Upsala J Med Sci 79: 72-83, 1974 Quantitative Determination of Pulmonary Microembolism ' in the Dog Comparison of In vivo and In vitro Methods CHRISTER BUSCH From the Department of Forensic Medicine, University of Uppsala, Uppsala, Sweden ABSTRACT External monitoring and in vitro measurement of pulmo- nary microembolism using 5'Cr-labelled platelets and 1251-labelled fibrinogen were compared. The micro- enbolism was induced in dogs by intravenous infusion of thrombin and the radioactivities were recorded by an external detector as well as at intervals in biopsy material. External detection proved to give lower esti- mates of the amount of trapped radioactivity than in vitro measurements. This was particularly so for 51Cr. Thus only about 2/13 of the in vitro amount of trapped "Cr and about 1/3 of the I z 5 I radioactivity were revealed by external detection. External detection, however, records the changes continuously and there- fore offers advantages over the biopsy technique, which requires open chest surgery and artificial respiration with consequent alterations in intrathoracic pressure and circulatory conditions. INTRODUCTION In previous communications from this laboratory, methods for the quantitation of intravascular plate- let and fibrin deposition in various organs of rats, rabbits, dogs and humans have been presented ( 3 , 4 , 5, 6 , 7 , 12, 13). Isotope-labelled platelets and fibrinogen were injected in advance and the organ deposition following intravenous infusion of throm- bin was measured in vivo by external detec- tors o r i n vitru by radioactivity measurements in tissue specimens. The in vivo technique was used for continuous registration of deposition of fibrin and platelets in, and their elimination from, the lung after intra- venous (i.v.) infusion of thrombin in normal dogs with inhibited fibrinolysis. The method proved to be very useful in elucidating the dynamics of pul- monary microembolism, thought it can be postulat- ed that in viho measurement o f lung tissue ra- dioactivity should give more accurate quantitative information. The aim of the present study was to determine the quantitative differences between the in vivo and the in vitro techniques and hence to define the applicability of each of them. Thus, the pulmo- nary distribution of radioactivity before and after an i.v. infusion of thrombin into dogs given 99mTc-albumin, T r - p l a t e l e t s and T-fibrinogen was studied by the two techniques. MATERIAL AND METHODS Animals. Twenty-four healthy German pointer dogs of both sexes, weighing 15-24 kg, were used. Labelling of platelets. The day before the experi- ment autologous platelets werelabelledwith 5 C r by a modi- fication (4) of the method described by Aas & Gardner ( I ) . Each dog was given a suspension of labelled plate- lets containing 1 5 4 0 pCi Labelledfibrinogen. Human fibrinogen prepared accord- ing to Blomback & Blomback (2) was used. The label- ling was performed according to Rosa et al. ( I I ) . Each dog was given an i.v. injection of about 60 pCi of 9- labelled fibrinogen (5 pCi/mg fibrinogen) 2 0 4 0 min before commencement of the thrombin (or saline) infusion. In previous experiments no differences in the kinetics of human and dog fibrinogen were observed (4). Labelled albumin. Human serum albumin labelled with 9 4 m T ~ (AB Atomenergi, Studsvik) was used to determine the plasma volume per gram of lung tissue. Immediately after passage through an ion exchange column (Dowex I-X8, 50-100 'mesh) about 100 pCi was injected i.v. ( 1 mCi/mg). This was allowed to mix with the plasma vol- ume for 5 rnin before the animals were killed (9). Thrombin. Bovine thrombin (Topostasin@, Roche) dissolved in saline was used. Blood samples were collected through a catheter inserted in the femoral artery (see below). One-ml portions were taken into small Ellermann plastic tubes for determination of radioactivity. by i.v. injection. Upsola