amountsvaried widely. In excellent correlation with this observation is the histologicevidence of pigment precipitation in the lower nephron segments of allpatients with fatal nephropathy. These related phenomena deserve furtherconsideration.
The Nature of the Pigment
Numerous clinical studies9 10 11 haveestablished the two pigments found in the urine as hemoglobin and the closelyrelated myoglobin. In histologic sections they react identically with allmethods employed (the benzidine reaction, hematoxylin and eosin, Masson`strichrome and Mallory`s phosphotungstic acid hematoxylin stains). It hasgenerally been considered a safe inference that the pigment casts foundin the tubules were precipitates of the same material spectroscopicallyidentified in its soluble form in the urine, though the possibility thatthey might be oxidized or otherwise chemically altered forms could notbe excluded. Harrison et al.12 provided spectroscopic proofthat in the lesions produced in dogs by arsine hemolysis the casts consistedpredominantly of methemoglobin. No pigments such as hemochromagen or hematinwhich would be insoluble at the pH of urine were formed.
Efforts to identify myoglobin by chemical meansin the kidneys of patients who died of the crush syndrome were disappointinglyunsuccessful. (See Ap-
9BYWATERS, E. G. L.; DELORY, G. E.; RIMINGTON, C., and SMILES, J.: Myohaemoglobin in urine of air raid casualties with crushing injury. Biochem. J. 35: 1164-1168, November 1941.
10MAEGRAITH, B. G.; HAVARD, R. E., and PARSONS, D. S.: Renal syndrome of wide distribution induced possibly by renal anoxia. Lancet 2: 293-296, September 8, 1945.
11MINAMI, S.: Über Nierenveränderungen nach Verschüttung. Virchow`s Arch. f. path. Anat. 245: 246-267, 1923.
12HARRISON, H. E.; BUNTING, H.; ORDWAY, N. K., and ALBRINK, W. S.: The pathogenesis of renal injury produced in the dog by hemoglobin or methemoglobin. J. Exp. Med. 86: 339-356, October 1947.
243
pendix C for the method used.) The data are presentedin the following tabulation.
Extraction of Myoglobin fromKidneys
|
Type of case |
Case No. |
Gm. of myoglobin per Kg. of tissue |
|
Control (sudden death) |
A |
0.40 |
|
Shock, no pigment casts |
19 |
2.45 |
|
Shock, pigment nephropathy |
65 |
1.18 |
|
Crush syndrome |
69 |
0.45 |
It is obvious that the kidneys in the crush cases, which one might reasonably expect to be loaded with myoglobin, did not show significantly more than those in cases of sudden death selected as controls; actually they contained considerably less myoglobin than did the kidneys in shock cases with or without nephropathy in which the urine never contained myoglobin. It seems probable that this failure to recover myoglobin correlates with the fairly rapid alteration in the staining reactions of the pigment in the lower nephron segments already described and may be attributed to a chemical degradation. Under field conditions it was impractical to attempt further identification of these products.
Acidity of the Urine and Precipitationof Pigment in the Tubules
One widely accepted explanation of the developmentof a pigment nephropathy is that the pigment is precipitated from solutionwhen the glomerular filtrate becomes acid in reaction upon reaching thedistal convoluted tubule. It is of interest to compare the acidity of theinitial pigment-containing urines from patients in whom nephropathy wasproved with those from patients with no evidence of a renal lesion otherthan pigment excretion. A control group in which the urine was consistentlyfree from benzidine-positive material is also compared in Table 93 withthe two pigment-containing groups.
244
The ranges overlap widely, and the differencesin the means are slight and not significant. Acidity of the urine can beeliminated as the determining factor in precipitation of pigment in thelower nephron. With acidity eliminated, we are left without explanationfor the phenomenon of pigment precipitation. Presumably it is due to somechange in renal physiology, particularly prone to occur in, but not limitedto shock. Possibly it is nothing more remarkable than stagnation of theglomerular filtrate secondary to the renal ischemia which is known to bepresent. Possibly it is due to increased concentration of some unknown,perhaps abnormal, metabolic product as suggested by Oliver.13
TABLE 93.-URINEpH IN PIGMENT GROUPSAND CONTROL GROUP
One is tempted to speculate upon the possibilityof a reciprocal relationship between the degree of shock and pigment concentrationin relation to nephropathy: When pigment excretion is great, as in thecrush syndrome, some burns, and transfusion accidents, little or no shockis necessary; when shock is severe and prolonged, minimal pigment excretionmay be required. Many of our data would appear to support this concept,but the exceptions are too numerous and striking to permit such a conclusion.
