by a Clearance Technic *

Transcription

1 Journal of Clinical nvestigation Vol. 5, No. 1, 1966 Gastric Secretion in Relation to Mucosal Blood Flo Studied by a Clearance Technic * UGN D. JACOBSON,t RAY H. LNFORD, AND MORTON. GROSSMAN (From the Veterans Administration Center and the Department of Physiology, University of California at Los Angeles Center for the Health Sciences, Los Angeles, Calif.) There is a considerable body of evidence suggesting that secretion and blood flo in the stomach vary together (1-6) and an equally impressive array of studies that speaks against such a relationship (7-11). n attempting to evaluate a possible relationship beteen gastric blood flo and secretion e are confronted ith problems that are basically methodological (12). Nearly all of the investigations cited have been either acute experiments on anesthetied and traumatied animals or chronic observations of unanesthetied humans in hich nonspecific gastric mucosal changes have been used to gauge gastric blood flo. n the acute experiments both blood flo and secretion are subject to the profound but unmeasured influences of anesthetics and surgical stress, hich can camouflage a relationship. n the chronic studies the color or temperature of the mucosa or the sie of microcirculatory vessels has been used as an "index" of mucosal blood flo. Unfortunately these indexes do not measure blood flo: mucosal color and vessel sie are estimates of mucosal blood volume, and mucosal temperature reflects deep body temperature, not blood flo. n short, the available evidence on the subject of a relationship beteen blood flo and secretion in the stomach comes from experiments in hich either the state of the preparation or the methods of measuring blood flo as unreliable. We established four requirements that e felt ould have to be satisfied for a study of gastric secretory and circulatory function to yield infor- * Submitted for publication April 1, 1965; accepted September 16, A partial report of this ork as read before the Western Section Meeting of the American Federation for Clinical Research, January 29, t Address requests for reprints to Dr. ugene D. Jacob- ;on, Veterans Administration Center, Wilshire and Satelle Blvds., Bldg. 11, Room 231, Los Angeles, Calif mation that ould not be subject to the preceding criticisms: 1) the animal must be unanesthetied and eeks past surgery; 2) both gastric secretion and blood flo must be measured simultaneously, and the measurement of one function must not interfere ith measurement of the other; 3) frequent or continuous measurements of blood flo and secretion must be obtained for considerable periods to obviate the effect of normal fluctuations; and ) the method of measuring blood flo must quantify gastric mucosal flo, since this can vary independently of total blood flo to the stomach. For these reasons e selected a gastric clearance technic first described by Shore, Brodie, and Hogben (13). The method is based upon the selective permeability of lipoidal membranes to the undissociated form of compounds. Since there is a sieable ph gradient from plasma to gastric juice, eakly basic drugs ith a suitable pka (5 to 1) ill be transported from the plasma and ill accumulate in the secretions of the stomach. The theoretical ratio (R) of the concentration of these drugs in gastric juice/plasma, if e assume no limitation on the rate at hich the drug is delivered to the membrane by the circulation, can be calculated from the equation, R 1 + JOPKa-gastric juice ph 1 + lopka-plasma ph For eak bases ith pka less than 2, the theoretical R (diffusion limited only) as identical ith the observed R. Hoever, for aminopyrine (pka = 5) the theoretical R ould be approximately 1, hereas the observed R as, suggesting that blood flo is not infinitely fast and that it limits the transit of aminopyrine by the rate at hich it delivers the drug to the membrane (13). f blood flo is a factor that limits R, then aminopyrine is suited as a clearance marker for estima-

2 2. D. JACOBSON, R. H. LNFORD, AND M.. GROSSMAN tion of mucosal blood flo and should meet the requirements of the Fick principle. extracted, 5 to g per hour; bethanechol (Urecholine),2 laboratory (16) and expressed as et eight of mucosa A basic assumption of the use of this clearance.5 mg per hour; vasopressin (Pitressin),3 U per hour; epinephrine hydrochloride (expressed as the base), technic for estimating blood flo is that the mucosa.25 to 2.5,ug per kg per minute; and secretin, completely clears the drug from the circulation on 76 U injected every 15 minutes intravenously. The dosages of the last four agents ere chosen to induce inhibi- a single passage through that tissue hich is exposed to acid gastric juice. This assumption cannot be tested directly, since venous blood from the tion of secretion ithout also producing undesired symptoms (collapse, retching, excitement). Chemical