J Med Sci 79 Quantitutive determination o j p u l m o n a v y microembolism in dog 7 3 fin sections were stained with haematoxylin-eosin and the Mallory PTAH method for demonstration of fibrin. Radioautogruphs were prepared from sections adja- cent to those used for light microscopy; they were dipped in Ilford G5 emulsion and the film was exposed for 3 weeks a t 4°C. The sections were stained with haema- toxylin-eosin. % c m 0 1 2 3 4 5 6 c m Fig. 1 . Absorption of 51Cr-radioactivity (left) and of 1251 (right) in water at various distances from the ori- fice of the collimator. Isocount lines are expressed as percentage of the counting rate at t h e orifice of the collimator. Haemutocrit (hct) was determined in triplicate in microhaematocrit tubes after centrifugation at 10 000 g for 5 min. No correction for trapped plasma was made. Calculations. In both in vivo and in vitro methods the amount of radioactivity of each isotope in the lung (IL(r)) was calculated as IL(t)=Llt) -Clfi where L ( t ) is the counting rate over the lung (or the radio- activity per gram lung tissue) at time t and C ( t ) the amount of circulating radioactivity recorded by the detector (or in the lung tissue sample) at time t , estimated as where P ( o ) and P ( t ) are the plasma radioactivities at times the aid of the haematocrits. ments the following standardization was introduced: o and t , calculated from whole blood activities with To allow comparison of results from different experi- Thus, the amount of radioactivity estimated with either of the described methods was expressed in per cent of the preinfusion radioactivity over (or in) the lung. This calculation requires two assumptions, first that the pulmonary blood volume does not change during the experiment, and second that the extravascular radio- activity can be neglected. These assumptions are dis- cussed later. Statisticul m e t h o d s Mean values, standard deviations, linear regressions and correlation coefficients were calculated by con- ventional statistical methods. Differences between mean values were tested by means of Student’s /-test. General experimental procedure The dogs were anaesthetized by intravenous adminis- tration of thiopental sodium (Pentohalsodiumm, Abbott), 10 mg per kg body weight (b.w.). A cuffed endotracheal tube was inserted and the dogs were placed supine. A polyethylene catheter was advanced into the right atrium via the right external jugular vein and another was inserted in the femoral artery. In previous experiments the dogs were anaesthetized with a- chloralose dissolved in 10 ml of saline per kg b.w. (4) and to obtain comparative conditions in the present experiments each dog was given this amount of saline before the thrombin infusion (see below). In addition the dogs were pretreated i.v. with 100 mg per kg b.w. of tranexamic acid (AMCA, Kabi. Stockholm) to prevent fibrinolysis. An infusion of thrombin, 310 NIH units per kg b.w., was given in 30 min through the jugular vein catheter. The dogs submitted to a left- sided thoracotomy were given only 150 NIH units per kg b.w. because with the higher dose the mortality rate in open chest conditions is high. Fig. 2. Schematic diagram of the position of the dog’s chest a n d o f t h e detector. LB=lead bricks, SB=sand bags,LL= left lung, R L =right lung, H=heart, LC=lead collimator, Morphological m e t h o d s Light microscopy. Specimens were taken from differ- , ent parts of both lungs. Formalin fixed 5 thich paraf- C = N a I (TI)crystal,A=aorta. Upsala J Med Sci 79 74 Christer Busch 1 0 0 0 1 r = o 803 V W c W Fig. 3 . Relationship between the estimates of the amount of 51Cr-radioactivity trapped in the lungs a s measured in vivo and in x 1 0 0 1, . 2 0 0 L O O 600 * * 1000 n 2 300 w 2 0 0 2 0 0 0 I L ( ' I r e , vitro. 