JOURNAL
OF
SURGICAL
RESEARCH
24,
132- 141 (1978)
RESIDENT Distribution
of lntrarenal
JACK L. CRONENWETT, Department
of Surgepv. Section Resecirc-h Loboratory,
M.D., of General Vereruns
RESEARCH Blood
AWARD
Flow During
AND S. MARTIN University
Surgery. Administration
Bacterial
Sepsis
M.D.,
F.A.C.S.’
LINDENAUER, oj’,Michigan
Hospital,
Ann
Medical Arbor,
Center
Michigan
ond
Vasclrlnr
48109
Submitted for publication November 3, 1977
In contrast to hemorrhagic or cardiogenic shock, bacterial sepsis is often associated with hyperdynamic circulation, especially when hypotension is corrected by adequate volume administration [21]. Although renal blood flow (RBF) was historically felt to be depressed during sepsis, this confusion was due to inappropriate extrapolation from animal endotoxin shock models where the entire circulation was depressed [7]. In 1969, using a canine model of bacterial sepsis, Hermreck and Thal first documented significantly increased RBF in most dogs during sepsis [5]. Major contributions were then made by Lucas and associates, who extended this observation to patients in septic shock [II, 16, 181. These authors reported that septic patients often exhibit polyuria which persists despite hypovolemia and hypotension and is therefore inappropriate [I 1, 181. If unrecognized, this polyuria, combined with increased cardiac output, may be interpreted as evidence of volume overload and elicit the therapeutic response of fluid restriction. Unfortunately, such management often leads to renal failure and death, in contrast to patients in whom adequate volume is restored [ 181. Since these observations were reported, several attempts have been made to define the mechanism of inappropriate polyuria [2, 6, 151. The present study employed a canine bacterial sepsis model similar to that reported by Postel et ul. in which live Pseudomonas organisms were infused into I Reprint requests should be addressed to S. Martin Lindenauer. M.D.. B4962 CFOB, University Hospital, 1405 East Ann Street, Ann Arbor. Michigan 48104. 0022-4804/78/0243-O 132$0 I .00/O Copyright 0 1978 by Academic Press. Inc. All rights of reproduction in any form reserved.
awake dogs [ 141. Microspheres were used to investigate the possibility that changes in the intrarenal distribution of blood flow could account for the clinical observations noted above. METHODS
Fifteen
female
dogs (12-30
kg; mean,
19 2 5 kg) were allowed free access to food
and water until they were briefly anesthetized with intravenous thiamylal sodium (10 mg/kg). During this light anesthesia a transfemoral 7F side-hole cardiac catheter was positioned in the left ventricle by pressure recording. A transfemoral 18.5gauge polyethylene catheter was placed in the abdominal aorta near the level of the renal arteries. A similar catheter was placed in the brachial artery for pressure recording. A Swan-Ganz thermodilution catheter was inserted into the pulmonary artery via the external jugular vein and a Foley catheter was inserted into the bladder. The incisions for catheter placement were closed and the dogs were allowed to recover completely from the anesthesia. When the animals were completely awake they were placed in an upright Pavlov stand for the remainder of the experiment. Mean arterial pressure and rectal temperature were continuously recorded on a Hewlett-Packard Series 8800 recorder. Cardiac output was measured intermittently by thermodilution using an Edwards Model 9500 computer. Urine output was collected at 15-min intervals and analyzed for creatinine and sodium concentration. Simultaneous serum samples for
132
CRONENWETT
AND
LINDENAUER:
INTRARENAL
BLOOD
FLOW
133