determinations. Gastric juice samples from stomach drains nonmucosal structures as ell as the pouch ere collected every 15 minutes, and the volume as recorded and the concentration of HC1 deter- the mucosa. Hoever, as ill be developed in this presentation, several considerations make this mined by titration ith.2 N NaOH to ph 7. ith a Radiometer autoburette. Six-ml brachial venous blood sam- assumption acceptable. ples ere aspirated into oxalated tubes at hourly intervals throughout each experiment and centrifuged; the plasma Methods as preserved. The concentration of aminopyrine in Unanesthetied dogs. Vagally denervated gastric fundic (Heidenhain) pouches ere prepared in 12 mongrel the method of Brodie and Axelrod (17). This technic each gastric juice and plasma sample as determined by dogs of both sexes eighing from 13 to 17 kg. No experiments ere performed during the first 3 eeks after ple, extracting aminopyrine into dichloroethane solvent, consists essentially of alkaliniing a 1-ml biological sam- surgery. Frequent hematocrit determinations ere obtained in each dog because of the depressive effects of back into an aqueous acid phase. The concentration of ashing ith sodium borate, and extracting aminopyrine aminopyrine on canine bone marro (pancytopenia) drug is then determined by its absorption at 56 mtu in a (1). To of the animals developed a profound chronic Zeiss spectrophotometer. n our hands the limit of detection of aminopyrine as 1 og in a 1-ml sample of anemia after their initial exposure to aminopyrine, and these dogs ere not used further. The other ten animals shoed no significant alteration in hematocrit detive duplicate samples from our experiments, the stand- either gastric juice or plasma. n 3 pairs of consecuspite administration of as much as mg of aminopyrine ard deviation of the mean of paired samples as.6,ug per kg body eight during each of five separate experiments spaced over 3 months. per ml.5 These results compare favorably ith the limits of the method as originally described by Brodie and A typical experiment consisted of placing the dogs in Axelrod (17). This method measures not only aminopyrine but also one of its metabolites, -amino anti- a sling frame and starting a constant infusion of saline into a hind leg vein at a rate of.5 ml per minute by pyrine. n our experience, hoever, the concentration means of a peristaltic pump.' A loading dose (2 mg ratio of substances in gastric juice and plasma measured per kg) of aminopyrine as infused over 1 minutes, by this method remained remarkably constant under constant experimental conditions. Furthermore, Shore and and then a maintenance dose (5 mg per kg per hour) as continued throughout the experiment to stabilie co-orkers (13) measured aminopyrine concentrations in plasma aminopyrine concentrations. After 3 minutes, gastric juice and plasma under conditions similar to our during hich the pouch as not secreting spontaneously, experiments and observed an R for aminopyrine alone a small amount of histamine (.1 mg base per hour) or that as nearly the same as the R e found. porcine gastrin (5 g per hour) as added to the infused Our recoveries of aminopyrine from canine plasma material to initiate a submaximal secretory rate. After and gastric juice ere not significantly different from 1 hour the dose of the stimulant as doubled, folloed by recoveries from distilled ater (over 97%9), again agreeing ith the original report for this method (17). successive doublings of the dose each hour until the secretory rate no longer increased ith increments in Brodie and Axelrod also found that hole blood aminopyrine concentrations ere 85%o of the plasma value. drug dosage (maximal response). At this point bethanechol (.5 mg) as infused for 1 hour, along ith the The clearance, hich can be assumed to represent maximal dose of histamine or gastrin, to induce a supramaximal secretory response (15). n some experiments a relationship, C = GV/P, here C is the clearance (mil- pouch mucosal plasma flo, as calculated from the steady submaximal secretory response as maintained liliters per minute), G and P are gastric juice and plasma ith histamine or gastrin, and a secretory inhibitory 2 Merck, Sharp and Dohme, Philadelphia, Pa. agent (vasopressin, epinephrine, isoproterenol, or secretin) 3 Parke, Davis, Detroit, Mich. as added to observe the effect of these agents on both Vitrum, Stockholm, Seden. secretion and clearance. The dosages of the various agents infused ere as follos: histamine phosphate (expressed as histamine base), teen each duplicate pair of samples and n is the number 5 Calculated from -_f 7Yn, here d is the difference be-.1 to 3.2 mg per hour; porcine gastrin, prepared in this of pairs (18). The formulae for the coefficient of correlation and tests of significance are to be found in 1 Harvard Apparatus Co., Dover, Mass. Snedecor's text (19).