2 -l n c w BIOPSY Arterial blood samples were drawn at regular inter- vals for determination of radioactivity, plasma fibrinogen and platelet count. Radioactivity measurements Recording by exfernal detector. The detector consisted of a 1"xl" sodium iodine crystal with a photomulti- plier (Harshaw) and a 24 mm lead cylindrical collimator with an inner diameter of 50 mm. The distance between the crystal and the collimator orifice was 63 mm. The sensitivity of 51Cr- and 1251-radioactivity in water a t various distances from the front of the collimator was measured using a point source of radioactivity, giving the isoresponse curve a s shown in Fig. 1 . With this detector the radioactivity was monitored continuously over the right lung a s described earlier (4, 5). The dogs were kept immobile by lead bricks. The detector was placed immediately cranial to the edge of the intercostal angel and immediately t o the right of the midline and was directed 10-20" laterally to avoid disturbance from the heart and major vessels. The chest was also rotated about 20" to the left to shift the mediastinal structures from the area used for detection (Fig. 2). The signals from the detector were passed t o a linear amplifier, analysed by a discriminator and recorded by a scaler (Canberra Industries, Connecticut, USA). I- u w I- 2 o 0 1 W n 1 a > z LL W + X W 1 2 5 1 r = 0.815 O 0 8 Two or three channels were used simultaneously and set for I z 5 I , 9 9 m T ~ and 51Cr respectively. Normally more than 400 counts per 100 seconds were recorded. The background was less than 1/6 of this value. The counting rates in each channel was corrected for over- lapping between the channels and for background. For 99mTc correction was also made for decay; this was not done for rzsI and W r , with their long halflives (60 and 28 days respectively). At autopsy the position of the organs in relation to the detector was checked. Radioactivity in tissue and blood samples was meas- ured in a gamma counting system (Gammatrix C , Stock- holm) with a well crystal and three windows set for 99mTc, 51Cr and Iz5I. Normally more than 2000 counts were registered in each channel. The background was usually less than 1/10 of the total counting rate. Cor- rections for overlapping between the channels, for decay of 99mTc and for background were made as described above. Recording by gamma camera.' One dog was placed as described above (supine with the thorax rotated ' This part of the investigation was performed with the kind and skilful assistance of K.-J. Vikterlof and K.-W. Beckman, Department of Radiophysics, Regional Hospital, Orebro, Sweden. Fig. 4 . Relationship between the estimates of the amount of 1251-radioactivity trapped in the lungs as measured in vivo and in I L ( ' I r e L vitro. 1 0 0 200 3 0 0 B I O P S Y Llpsala J Med Sci 79 Quantitative determination of pulmonary microembolism in dog 75 5 30 60 M I N THROMBIN F i g s . 5 and 6 . Pulmonary deposition of T r (Fig. 5 ) and 1Z5I (Fig. 6) (IL(r)re, as measured in vitro (-) and in v i v o (---) as well as the percentual changes recorded by the about 20" to the left) and was given about 150 pCi of 13'I-labelled human fibrinogen i.v. A 1 000-hole, diver- gent collimator was placed over the thorax in the same direction as t h e detector (described above). The radioactivity distribution was recorded by a gamma camera (Selektronik, Denmark). The information was stored, treated and presented by an on-line com- puter system (NUCAB 2530) incorporating a PDP 8/e computer with 12 k core memory and a cartridge memory for storing programme and pictorial informa- tion. The activity distribution was registered as con- secutive 64 x64 matrices during 100-second intervals. The entire right lung, the base of the same lung and the heart were selected as "regions of interest" and the radioactivity of these regions detected during the 100-second intervals was listed digitally and analogously plotted by the computer. E X P E R I M E N T S A N D R E S U L T S I . Comparison of in vivo and in vitro determi- nation of pulmonary