sodium, creatinine, white blood cell count, tion. Quantification of the challenge inoculum and hematocrit were obtained. After the was performed by the following procedure: animals were stable in the Pavlov stand, A l-ml aliquot from the inoculant was baseline data were collected for 60 min. At serially diluted with sterile 0.9% sodium the end of this sampling period intrarenal chloride solution and duplicate 0.5-ml blood flow distribution and total RBF were aliquots from each dilution were spread onto measured using a microsphere technique. blood agar plates. The resulting places were Carbonized microspheres labeled with 141 incubated overnight at 37°C and the colonies Ce (15 + 5 pm diameter in 20% dextran) were counted. Bacteria were infused intrawere used (Nuclear Products, 3M Co.). venously for 90 min in each experiMicrospheres (400,000) were injected through mental (septic) dog. Blood cultures (5 ml) the left ventricular catheter and flushed with were taken immediately prior to termina20 ml of heparinized saline over 10 sec. A tion of the infusion and injected into 45 ml blood reference sample was withdrawn sim- of Bectin-Dickinson tripticase soy broth. ultaneously from the abdominal aortic A l-ml aliquot of each blood culture was catheter using a Harvard withdrawal pump serially diluted and duplicate 0.5-ml aliquots at the rate of 20 mumin for 1 min, be- from each dilution were spread onto blood ginning 5 set before microsphere injection. agar plates. The resulting plates were incubated overnight at 37°C and the colonies Following this determination, six control animals received an intravenous infusion were counted. of sterile BHI broth (Bacto brain-heart Immediately prior to termination of the infusion, Difco Laboratories, Detroit, Michbacterial infusion, intrarenal distribution of igan) at the rate of 0.5 ml/min for 90 min. blood flow was determined in septic dogs This infusate was verified sterile by bac- using *5Sr microsperes. Approximately 1 hr teriologic culture in each instance. Data later the animals were sacrificed and the were collected during this 90-min infusion kidneys were removed and fixed in 10% period as described above. Immediately formalin solution. During the entire experiprior to termination of the infusion, in- ment fluid administration was restricted to trarenal blood flow distribution was again 0.5 ml/min of D,W or broth infusion. measured using YSr-labeled microspheres. Total RBF was calculated by the reference Nine experimental dogs received an in- sample technique [ 1,3]. Both entire kidneys fusion of live Pseudomonas aeruginosa from each animal plus the aortic reference organisms originally isolated from a dog blood samples were counted using a sodium manifesting clinical sepsis at the School of iodide crystal scintillation detector. Samples Veterinary Medicine, Michigan State Uniwere placed into plastic containers and versity, East Lansing, Michigan and des- rotated to eliminate geometric variance. The ignated PSaEL. This organism was maindetails of this calculation have been pretained in BHI broth and stored at 4°C viously described [3]. between experiments. Initially a culture of Following isotope counting of the entire PSaEL grown in BHI broth for 24 hr at kidney, each kidney was divided in half 37°C was diluted with sterile 0.9% sodium coronally. Thin sections were taken in chloride solution to provide varying concencoronal fashion from three different areas of trations of bacteria to test specific in vivo cortex of each half kidney. These blocks reactions. A standard inoculant was then of cortex (approximately 4 mm thick x 10 selected which consisted of aliquots of un- mm wide) were then cut into four equal diluted 24-hr BHI broth cultures of PSaEL layers (from outer to inner cortex) and which had an average concentration of 4.8 labeled as Layers 1 through 4. It was (2 0.06) x lOa organisms/ml. These organ- found that equal layers could be cut by isms were used within 90 min of preparahand after familiarity was obtained with this
134
JOURNAL
OF SURGICAL
RESEARCH:
technique. The six samples of each layer from each kidney were then pooled in preweighed air-tight vials and subsequently weighed prior to counting. Samples from both kidneys from each dog were counted separately as an index of variance. The sample vials of each cortical layer were then counted for both isotopes simultaneously using a lithium-drifted germanium detector. Using previously described calculations [3, 121, computer analysis resolved microsphere distribution into the percentage distribution of blood flow to each of the four cortical layers for each isotope. Using the correction for ellipsoidal geometry of the kidney, this distribution was then converted to percentage distribution to each of four equal volume cortical zones, labeled Zones 1 through 4 from the outer to inner cortex [ 121. Zonal perfusion rate, or absolute blood flow to each zone, normalized for relative zone mass, was calculated [3]. The 30-min interval prior to microsphere injection was selected for comparison of hemodynamic, serum, and urine variables. Baseline and infusion values were compared using Student’s two-tailed t test for TABLE