3 GASTRC MUCOSAL BLOOD FLOW MASURD BY AMNOPYRN CLARANC 3 aminopyrine concentration (milligrams per milliliter), and V is the gastric secretory rate (milliliters per minute). The ratio of G/P, hich is designated as R, is of more than passing interest (13, 2, 21). t expresses the limitation imposed by the circulation on the transit of eak base across the gastric mucosal membrane. n other ords, R tells us ho much plasma flo in the mucosa occurred per unit of secretory volume. For these reasons both clearance and R ill be compared ith gastric secretory rate in our results. Results The gastric clearance technic Studies in anesthetied dogs. n to acute experiments on anesthetied dogs e found that there as no significant arteriovenous concentration difference for aminopyrine in the folloing circulatory areas: small intestine, brain and head, kidney, heart and lungs, and forelimb. By contrast, in seven other acute experiments the stomach cleared 58% (range, to 75%) of the aminopyrine from the blood presented to it. These determinations ere made at the end of 3 minutes stimulation ith histamine. We compared the amount of aminopyrine secreted by the total stomach in 1 hour ith the amount of drug cleared simultaneously from the blood perfusing the stomach (product of stomach blood flo and arteriovenous concentration differ- HSTAMN.8mg/hr AMNOPYRN m 75mg/hr - ZroX = DOG NO X = DOG NO.51 X SCRTORY VOLUM Cr- 2 CONCNTRATN OF >_ W ACD CONCNTRATON. 15 ~~ W.) d 2 5< C)5 ~~ ~~ 2 3 TM (hours) FG. 1. RSULTS FROM TWO XPRMNTS SHOWNG THAT AMNOPYRN AS USD N THS STUDY HAD NO F- FCT ON THR TH VOLUM RAT OF SCRTON OR TH CONCNTRATON OF HCL N GASTRC JUC. TABL ffect of pouch solution ph and histamine on the amount of aminopyrine collected from the pouch ph of recovered solution Aminopyrine recovered in 3 minutes mg i.8 (S) i (histamine) 1.1 ±t 1.9 ence). This as performed during stimulation ith histamine in to acute experiments on anesthetied dogs ith isolated gastric circulation (22). n each experiment over 9%o of the extracted aminopyrine as recovered in the gastric juice. Studies in unanesthetied dogs. Aminopyrine as used in our experiments did not affect gastric secretion. n to dogs (Figure 1) a stable secretory rate as established ith histamine, and the infusion of aminopyrine did not alter either the volume rate of secretion or the acid concentration of gastric juice. When a constant stimulus to gastric secretion as used and a constant amount of aminopyrine infused to maintain stable plasma concentrations, e observed no significant changes in R values over many hours. This relative constancy of R as noted not only hen gastric secretions ere alloed to drain freely to the exterior but also hen pouch secretions ere kept in constant contact ith the mucosa. This steady state for R suggested that reabsorption of aminopyrine from the pouch must be minimal. This possibility as investigated in 22 experimental periods of 3 minutes each in three different animals. We kept a 5-ml solution containing 1 mg per ml aminopyrine in.1 N HCO (this corresponds to the concentrations of aminopyrine and HCO secreted during our experiments) in contact ith the pouch mucosa during each of the experimental periods. Recovery of aminopyrine from pouch solutions in these 22 periods as %o (S) of that instilled, hich as not significantly different from 1% (p >.1). The effect of varying the ph of the pouch solution on aminopyrine clearance as examined during seven experiments on six animals. A constant dose of aminopyrine as administered throughout each experiment, and the different solutions ere

4 . D. JACOBSON, R. H. LNFORD, AND M.. GROSSMAN 16 C *' 1 to -L J 8 6 (1) (1) 2 CD.S R 1 l-. C. 8 r W 6 -i (So -J a. 2 v 3 6 TM (hours) FG. 2. FFCTS OF HSTAMN N PROGRSSVLY NCRASNG DOSS AND URCHOLN ON GASTRC SCR- TORY RAT, MUCOSAL CLARANC, AND R N ON XPRMNT. R = the ratio of gastric juice/plasma aminopyrine concentrations. kept in contact ith the pouch mucosa by means of a reservoir. t can be seen from the results of these experiments in Table that there as little difference beteen the amount of aminopyrine crossing into the pouch filled ith solutions at lo ph (1. or.9), but the alkaline solution (ph 8.) contained very little drug. By contrast, histamine induced a marked increase in aminopyrine clearance despite the fact that the ph of the secreted juice as also.9. Apparently the amount of aminopyrine accumulating in the pouch depends upon at least to major factors: the presence of an acid gradient from plasma to gastric juice, and a mucosal factor, about hich e shall present evidence and hich e believe is the rate at hich the gastric mucosal circulation can deliver aminopyrine to the membrane for transport. An effort as made to maintain plasma aminopyrine concentrations at a steady level in each experiment. n a fe experiments, hoever, considerable fluctuation in plasma concentrations as noted, and this afforded us the opportunity to observe its