microembolism A . Six dogs were given 51Cr-labelled platelets and '251-labelled fibrinogen as described. T h e y were ventilated by a n Engstrom respirator and a left- sided thoracotomy was performed so as t o expose the base of t h e left lung f o r biopsies. T h e radio- % 300 200 1 0 0 1 2 5 6 30 THROMBIN 6 0 ~ MIN detector without correction for plasma decrease of the radioactivity (-). activity over the right lung was recorded con- tinuously and after administration af A M C A t h e thrombin w a s infused in 30 min. Small biopsies (0.3-1.0 g) were taken repeatedly a t intervals of 5-10 min during t h e infusion and in o n e dog for a further 50 min. T h e biopsies were taken from the peripheral parts of t h e lower lobe with the aid of vessel clamps and ligatures. T h e samples were weighed a n d t h e radioactivities determined. As c a n b e seen in Figs. 3 and 4, a n almost linear relationship between t h e estimates of trapped radioactivity of both slCr and lZ5I was obtained. T h e correlation coefficient was 0.803 for T r (n=19) and 0.815 (n=27) for ln51. T h e uptake reg- istered b y external detection was considerably lower than that measured in vivo. T h u s , f o r T r , the in vitro values were about 6.5 times higher than the in vivo estimates for lz5I about times higher. In Figs. 5 and 6 t h e estimates of pulmonary de- position of W r and lZ5I (IL(t)re,) in o n e of the dogs as measured in vitro and in vivo, as well a s t h e percentual changes recorded b y the external detector without correction f o r decrease of plasma radioactivity, a r e plotted against time. T h e quanti- Upsala J Med Sci 79 76 Christer Busch Table I . The rcrtios bet,t,een the sKr-rudioactivity per gram tissue at the end of the thrombin infusion and that per gram plasma before the infirsion (mean ?S.D.) No. of Dog no. Tissue 1 2 3 4 5 samples Skin 0.064k0.005 0.038+0.010 0.007k0.003 0.052r+0.014 5 Skeletal muscle 0.014t0.003 0.040?0.016 0.001 20.007 0.008k0.007 5 Bone-cartilage 0.153~0.021 0.092t0.022 0.065?0.012 0.162t0.053 5 Lung, ventral part 6.559-el.086 3.885tO.615 5.759?0.335 2.163t0.195 3.291 50.955 15 Lung, dorsal part 5.138k1.347 3.692k0.495 5,61820.394 2.135-eO.283 3.576?1.117 IS tative differences between the techniquesare shown, a s well as the unexplained deviation of the amount of label in the tissue specimens. B. Five dogs were treated in the s a m e way as those under “ A ” except that they were allowed to breathe spontaneously and a higher dosage of thrombin (310 NIH units per kg b.w.) was used. After completion of t h e thrombin infusion the dogs were killed with an overdose of thiopental sodium. Small pieces (0.3-1.0 g) were taken from the ventral and t h e dorsal parts of the right lung (15 samples from each region) as well a s from the skin, skeletal muscle (intercostal) and bone and cartilage from the ribs ( 5 samples from each tissue). Samples were also taken from different parts of the lung for morphological studies. T h e radioactivities of t h e tissue and corresponding plasma samples were determined. T h e ratio be- tween tissue radioactivity per gram and the pre- infusion plasma activity per gram was calculated for each sample. T h e radioactivity of both isotopes was seen to b e evenly distributed in different parts of the lung (Tables I and 11) and t h e microemboli were found t o b e localized in small pulmonary arterioles and capillaries (Fig. 7 and 8). No emboli were found in the large vessels o r in the heart at autopsy. T h e amount of radioactivity in the thoracic wall tissues did not increase during the thrombin infusion (Tables I and 11). 