VOL. 24, NO. 3, MARCH
1978
paired data. Control and septic animals were also compared using Student’s two-tailed t test for uncorrelated means. Kidney specimens from each animal were examined by light microscopy. RESULTS Quantitative blood cultures in septic dogs varied from 6.7 x lo4 to 1.2 x lo6 organisms/ml (3.7 % 3.5 x 105, X ? SD). Differences appeared to reflect individual dog variation, since the magnitude of bactermia was not correlated with changes in other measured variables. Cardiac index did not change significantly from baseline to infusion periods in either control or septic dogs (Table 1). Furthermore, no statistical difference was noted between control and septic animals during baseline or infusion. Mean arterial pressure did not change significantly in control animals. Septic animals, however, demonstrated significant hypotension during bacterial infusion, which was significantly lower than control animals (Table 1). Heart rate was statistically equivalent in control and 1 Baseline
Infusion
p Value
3.3 2 0.3 3.0 * 0.2
NS NS
Cardiac index ( l/min/m2)
Control Septic
3.6 2 0.2” 3.4 k 0.1
Mean arterial pressure (mm I-k)
Control Septic
140 2 5 138 2 5
Heart rate (bpm)
Control Septic
143 f 11 135 f 6
135 I?I 11 133 + 8
Rectal temperature 03
Control Septic
40.3 + 0.2 40.4 ?I 0.2
* I 40.3 + 0.2 41.1 k 0.3
Renal blood flowb (ml/min)
Control Septic
226 k 20 280 + 37
220 k 25 308 f 25
NS NS
Renal blood flow/per gram (ml/mm/g)
Control Septic
5.15 f 0.33 5.52 k 0.74
4.97 2 0.41 6.09 + 0.52
NS NS
a Mean b Renal *p < ** p
OF
SURGICAL
RESEARCH
24,
132- 141 (1978)
RESIDENT Distribution
of lntrarenal
JACK L. CRONENWETT, Department
of Surgepv. Section Resecirc-h Loboratory,
M.D., of General Vereruns
RESEARCH Blood
AWARD
Flow During
AND S. MARTIN University
Surgery. Administration
Bacterial
Sepsis
M.D.,
F.A.C.S.’
LINDENAUER, oj’,Michigan
Hospital,
Ann
Medical Arbor,
Center
Michigan
ond
Vasclrlnr
48109
Submitted for publication November 3, 1977
In contrast to hemorrhagic or cardiogenic shock, bacterial sepsis is often associated with hyperdynamic circulation, especially when hypotension is corrected by adequate volume administration [21]. Although renal blood flow (RBF) was historically felt to be depressed during sepsis, this confusion was due to inappropriate extrapolation from animal endotoxin shock models where the entire circulation was depressed [7]. In 1969, using a canine model of bacterial sepsis, Hermreck and Thal first documented significantly increased RBF in most dogs during sepsis [5]. Major contributions were then made by Lucas and associates, who extended this observation to patients in septic shock [II, 16, 181. These authors reported that septic patients often exhibit polyuria which persists despite hypovolemia and hypotension and is therefore inappropriate [I 1, 181. If unrecognized, this polyuria, combined with increased cardiac output, may be interpreted as evidence of volume overload and elicit the therapeutic response of fluid restriction. Unfortunately, such management often leads to renal failure and death, in contrast to patients in whom adequate volume is restored [ 181. Since these observations were reported, several attempts have been made to define the mechanism of inappropriate polyuria [2, 6, 151. The present study employed a canine bacterial sepsis model similar to that reported by Postel et ul. in which live Pseudomonas organisms were infused into I Reprint requests should be addressed to S. Martin Lindenauer. M.D.. B4962 CFOB, University Hospital, 1405 East Ann Street, Ann Arbor. Michigan 48104. 0022-4804/78/0243-O 132$0 I .00/O Copyright 0 1978 by Academic Press. Inc. All rights of reproduction in any form reserved.