effect on R. n to animals plasma aminopyrine concentrations varied 3 and 53%, respectively, from initial values during each experiment; R values in these same studies varied 12 and 15%o, and there as no directional correlation beteen changes in concentration and changes in R. Furthermore, in the many experiments reported in our study, although plasma aminopyrine concentrations varied beteen experiments over the range 15 to 53 ug per ml, at maximal secretory rates nearly all R values fell in the range 3 to. These findings suggest that in our experiments maximal transport (Tm) as not reached for the secretion of aminopyrine by the stomach, since increments in the concentration beyond that required to achieve Tm ould be associated ith a progressive decline in R. Finally, the constancy of R during brisk secretion in the face of idely

5 GASTRC MUCOSAL BLOOD FLOW MASURD BY AMNOPYRN CLARANC fluctuating plasma aminopyrine concentrations indicates a fairly constant extraction ratio across the gastric mucosa. The relationship beteen secretion and blood flo Histamine. Dose-response experiments ith histamine demonstrated the anticipated effect on gastric secretion in six dogs ith denervated pouches. Successive doubling of the dose infused caused a corresponding increment in the secretory rate that could be increased further by the addition of bethanechol (Figure 2). Coincident ith the increase in gastric secretory rate, there as an increase in the clearance value. The ratio (R) of gastric juice to plasma aminopyrine concentrations exhibited a decline from high values at lo secretory rates to a steady plateau value of 3 to at near maximal or supramaximal secretory rates. c at cn -J a. 2 r F 12F lop of so am am Om * * *s as 6 so GASTRC SCRTORY VOLUM RAT (ml/15min) FG. 3. PLOT OF SCRTORY RAT AGANST CLARANC FOR SX XPRMNTS N WHCH HSTAMN WAS M- PLOYD. a 16r Z a. - 3 i' 8 F O Z 8 CD 6 o >- L cr 2 i )F. )F 2 F *6 fts : * HSTAMN A HSTAMN + URCHOLN a L A i A *. 8 i A A A A GASTRC SCRTORY VOLUM RAT (ml/15 min) FG.. R PLOTTD AS A FUNCTON OF SCRTORY RAT N XPRMNTS N WHCH HSTAMN AND URCHOLN WR USD. Concentration of HCO in gastric juice did not vary greatly in our experiments. At all but the loest secretory rates HCl concentration as 18 to 163 mq per L during stimulation ith histamine and 133 to 16 mq per L ith gastrin. Therefore, the output of acid as almost a linear function of the volume rate of secretion of gastric juice, and the relation beteen acid output and plasma clearance as of the same form as the relation beteen the volume rate of secretion and clearance. A graphic comparison of gastric secretory volume rate and plasma clearance in the six experiments of this type is shon in Figure 3. t can be seen that secretion and clearance exhibit a significant positive correlation (r =.95, p <.1) over the range of minimal to maximal secretory response to histamine. A comparison of secretory rates and R in the same experiments (Figure ) again demonstrates the high R values at lo secretory rates, the decline in R as secretion increases, and the plateau of R in the neighborhood of 3 as maximal and supramaximal secretory rates are achieved in the pouch. Gastrin. Simultaneous gastric secretory a-id circulatory responses to different doses of gastlin A 5

6 6. D. JACOBSON, R. H. LNFORD, AND M.. GROSSMAN The effect of vasopressin on gastric secretion, muc W._ - -a > 6 : r- Cl) 2 GASTRN 1Ogm/hr GASTRN gm/hr URCHOLN 1.5 mg/hr ) Om - - Gastric Secretory Volume Rate Plasma Clearance R c W 1 2.U) 5a. c N._ "- cr n C. 2 r 18F F _ TM(hours) FG. 5. RSPONSS OF SCRTORY RAT, CLARANC, AND R TO GASTRN AND URCHOLN N ON XPRMNT GASTRC SCRTORY VOLUM RAT ( FG. 6.. TORY RAT N SX XPRMNTS USNG GASTR] LAT SCRTON. and to gastrin and bethanechol in one animal are shon in Figure 5. The maximal secretory response to gastrin as attained at a dosage of 1 g per hour; no further secretory increment as observed hen g per hour as administered; hoever, addition of bethanechol to gastrin induced a supramaximal secretory response. The changes of mucosal blood flo, as estimated by the clearance, corresponded quite ell to the changes in secretion, both in direction and magnitude. R * S - shoed a decrease as secretion as increasing from lo levels and then remained fairly steady at 3 to throughout the remainder of the experiment. * Graphic comparison of secretory volume rate :1 and plasma clearance in six experiments ith gastrin is shon in Figure 6. The scatter is some- * s hat greater than for histamine (r =.85, p <.1). Hoever, as ith histamine, hen secretory rate increased, clearance increased. A plot of R as a function of secretory rate for all gastrin experiments reveals, as ith histamine, the same rapid decay in R as secretory rate rose 1 12 from minimal values ith a plateau at 3 to in :ml/15min) the maximal and supramaximal secretory range MUCOSAL CLARANC AS A FUNCT[ON OF SCR- (Figure 7). DO [N TO STMU- Gastric secretory inhibitors, a) Vasopressin..