11. Distribution of 99mTc-albumin, 51Cr-platelets and Iz5Z-fibrinogen before the thrombin infusion T h r e e dogs given ”Cr-platelets and l z 5 1 - f i brino- gen were injected i.v. with 100 p C i gYmTc-albumin. Ionogenic 99mTc was removed by a n ion exchange column (Dowex 1-X8, 50-100 mesh) immediately before the infusion. After S-min mixing, which was followed by t h e external detector, the dogs were killed with a n overdose of thiopental sodium. At this time the 99mTc-albumin should have been distributed homogeneously in the vascular com- partment (9). T h e lungs were removed and I5 small pieces from the ventral and 15 from the dorsal parts of t h e right lung were transferred into plastic tubes after gentle drying o n filter paper. Furhermore, tissue samples from the thoracic wall (0.5-1.0 g), liver, kidney and spleen were collected analogously with those mentioned a b o v e , and the radioactivity was measured. T h e content of each label per gram was determined in the tissue specimens and in plasma. Table 11. The ratio between the ~Z51-radioactivity per gram tissue at the end of the thrombin infusion and that per gram plasma immediately before the thrombin infusion ( m e a n 5 S . D . ) Dog no. No. of Tissue I 2 3 4 5 samples 0.043?0.004 5 Skin 0.018~0.001 0.01450.002 0.01 1 k0.002 Skeletal muscle 0.014?0.001 0.01520.002 0.011 ?0.001 0.02520.002 5 Bone-cartilage 0.059?0.013 0.030k0.009 0.044?0.004 0.019?0.010 5 Lung, ventral part 2.08920.284 1.115kO.176 1.499?0.170 2.65121.402 1.32650.220 15 Lung, dorsal part 1.78320.326 1.14150.151 1.34620.362 2.926t0.965 1.57450.242 IS Upsala J Med Sci 79 Quantitative &termination of pulmonary microembolism in dog 77 F o r none of t h e labels did the tissue/plasma ra- tios show a significant difference between different parts of the lung (Tables 111-V). T h e ratio for lz5I were lower than those of 99mTc in the lungs, liver a n d kidneys. T h e 51Cr ratios were higher than those for 1451 and 9 Y m T ~ in t h e spleen, liver and b o n e c a r t i l a g e a s well as in the lung (Table 111). 111. Statistical evaluation of in vivo external de- tection of pulmonary microembolism A . Three dogs' were injected with 5LCr-labelled platelets and Lz51-labelled fibrinogen as in previous experiments. After a n injection of 20 ml saline thrombin was infused as before into the jugular vein in 30 min (310 NIH units per kg b.w.). Ex- ternal detection was carried out o v e r right lung for a further 5 1/2 hours. During this time t h e animals were given another 50 ml of saline. B. Five dogs1 were injected with AMCA 100 mg per kg b.w. about 15 min before an infusion of , Fig. 7. Fibrin microemboli in pulmoi arterioles and capillaries (PTAH, x 1 iary 150). thrombin identical t o that above. After t h e 30-min infusion another 100 mg of AMCA per kg b.w. dissolved in 50 ml of saline was given o v e r 5 1/2 hours. T h e radioactivity was monitored a s in group A. The amount of 51Cr and lz5I radioactivity trapped in t h e lung a s measured by t h e in vivo method (IL(t)re,) was plotted against time (Figs. 9 a n d 10). C. T w o dogsZ were given 20 ml saline and another t w o dogs 50 mg AMCA per kg b.w. a s a single, rapid i.v. injection. 1251-fibrinogen was also given in advance and external detection was then performed as described (Fig. 1 I ) . One of t h e main purposes of the in vivo method was t o measure differences in the rate of elimi- nation of the labelled microemboli from the I The results of these experiments have been pre- sented in detail elsewhere (4). * The results of these experiments have been pre- sented in detail elsewhere ( 5 ) . Upsala J Med S c i 79 7 8 Christer Busch lung. In Figs. 9 and 10 it is visually evident that AMCA treatment delayed the elimination of both 1251-fibrin and 5*Cr-platelets. This effect was most pronounced from one to 3 hours after completion of the thrombin infusion. The tails of the curves show high coefficients of variation and/ or low numbers of observation. Fig. 8 . Radioautograph of microemboli (within arrows) in small pulmonary vessels (haematoxylin-eosin, ~ 3 0 0 ) . Fig. I I shows that even a lower dose of AMCA than was used in previous experiments and given as a single rapid injection caused retardation of the elimination as compared with dogs with nor- mal fibrinolysis. In dogs given 200 mg AMCA per kg b.w. the pre-infusion counting rate was usually not regained during the 6-hour observation period. Table 111. The ratios between 51Cr-rudioa~tivity per grum tissue and per grum piusma before the thrombin infusion (mean fS.D.) Dog no. N o . of Tissue I 2 3 samples Skin Skeletal muscle Bone-cartilage Liver Kidney Spleen Lung, ventral part Lung, dorsal part 0.03 1 k0.021 0.03120.014 0.19520. 