awake dogs [ 141. Microspheres were used to investigate the possibility that changes in the intrarenal distribution of blood flow could account for the clinical observations noted above. METHODS
Fifteen
female
dogs (12-30
kg; mean,
19 2 5 kg) were allowed free access to food
and water until they were briefly anesthetized with intravenous thiamylal sodium (10 mg/kg). During this light anesthesia a transfemoral 7F side-hole cardiac catheter was positioned in the left ventricle by pressure recording. A transfemoral 18.5gauge polyethylene catheter was placed in the abdominal aorta near the level of the renal arteries. A similar catheter was placed in the brachial artery for pressure recording. A Swan-Ganz thermodilution catheter was inserted into the pulmonary artery via the external jugular vein and a Foley catheter was inserted into the bladder. The incisions for catheter placement were closed and the dogs were allowed to recover completely from the anesthesia. When the animals were completely awake they were placed in an upright Pavlov stand for the remainder of the experiment. Mean arterial pressure and rectal temperature were continuously recorded on a Hewlett-Packard Series 8800 recorder. Cardiac output was measured intermittently by thermodilution using an Edwards Model 9500 computer. Urine output was collected at 15-min intervals and analyzed for creatinine and sodium concentration. Simultaneous serum samples for
132
CRONENWETT
AND
LINDENAUER:
INTRARENAL
BLOOD
FLOW
133
sodium, creatinine, white blood cell count, tion. Quantification of the challenge inoculum and hematocrit were obtained. After the was performed by the following procedure: animals were stable in the Pavlov stand, A l-ml aliquot from the inoculant was baseline data were collected for 60 min. At serially diluted with sterile 0.9% sodium the end of this sampling period intrarenal chloride solution and duplicate 0.5-ml blood flow distribution and total RBF were aliquots from each dilution were spread onto measured using a microsphere technique. blood agar plates. The resulting places were Carbonized microspheres labeled with 141 incubated overnight at 37°C and the colonies Ce (15 + 5 pm diameter in 20% dextran) were counted. Bacteria were infused intrawere used (Nuclear Products, 3M Co.). venously for 90 min in each experiMicrospheres (400,000) were injected through mental (septic) dog. Blood cultures (5 ml) the left ventricular catheter and flushed with were taken immediately prior to termina20 ml of heparinized saline over 10 sec. A tion of the infusion and injected into 45 ml blood reference sample was withdrawn sim- of Bectin-Dickinson tripticase soy broth. ultaneously from the abdominal aortic A l-ml aliquot of each blood culture was catheter using a Harvard withdrawal pump serially diluted and duplicate 0.5-ml aliquots at the rate of 20 mumin for 1 min, be- from each dilution were spread onto blood ginning 5 set before microsphere injection. agar plates. The resulting plates were incubated overnight at 37°C and the colonies Following this determination, six control animals received an intravenous infusion were counted. of sterile BHI broth (Bacto brain-heart Immediately prior to termination of the infusion, Difco Laboratories, Detroit, Michbacterial infusion, intrarenal distribution of igan) at the rate of 0.5 ml/min for 90 min. blood flow was determined in septic dogs This infusate was verified sterile by bac- using *5Sr microsperes. Approximately 1 hr teriologic culture in each instance. Data later the animals were sacrificed and the were collected during this 90-min infusion kidneys were removed and fixed in 10% period as described above. Immediately formalin solution. During the entire experiprior to termination of the infusion, in- ment fluid administration was restricted to trarenal blood flow distribution was again 0.5 ml/min of D,W or broth infusion. measured using YSr-labeled microspheres. Total RBF was calculated by the reference Nine experimental dogs received an in- sample technique [ 1,3]. Both entire kidneys fusion of live Pseudomonas aeruginosa from each animal plus