7 GASTRC MUCOSAL BLOOD FLOW MASURD BY AMNOPYRN CLARANC cosal plasma clearance, and R during background stimulation ith a fixed dose of histamine as observed in three animals, and results from one experiment are shon in Figure 8. The expected inhibition of secretion induced by the administration of vasopressin as associated ith a parallel decline in clearance and a slight rise in R. After cessation of the vasopressin, secretion and mucosal blood flo increased, hereas R exhibited a return toard prevasopressin values. A plot of secretion versus clearance in the experimental periods during hich vasopressin and histamine ere administered together is shon in Figure 9. Since vasopressin inhibited secretion so effectively, the range of values is restricted to lo secretory and blood flo rates. Nevertheless, comparison of these values ith similar secretoryclearance points during administration of histamine alone (Figure 3) reveals little overlap beteen groups. With vasopressin and histamine higher secretory rates ere associated ith an increased blood flo (r =.95, p <.1). b) soproterenol. nhibition by isoproterenol of submaximal gastric secretory rates maintained ith infusion of histamine as investigated in seven animals. Results from one experiment using to doses of the sympathomimetic dilator appear in Figure 1. At the loer dose the drug did g 18 ( W 16 * GASTRN 1- * GASTRN+ URCHOLN W 1 U o U W 12 a. 1 _ 8 - a- 6 - *r * A ' - A A A UA 2 - U&. 2 C. <' GASTRC SCRTORY VOLUM RAT (ml/15 min) FG. 7. R PLOTTD AS A FUNCTON OF SCRTORY RAT N XPRMNTS USNG GASTRN ALON OR N COMBNA- TON WTH URCHOLN. not depress secretion, and the clearance as markedly increased. With the higher dose both secretion and clearance declined; hereas secretion fell 7 HSTAMN.6 mg/hr. 6 in 5 Volume Rate i -* PlasmaClearance of- OR Gastric Secretory units/hr _ C 3 Lii -2 R. 1 Lii U) en 1 A () TM (hours) FG. 8. NHBTON OF BOTH SCRTON AND MUCOSAL CLARANC BY VASOPRSSN DURNG STMULATON WTH HSTAMN.

8 8 c C.) a. cn 7- F n 5r 5F 35 3k 25F 2F 15F 11-5 F OL L * * J. Xi X,2fs xytx *. D. JACOBSON, R. H. LNFORD, AND M.. GROSSMAN * HSTAMN + SOPROTRNOL x HSTAMN + PTRSSN VOLUM RAT OF GASTRC SCRTON (ml /15 min) FG. 9. MUCOSAL CLARANC AS A FUNCTON OF SCR- TORY RAT DURNG NHBTON OF HSTAMN-STMULATD SCRTON BY PTRSSN OR SOPROTRNOL.. ell belo preisoproterenol values, clearance as approximately reduced to preisoproterenol values. Thus, the clearance diminished less than secretion in response to the higher dose of isoproterenol; hoever, during the preceding period hen the loer dose as used, the clearance as high relative to the rate of secretion. After ithdraal of isoproterenol, both secretion and clearance increased, ith secretion not reaching control values and mucosal plasma flo exceeding control levels. R as abruptly increased from the control plateau by isoproterenol, and the progressive increase in R continued long after cessation of infusion of dilator catecholamine. Results in the other six experiments ere similar, i.e., isoproterenol could increase clearance ithout altering secretory rate; once secretion as inhibited, hoever, clearance also decreased from values obtained hen secretion as not affected by isoproterenol. A plot of values for secretion versus clearance during the periods hen isoproterenol and histamine ere administered together is shon in Figure 9. For this drug combination, as ith all preceding agents, secretory rate changes ere associated ith clearance changes in the same direction (r =.98, p <.1). c._ 3.5 V 3.- Gastric Secretory M Volume Rate. _-@ -O R Plasma Clearance R 1-15 o>!5.. d n V TM (hours) FG. 1. RSPONSS OF SCRTON, CLARANC, AND R TO HSTAMN ALON AND N COMBNATON WTH TWO DOSS OF SOPROTRNOL, ON OF WHCH NHBTD SCRTON C Ẹ _ C-) c -J (n -j a.