160 3.92450.643 0.262 20.0 19 25.074 0.491 k0. 193 0.660 20.405 0.014?0.007 0.015?0.003 0.04620.018 0.69720.006 0.174-tO.013 12.364 0.191 20.038 0.163 50.050 0.017k0.003 0.01320.002 0.064 k0.023 0.56320.018 0.134k0.019 17.823 0.26220.046 0.22520.023 5 5 5 2 2 1 15 15 Upsala J Med Sci 79 Quantitative determination of pulmonary microembolism in dog 19 Table I V . The ratios between ‘251-radioactivity per gram tissue and p e r gram plasma before the thrombin infusion ( m e a n + S . D . ) ~~ ~~ - Dog no. N o . of Tissue I 2 3 samples Skin Skeletal muscle Bone-cartilage Liver Kidney Spleen Lung, ventral part Lung, dorsal part 0.02520.002 0.01 520.003 0.02720.00 I 0.13720.001 0.17620.013 0.097 0.160?0.024 0.156?0.022 0.020 20. 004 0.01620.003 0.03 1 20.002 0.108~0.001 0.22620.004 0.170 0. I1520.01 I 0.11520.019 Differences in the elimination rate can also be expressed as differences in the percentual amount of the trapped radioactivity at various time points after the maximum value. Such figures (IL(t)re,/ 1L(0.5)re,) for groups A and B are shown in Ta- ble VI. Thus, significant differences in the ratios were found at 60, 90 and 120 min after commencement of the thrombin infusion for both isotopes, while the tail values did not differ significantly. IV. Recording by gamma camera One dog was given 150 pCi 1311-fibrinogen and a 30-min thrombin infusion. After about 1 hour’s registration of the radioactivity changes about 1 mCi of 99mTc-labelled macroaggregated human serum albumin (AE3 Atomenergi, Studsvik) was injected into the jugular vein catheter S O that the contours of the lungs and heart were visualized. The entire right lung, the base of the same lung and the heart were taken as “regions of interest” and the radioactivity recorded i n 0.023+0.002 5 0.018 20.002 5 0.05120.012 5 0.191 20.014 2 0 . 1 6 2 ~ 0 . 0 1 1 2 0.157 I 0.1 1320.013 15 0.115t0.013 15 these regions during the 100-second intervals was listed digitally and plotted analogously by the computer. The results are given in Fig. 12. It is evident that the gamma camera technique was less effici- ent in revealing the true increase of pulmonary radioactivity than the lead collimated detector placed over a smaller field of the right lung. The pictures also show that the shifting of the chest to the left also shifts the mediastinal struc- tures, thus giving a large free field for detection at the base of the right lung. DISCUSSION The present study has shown that quantitative determination of labelled pulmonary microemboli by external detection in dogs gives lower estimates of the trapped radioactivity than in vitro measure- ment. The differences are probably due to the fact that the radioactivities measured emanate from Table V. The ratios between 99mTc-radioactivity per gram tissue and per gram plasma before the thrombin infusion (mean L S . D . ) Dog no. N o . of Tissue I 2 3 samples Skin 0.0 18 20.003 Skeletal muscle 0.01 5 20.004 Bone-cartilage 0.032t0.014 Liver 0.127?0.002 Kidney 0.292 tO.0 I0 Spleen 0.070 Lung, ventral part 0.18020.025 Lung, dorsal part 0.161 20.018 0.041 20.018 0.021 20.005 0.040 20.004 0. 