Clinicopathologic Correlation
Any attempt to correlate form and function inthe kidney is hazardous in the extreme, but its fascination is irresistible.The first effect of shock upon the function of the kidney is oliguria,followed quickly by retention of metabolic products which are both nitrogenous,such as urea and creatinine, and inor-
13OLIVER, J.: New directions in renal morphology; method, its results and its future. Harvey Lect. (1944-1945) 40: 102-155, 1945.
245
ganic, such as phosphates. Both of these effectswere manifested within a few hours of the onset of shock, whereas the firstrecognizable morphologic change, the appearance of lipoid vacuoles in theascending loops, developed about 18 hours after onset, and pigment castsappeared still later. The initial renal insufficiency is, therefore, functionalrather than structural in basis. The studies of Cournand and his associates14haveestablished the early onset of renal ischemia in shock and adequately explainthe initial functional changes.
Ischemia with consequent anoxemia also servesas a logical explanation of the appearance of demonstrable lipid in theparenchymal cells. As will be shown in Chapter XII, a similar phenomenonin shock was found in the liver, the heart, and other organs in which therelationship of fatty degeneration to anoxemia has received more attention.Further knowledge of renal circulation and of the relative susceptibilityof various portions of the nephron to anoxemia is necessary to an understandingof the localization of the changes. The hypothesis of Trueta et al.15that blood is shunted from the cortex into the vasa recta of the pyramidsis not a satisfactory explanation, since the degenerative changes weremaximal in the very areas which, according to his theory, receive the largestquantities of blood.
The lipoid vacuolation became more intense onthe second and third days after the onset of shock. This, however, cannotbe interpreted as evidence that the initial causative factor necessarilypersisted. A form of "chain reaction" which required time for its completedevelopment may have been inaugurated by the initial insult.
An interesting feature of the development of the"shock kidney" is the apparent delay in the appearance of the pigmentedcasts. Pigment casts were present in significant numbers in only two ofthe nine Board cases in which death occurred within 48 hours of injury.Even in Cases 32 and 34, in which death occurred between the second andthird days, only traces of pigment could be found. It is obvious from examinationof the histories of this group that none of them, with the possible exceptionof one (Case 34), was ever adequately resuscitated from shock. In the caserecords of scores of other fatal shock cases
14COURNAND, A.; RILEY, R. L.; BRADLEY, S. E.; BREED, E. S.; NOBLE, R. P.; LAUSON, H. D.; GREGERSEN, M. I., and RICHARDS, D. W.: Studies of the circulation in clinical shock. Surgery 13: 964-995, June 1943.
15TRUETA, J.; BARCLAY, A. E.; DANIEL, P.; FRANKLIN, K. J., and PRICHARD, M. M. L.: Renal pathology in the light of recent neurovascular studies. [Preliminary communication.] Lancet 2: 237-238, August 17, 1946.
246
reviewed in the Pathology Section of the 15thMedical Laboratory as controls, unless there had been a transfusion accident,pigment casts were seldom recorded when death had occurred short of 24hours, although in one burn case they were numerous at 14 hours. They werenot seen with any regularity until from 36 to 48 hours had elapsed sinceinjury.
A simple explanation for this phenomenon whichinvokes no new mechanism would be the suppression of glomerular filtrationduring shock, either because of the depressed systemic blood pressure orbecause of specific renal ischemia. The presence of pigment casts, therefore,may logically be considered histologic evidence of glomerular function,either continued or at least temporarily resumed.
Functional Significance of the PigmentCasts
Although it is certain that the pigment castsplay no role in the initiation of renal insufficiency following shock,no such dogmatism is possible in defining their possible effect in thelater stages of the lesion. Controversy has hinged on the thesis that thecasts cause suppression of urine primarily by mechanical blockage of thetubules. The morphologic evidence for this is inconclusive. It is suggestedby the high proportion of apparently blocked tubules in many cases andby the dilatation of portions of the nephron proximal to the cast. In ourcases dilatation of moderate degree was a fairly constant phenomenon fromthe third day onward. It was not seen in the absence of casts, and thepossibility of a mechanical block cannot be denied. The number of pigmentedcasts varied widely from case to case, but their number was difficult toestimate in routine sections and we did not use serial sections or microdissectionto confirm our impressions of the proportion of occluded nephrons. It isdoubtful that they were numerous enough to explain the almost completeurinary suppression in certain cases. Experimental evidence that the glomeruliof nephrons whose tubules are plugged with pigmented casts are nonfunctionalwas provided by Harrison et al.16 in the study on dogs previouslycited.