9 GASTRC MUCOSAL BLOOD FLOW MASURD BY AMNOPYRN CLARANC 9 GASTRN 1 gm/hr r.dcou=ddkc.' J 3 R (n J (n< o D L TM (hours) FG. 11. RSPONSS OF SCRTORY RAT, CLARANC, AND R DURNG NHBTON OF GASTRN-STMULATD SCRTON BY PNPHRN. c) pinephrine. The effects of epinephrine during background stimulation ith gastrin) on gastric secretion, clearance, and R ere observed in three dogs. Typical results (Figure 11) ere an inhibition of both secretion and clearance and an increase in R during the interval of epinephrine infusion, and a return toard precatecholamine values after cessation of epinephrine. The scatter of values for secretion versus clearance during experimental periods hen the to drugs ere delivered together is shon in Figure 12. Changes in secretory rate ere associated ith directionally corresponding changes in clearance (r =.9, p < d) Secretin. The action of secretin on gastric secretion, clearance, and R as observed in three animals, and results from one experiment appear in Figure 13. The hormone inhibited secretion and clearance and induced a rise in R. These effects ere not entirely gone 1 hour after the last secretin injection. Results from these three experiments (Figure 12) demonstrate that during periods of administration of both secretin and,gastrin changes in secretion correlated ell ith changes in clearance (r =.93, p <.1). - 7 r - en C. -j -J en (a F 8 6F 2 A A *A L A * * GASTRN + SCRTN A GASTRN + PNPHRN o VOLUM RAT OF GASTRC SCRTON (ml/15 min) FG. 12. CLARANC PLOTTD AS A FUNCTON OF SCR- TORY RAT DURNG NHBTON OF GASTRN-STMULATD SCRTON BY PNPHRN OR SCRTN.

10 1. D. JACOBSON, R. H. LNFORD, AND M.. GROSSMAN c._ n - %~~~~~1 ~8unts evr R 6 VoliumeRateteuunl,15 minutes i so6 Plasma Clearance - M ẉ. en r- () 1 i 8 d_ JCFO~~~~~~~~~~~~~~~~~~~~~~~~- _j TM (hours) FG. 13. RSPONSS OF SCRTORY RAT, CLARANC, AND R DURNG NHBTON OF GASTRN-STMULATD SCRTON BY SCRTN. Discussion The findings of our study indicate that secretion and blood flo are related in the stomach. There are, hoever, other reports that present evidence against such a relationship, namely that knon secretory stimuli did not increase gastric blood flo and knon inhibitors of secretion failed to decrease blood flo to the stomach (7-11). n four of these reports (7-1) anesthetied and even hypotensive animals ere used, hich may account for some of the discrepancies among these papers: histamine as observed to increase (11), to decrease (8), and not to alter gastric blood flo (7, 1); epinephrine as found to increase (1, 11) and to decrease (7, 8) blood flo to the stomach. Only the investigation by Delaney and Grim (11) as conducted on unanesthetied dogs, and they performed no secretory studies. Thus, their finding that secretin and epinephrine (secretory inhibitors) did not diminish gastric blood flo is of questionable value, since these drugs ere employed in the unstimulated stomach. Finally, Harries (23) demonstrated that isoproterenol decreased gastric secretion, hich, he maintained, disproved a relationship beteen secretion and blood flo, since isoproterenol is a splanchnic vasodilator. Harries did not, hoever, measure the effect of isoproterenol on gastric blood flo. Our results strongly support a positive relationship beteen gastric secretion and blood flo. n a large number of experiments ith seven different drugs that affect gastric secretion, e consistently observed this relationship: as gastric secretion changed, gastric mucosal blood flo (as estimated by clearance) changed correspondingly. Results ith the stimulatory agents histamine, gastrin, and bethanechol appear to us to be the most clear-cut; dose-response relationships obtained about as ell for clearance as for secretory rate. n the experiments ith gastric secretory inhibitors, results ere more complex, as for example the different effects of vasopressin and isoproterenol on R hen both agents depressed secretion and clearance. The results from these inhibitory experiments, hoever, still support a positive correlation beteen secretion and blood flo (Figures 9 and 12). The ratio (R) of gastric juice to plasma aminopyrine concentrations declined to a plateau value in the range of 3 to hen secretion achieved moderate to maximal rates, but as considerably higher than hen small doses of histamine or