12720.001 0.335 c0.010 0.149 0.15520.010 0.15020.009 0.02220.002 5 0.01 9?O.O01 5 0.056 20.0 I 2 5 0.217?0.013 2 0.314t0.014 2 - 1 0.14520.016 15 0.14020.013 15 Upsala J Med Sci 79 80 Christer Busch 9 =SALINE + THRMBB 0 = AMCA + THROMBIN 10 ZOO =SALINE +THROMBIN 0 =AMCA + THROMBIN $ L 2 150 . L U ? *- k? 9 2 Figs. 9 and 10. Pulmonary deposition (IL(t)re,) of 5'Cr-radioactivity (above) and lZ51-radioactivity (beiow) in dogs pretreated with AMCA, 200 mg per kg b . w . ( 0 ) and dogs with normal fibrinolysis (0). - 100 ,- 5 0 HOURS THROMBIN different sources in the two techniques. Thus, the external detector records activity originating in all thoracic tissues, the limits and extent of the tissue field depending on the collimating properties of the detector and on the energy of the radio- active isotope employed (Fig. I ) . The counting rate is an expression of the sum of radioactivity from the chest wall, peripheral lung tissue (where the deposition actually occurs), large vessels and perhaps also from the right heart (in spite of the preventive measures). The biopsy specimens consist of peripheral lung tissue only, thus with no disturbance form the blood of large vessels. Thur the decrease of radio- activity in the blood of large vessels gives a smaller net accretion when measured with the external technique than when measured in vitro on tissue samples. This difference was more pronounced for 51Cr than for Only about 2/13 of the in vitro estimate was recorded externally, as compared with about 1/3 for the It is conceivable that the 51Cr radioactivity with its higher gamma energy (0.320 MeV), emanates to a greater extent from the large central vessels than the radio- activity of Iz5I (0.035 MeV) (Fig. I ) . The pulmonary microemboli were shown to be homogeneously distributed in the arterioles and capillaries, while at autopsy no emboli were found in the heart or large pulmonary arteries. Further- more, no significant increments of the 51Cr or lZ5I radioactivities were observed in the various parts of the chest wall. Thus an uneven distribution of deposits in the tissues mentioned did not influence the measurements. The related differences are also illustrated in Upsala J Med Sci 79 Quantitative determination of pulmonary microembolism in dog 8 1 w D > r > t - - 0 - z? 4 z & t o be about 30% (6). In the present experiments this amount should be considerably smaller, since only 2 0 4 0 min elapsed between the administratiop of t h e labelled fibrinogen and the thrombin infusion. A large amount of 51Cr radioactivity was recover- I-, , , , , , , , ed in the spleen, liver and bone-cartilage. Further- 2 3 4 more, the lung/plasma ratio of 51Cr was higher HOURS THROMBIN 1 Fig. / I . Pulmonary uptake and elimination of lZ5I in dogs pretreated with A M C A , 50 mg per kg b.w. (-) and in dogs pretreated with saline (- - -). Figs. 7 and 8, where the estimates of pulmonary trapping of t h e radioactivities are presented as well as the changes recorded by the detector without correction f o r plasma decrease. In spite of the occasional variations in t h e tissue sample radio- activities the in vitro and in v i v o methods a r e seen to describe t h e same time course, which further indicates that they reflect the same patho- physiological events. T h e distribution of the different isotopes be- fore t h e thrombin infusion in the present study revealed some interesting facts. Firstly, the tissue/ plasma ratios for all isotopes were homogeneously distributed in the different parts of the lung. Second- ly, the ratios of 99mTc were higher than those of Iz5I in t h e lungs. kidneys and liver. This than that of the two other isotopes indicating that this isotope and/or labelled platelets o r fragments of platelets a r e retained in these organs on re-infusion of t h e suspension of Y r platelets. T h e accumulation of 51Cr in the liver, spleen and bone might b e d u e t o the action of the re- ticuloendothelial system, while the accretion in the lung might indicate trapping of aggregates formed during the labelling procedure. These high lung/ plasma ratios introduce a n error in the estimation of the amount of circulating 51Cr activity be- fore the thrombin infusion ( P ( t J 0 ) . Since, how- ever, the tissue/plasma