the aortic reference organisms originally isolated from a dog blood samples were counted using a sodium manifesting clinical sepsis at the School of iodide crystal scintillation detector. Samples Veterinary Medicine, Michigan State Uniwere placed into plastic containers and versity, East Lansing, Michigan and des- rotated to eliminate geometric variance. The ignated PSaEL. This organism was maindetails of this calculation have been pretained in BHI broth and stored at 4°C viously described [3]. between experiments. Initially a culture of Following isotope counting of the entire PSaEL grown in BHI broth for 24 hr at kidney, each kidney was divided in half 37°C was diluted with sterile 0.9% sodium coronally. Thin sections were taken in chloride solution to provide varying concencoronal fashion from three different areas of trations of bacteria to test specific in vivo cortex of each half kidney. These blocks reactions. A standard inoculant was then of cortex (approximately 4 mm thick x 10 selected which consisted of aliquots of un- mm wide) were then cut into four equal diluted 24-hr BHI broth cultures of PSaEL layers (from outer to inner cortex) and which had an average concentration of 4.8 labeled as Layers 1 through 4. It was (2 0.06) x lOa organisms/ml. These organ- found that equal layers could be cut by isms were used within 90 min of preparahand after familiarity was obtained with this
134
JOURNAL
OF SURGICAL
RESEARCH:
technique. The six samples of each layer from each kidney were then pooled in preweighed air-tight vials and subsequently weighed prior to counting. Samples from both kidneys from each dog were counted separately as an index of variance. The sample vials of each cortical layer were then counted for both isotopes simultaneously using a lithium-drifted germanium detector. Using previously described calculations [3, 121, computer analysis resolved microsphere distribution into the percentage distribution of blood flow to each of the four cortical layers for each isotope. Using the correction for ellipsoidal geometry of the kidney, this distribution was then converted to percentage distribution to each of four equal volume cortical zones, labeled Zones 1 through 4 from the outer to inner cortex [ 121. Zonal perfusion rate, or absolute blood flow to each zone, normalized for relative zone mass, was calculated [3]. The 30-min interval prior to microsphere injection was selected for comparison of hemodynamic, serum, and urine variables. Baseline and infusion values were compared using Student’s two-tailed t test for TABLE
VOL. 24, NO. 3, MARCH
1978
paired data. Control and septic animals were also compared using Student’s two-tailed t test for uncorrelated means. Kidney specimens from each animal were examined by light microscopy. RESULTS Quantitative blood cultures in septic dogs varied from 6.7 x lo4 to 1.2 x lo6 organisms/ml (3.7 % 3.5 x 105, X ? SD). Differences appeared to reflect individual dog variation, since the magnitude of bactermia was not correlated with changes in other measured variables. Cardiac index did not change significantly from baseline to infusion periods in either control or septic dogs (Table 1). Furthermore, no statistical difference was noted between control and septic animals during baseline or infusion. Mean arterial pressure did not change significantly in control animals. Septic animals, however, demonstrated significant hypotension during bacterial infusion, which was significantly lower than control animals (Table 1). Heart rate was statistically equivalent in control and 1 Baseline
Infusion
p Value
3.3 2 0.3 3.0 * 0.2
NS NS
Cardiac index ( l/min/m2)
Control Septic
3.6 2 0.2” 3.4 k 0.1
Mean arterial pressure (mm I-k)
Control Septic
140 2 5 138 2 5
Heart rate (bpm)
Control Septic
143 f 11 135 f 6
135 I?I 11 133 + 8
Rectal temperature 03
Control Septic
40.3 + 0.2 40.4 ?I 0.2
* I 40.3 + 0.2 41.1 k 0.3
Renal blood flowb (ml/min)
Control Septic
226 k 20 280 + 37
220 k 25 308 f 25
NS NS
Renal blood flow/per gram (ml/mm/g)
Control Septic
5.15 f 0.33 5.52 k 0.74
4.97 2 0.41 6.09 + 0.52
NS NS
a Mean b Renal *p < ** p