11 GASTRC MUCOSAL BLOOD FLOW MASURD BY AMNOPYRN CLARANC gastrin ere employed. This plateau value for aminopyrine is similar to hat Shore and his associates (13) reported ith a maximal stimulatory dose of histamine. We also noted that during inhibition of secretion, especially ith the catecholamines and secretin, R became markedly elevated. We have interpreted R as an indicator of the limitation imposed by the circulation on the transit of poorly ionied drugs across the stomach. R may be interpreted further as the number of unit volumes of mucosal blood that have floed during the formation of one unit volume of secretion under a given set of conditions. Thus, a high R indicates that blood flo is disproportionately high in relation to secretion, a condition seen at lo secretory rates ith lo doses of stimulants, or ith secretory inhibitors (e.g., secretin, catecholamines) that do not limit blood flo, or ith an agent that vasodilates ithout inhibiting secretion (lo dose of isoproterenol in Figure 1). At high secretory rates induced by high doses of stimulants, R declines to a fixed plateau value, probably representing the situation in hich secretion is blood flo limited, and further increments in secretory rate ill require 3 to ml blood flo for each ml of gastric juice that ill be secreted. f our inference that secretion and mucosal blood flo are normally related is valid, the next consideration is one of assessing hether secretory changes determine blood flo or the reverse. Our evidence does not allo a definitive anser to the nature of the secretory-circulatory relationship in the stomach; hoever, from our experiments ith isoproterenol (Figure 1), it seems that blood flo can increase significantly ithout a corresponding increment in secretion. On the other hand, e have not observed significant secretory changes ithout corresponding directional changes in clearance. Secretion and blood flo in the stomach appear to be related in to major ays: 1) secretory changes determine directionally corresponding changes in the gastric mucosal circulation, and 2) various conditions of stimulation and inhibition can act to alter the amount of blood flo associated ith a unit volume of secretion. n other ords, it appears as if secretory changes induce corresponding circulatory alterations, but the reverse relationship, that is, alteration of secretion by a change in blood flo, is seen only hen blood flo is reduced to a limiting value. 11 The major assumption of this clearance technic, namely that extraction is high and stable, remains to be proved. The extraction ratio for aminopyrine cannot be determined directly because pure gastric mucosal venous blood cannot be sampled. f the extraction ratio for aminopyrine varied greatly ith different experimental conditions, then the clearance of aminopyrine ould not be a valid method for measuring gastric mucosal blood flo, and the subscripts t, m, and nm represent togastric extraction of aminopyrine as 58%. Using this value e can relate extraction to flo in the stomach, as t = (m Fm + nm Fnm)/ (Fm + Fnm), here is extraction, F is plasma flo, and the subscripts t, m, and nm represent total, mucosal, and nonmucosal, respectively. f e assume that nonmucosal gastric tissue does not extract aminopyrine, then it follos that ith an observed t of 58%, either m is 1% (or close to it), or Fnm is extremely lo. Thus, if m ere 1%o, Fnm ould be 2% of total gastric blood flo. This partition of blood flo beteen mucosa and nonmucosa seems reasonable, since the mucosa is about half the stomach mass. Furthermore, this distribution agrees ell ith distribution figures reported by Delaney and Grim (2) using 2K uptake and by Shore and his colleagues (13) employing the clearance of aniline and is close to the estimate made by Mall (25) on anatomical grounds. n any event, if this analysis is correct, m cannot vary beyond the range of 58 to 1%o. n our experiments the ratio of gastric juice to plasma aminopyrine concentrations (R) as found to vary from 2 to 175 (Figures and 7). This eightfold range of R values, and the even larger range of mucosal blood flo values could be explained on the basis of a variation in the relationship beteen mucosal blood flo and secretion or a variation in extraction ratio. As the preceding discussion indicates, an eightfold variation in extraction ratio is not possible. t is possible that aminopyrine is actively secreted by the gastric mucosa. Our finding that stimulation of gastric secretion increases the amount of aminopyrine accumulating in the lumen of the stomach ould support this possibility. The data appear at face value to sho that secretory rate and clearance are inseparable. f this possibility ere correct, a corollary ould be that the gastric mucosa presents high resistance to pas-

12 12. D. JACOBSON, R. H. LNFORD, AND M.. GROSSMAN sive diffusion, clears aminopyrine incompletely from the blood, and accomplishes this clearance mainly by active transport. This alternative explanation cannot be refuted directly; hoever, the diversity of chemical compounds that can be concentrated by the stomach makes it unlikely that there is an active transport mechanism involving specific chemical reactions (13). Furthermore, e have found that altering the ph of solutions instilled into the pouch lumen markedly changes the amount of aminopyrine that ill cross from blood to lumen in the nonsecreting stomach. Our experiments ith lo doses of isoproterenol (Figure 1) sho one situation, at least, in hich clearance can be raised ithout increasing