It is interesting that in other forms of renal disease, chronic glomerular nephritis and lipoid nephrosis for instance, casts form in the nephrons but are readily swept in great numbers into the urine. In pigment nephropathy, al-
16See footnote 12.
247
though casts may regularly be found in the sediment,it is rare that large numbers of them are passed.
Tubular Degeneration
The initial evidence of tubular degeneration hasinvariably been fat vacuoles in the ascending limbs of Henle`s loops andless regularly in the distal convoluted tubules. This change evidentlyantedated and was independent of pigment precipitation, since it was frequentlyseen in the absence of the latter. By the fourth and fifth days lipid tendedto disappear from the tubules, but frank necrosis and regeneration of theepithelium became evident and affected the same segment of the nephron.These phenomena were never observed in the shock cases in the absence ofpigment precipitation, although in sulfonamide injury to the kidney theywere occasionally seen without pigment. Although the pigment deposit appearsto be a sine qua non, the mechanism of its effect is obviously indirect,since the maximal degenerative change occurred at a higher level in thenephron than did the first pigment deposits. Degeneration was most pronouncedin the ascending limbs, whereas pigment deposits began in the distal convolutedtubules.
The functional significance of this tubular degenerationis of interest. In lipoid nephrosis extensive tubular degeneration is associatedwith oliguria. It has been suggested that the injured tubular cells permitrediffusion of the glomerular filtrate back into the vascular system. Inpigment nephropathy tubular degeneration was constant, and such a mechanismmay be involved in this syndrome as well. One point of difference betweenthe two syndromes should be noted, however, before accepting the hypothesis.In lipoid nephrosis the gravity of the urine is well maintained or evenconcentrated above usual limits. In pigment nephropathy fixation at a lowlevel is almost constant. This could be readily explained by the theoryof mechanical blockage, since only a few unobstructed nephrons can stillfunction.
The focal rupture of tubules, so frequently notedfrom the fifth day onward, has, in the author`s opinion, little functionalsignificance. It is a late phenomenon, whereas the entire syndrome--oliguria,azotemia, pigment excretion, and hypertension, together with fixation ofgravity--is established before it appears. Furthermore, if actual extravasationof glomerular filtrate occurred through these ruptures, it seems inevitablethat interstitial edema would be a more conspicuous feature of the histologicprocess.
248
SUMMARY
A clinicopathologic survey of necropsy findingsin 60 severely wounded casualties provided 38 examples of lower nephronnephrosis in all stages of development. The initial change, appearing from18 to 24 hours after injury, was found to be lipoid degeneration of a segmentof the nephron, particularly the ascending limb of Henle`s loop, less markedlythe distal convoluted tubule. The second stage was the precipitation ofpigment, either hemoglobin or myoglobin, in the distal convoluted and collectingtubules, seldom occurring short of 24 hours after injury, more commonlybetween 32 and 72 hours. Moderate dilatation of proximal and sometimesdistal convoluted tubules followed pigment precipitation. Sometimes onthe third, regularly on the fourth and fifth days, necrosis and regenerationof epithelium in the ascending limbs and distal tubules became evident,and simultaneously lymphocytes appeared between the tubules and about thevessels. From five days onward, rupture of tubules with herniation of theircontents into the stroma and consequent granuloma formation became frequent.The last development in many cases was the formation of nonocclusive muralthrombi in many of the small, thin-walled veins.
In clinicopathologic correlation it was shownthat two factors were constant in the nephropathy cases, inconstant inthe remainder. These are (1) the excretion of benzidine-positive pigmentin the urine, and (2) (with one apparent exception) a state of shock or,at least, a related physiologic abnormality.
Renal insufficiency was found to antedate allstructural change, but was never progressive in the absence of a demonstrablepigment nephropathy. The effects of myoglobin and of hemoglobin precipitationupon the kidney were indistinguishable. Our evidence does not support thehypothesis that acidity of the urine is the factor responsible for pigmentprecipitation in the nephrons.
Although pigment precipitation is an essentialfactor in the development of the "shock kidney," the mechanism of its effectupon the kidney remains unexplained. The possibility that it is in partmechanical cannot be disregarded, but the complex chain of degenerativeand inflammatory phenomena demands another explanation. Controlled experimentationwill be necessary before many of these uncertainties can be resolved.
For Cases of Special Interest in this chapter,see page 282.