secretion. Finally, as e have discussed previously, clearance from the mucosa cannot be far from complete. Summary The relationship beteen secretion and mucosal blood flo as investigated in the stomach of unanesthetied dogs provided ith denervated pouches. Gastric mucosal blood flo as estimated by aminopyrine clearance. Secretion and mucosal blood flo ere studied during administration of varying amounts of histamine or porcine gastrin, and of combinations of these stimuli ith bethanechol, vasopressin, isoproterenol, epinephrine, or secretin. Under all conditions of stimulation and inhibition, clearance varied directly ith secretory rate; hoever, for some situations (e.g., isoproterenol plus histamine) mucosal blood flo per unit of secretion as higher than for other situations (e.g., vasopressin plus histamine). From these findings e conclude that secretion and mucosal blood flo are related in the stomach. Acknoledgments The authors gratefully acknoledge the technical assistance of Messrs. William Day, Alvin Chang, and Andre Billings. References 1. Cutting, W. C.,. C. Dodds, R. L. Noble, and P. C. Williams. Pituitary control of alimentary blood flo and secretion. The effect of alterations in blood flo on gastric secretion. Proc. roy. Soc. B 1937, 123, Wolf, S., and H. G. Wolff. Human Gastric Function. An xperimental Study of a Man and His Stomach. London, Oxford University Press, 193, chap. and V. 3. Thompson, J.., and J. R. Vane. Gastric secretion induced by histamine and its relationship to the rate of blood flo. J. Physiol. (Lond.) 1953, 121, 33.. Peters, R. M., and N. A. Womack. Hemodynamics of gastric secretion. Ann. Surg. 1958, 18, Jacobson,. D., J. B. Scott, and. D. Frohlich. Hemodynamics of the stomach.. Relation beteen gastric secretion and blood flo. Amer. J. dig. Dis. 1962, 7, Peter,. T., D. M. Nicoloff, H. Sosin, A.. Walder, and. H. Wangensteen. Relationship beteen gastric blood flo and secretion. Fed. Proc. 1962, 21, Lim, R. K. S., H. Necheles, and T. G. Ni. The vasomotor reactions of the (vivi-perfused) stomach. Chin. J. Physiol. 1927, 1, Boenheim, F. tuber das Minutenvolumen des Magens und seine Beeinfluissung durch Blutdruck, durch Vagusreiung, durch Histamin und durch Organextracts. Z. ges. exp. Med. 193, 71, Necheles, H., P. Levitsky, R. Kohn, M. Maskin, and R. Frank. The vasomotor effect of acetylcholine on the stomach of the dog. Amer. J. Physiol. 1936, 116, Cumming, J. D., A. L. Haigh,. H. L. Harries, and M.. Nutt. A study of gastric secretion and blood flo in the unanesthetied dog. J. Physiol. (Lond.) 1963, 168, Delaney, J. P., and. Grim. xperi-mentally induced variations in canine gastric blood flo and its distribution. Amer. J. Physiol. 1965, 28, Jacobson,. D. The circulation of the stomach. Gastroenterology 1965, 8, Shore, P. A., B. B. Brodie, and C. A. M. Hogben. The gastric secretion of drugs: a ph partition hypothesis. J. Pharmacol. exp. Ther. 1957, 119, Butt,. M., A. M. Hoffman, and S. N. Soll. xperimental production of neutropenia ith aminopyrine. Arch. intern. Med. 1939, 6, Gillespie,.., and M.. Grossman. Potentiation beteen Urecholine and gastrin extract and beteen Urecholine and histamine in the stimulation of Heidenhain pouches. Gut 196, 5, Gregory, R. A., and H. J. Tracy. The constitution and properties of to gastrins extracted from hog antral mucosa.. The isolation of to gastrins from hog antral mucosa. Gut 196, 5, Brodie, B. B., and J. Axelrod. The fate of aminopyrine (Pyramidon) in man and methods for the estimation of aminopyrine and its metabolites in biological material. J. Pharmacol. exp. Ther. 195, 99, Youden, W. J. Statistical Methods for Chemists. Ne York, John Wiley, 1951, p. 16.

13 GASTRC MUCOSAL BLOOD FLOW MASURD BY AMNOPYRN CLARANC Snedecor, G. W. Statistical Methods Applied to xperiments in Agriculture and Biology, 5th ed. Ames, oa State College Press, 1956, pp. 6, Milne, M. D., B. H. Scribner, and M. A. Craford. Nonionic diffusion and excretion of eak acids and bases. Amer. J. Med. 1958, 2, Van Os, G. A. J.,. J. Ariens, and A; M. Simonis. Drug transference: distribution of drugs in the organism in Molecular Pharmacology,. J. Ariens, d. Ne York, Academic Press, 196, vol. 1, pp Jacobson,. D.,. S. Dooley, J. B. Scott, and. D. Frohlich. ffects of endotoxin on the hemodynamics of the stomach. J. clin. nvest. 1963, 2, Harries,. H. L. The mode of action of sympathomimetic amines in inhibiting gastric secretion. J. Physiol. (Lond.) 1957, 138, 8 P. 2. Delaney, J. P., and. Grim Canine gastric blood flo and its distribution. Amer. J. Physiol. 196, 27, Mall, F. The vessels and alls of the dog's stomach. Johns Hopk. Hosp. Rep. 1896, 1, 1. SPCAL NOTC TO SUBSCRBRS Post Offices ill no longer forard the Journal hen you move. Please notify The Journal of Clinical nvestigation, Business Office, 1 Stoughton Street, Boston, Mass. 2118, at once hen you have a change of address, and do not omit the Zip Code number.