ratio after the thrombin infusion ( L ( t ) / P ( t ) , ) was about 14 times that be- fore the infusion ( L ( t ) o / P ( t ) o ) , and the correction made f o r circulating radioactivity ( C ( t ) w a s rather small because of t h e decreased plasma radioactivity ( P ( t ) ) , this error was neglected. The increase in counting rate over the lung during the thrombin infusion might theoretically be d u e t o a n increase in t h e pulmonary plasma volume. In a n earlier study using 12sI-labelled Table VI. The ratio between the cimount of 5LCr and Iz5I tripped in the lungs a t various time points after the thrombin infusion a t the amount at the end o f t h e infusion (IL(t)re,/lL(0.5)re, ( m e a n M . D . ) 1251 51Cr Saline+ A M C A + Saline + A M C A + Minutes after thrombin thrombin thrombin thrombin thrombin infusion &kS.D. - X,?S.D. P - X 3 2 S . D . - X,kS.D. P - - start of the - - 60 0.327k0. 122 0.869k0.046 <0.001 0.427k0.052 0.698k0.069 <0.01 90 0.274k0.082 0.781 k0.066 <0.001 0.314k0.045 0.54220.097 <0.02 120 0.312k0.146 0.672k0.099 <0.02 0.15220.092 0.36420.062 c0.05 180 0.397k0.200 0.59720.123 Non-sign. 0.154k0.058 0.244k0.037 Non-sign. 240 0.379k0.245 0.574+0.133 Non-sign. 0.13220.070 0.212?0.055 Non-sign. 6 - 142855 Upsala J M e d Sci 79 82 Christer Busch W 3 a > LL 0 z 75l 75 i 0 I I 500 1000 1500 2000 2500 3000 3500 SECONDS 751 T H ROM 81 N albumin and external detection, no such increment was observed, however (4). The gamma camera system did not prove to be more sensitive than recording by the conventional detector. The former requires high doses of high energy radioisotopes for adequate discrimination and again the influence of decreasing plasma radio- activity, especially in the large vessels, must affect the recordings. This is supported by the fact that some improvement was obtained by se- lecting only the base of the right lung as the “region of interest”. 1311 is not an ideal radioisotope for studies with gamma camera. It is possible that the use of more suitable isotopes such as Iz31 which has not been available to us could increase the sensi- tivity and hence make monitoring by gamma camera suitable and possible for study also of human microembolism. The results of the external detection in this study are i n accordance with those of Saldeen (13), who used a similar technique on rabbits given 1311-labelled fibrinogen and thrombin i.v. Only a small, though significant increase was detected over the lung. Coombey & Tyler (8), using 1251-fibrinogen, infused 50 units of throm- bin via a sublingual vein i n 24 min into rats. The Fig. 12. Gamma camera pic- tures of the anatomical rela- tionship in the dog’s chest and the “regions of inter- est” ( , left). The radio- activity changes in each re- gion (% of preinfusion values) are shown to the right. radioactivity was recorded externally over the lung. The increase was about 30% above the baseline level. According to o u r experience from rats (6), however, such a thrombin infusion will give a true in vitro. increase in lung radioactivity of several hundred percent, which is another example of the differences between in vitro and in vivo tech- niques. In conclusion, external determination is a re- producible method particularly suited for con- tinuous study of the dynamics of pulmonary micro- embolism. The method can also be used for simul- taneous registration of the radioactivity in other organs such as the liver and kidneys (5). It pro- vides definite advantages over the biopsy tech- nique, which limits the number of observations and occasionally shows considerable variations between different samples. These variations might be due to altered intrathoracic pressure and cir- culatory conditions following the thoracotomy and artificial respiration. The sensitivity of the external method is limited, however, and in states with an expected low de- gree of intravascular coagulation the biopsy tech- nique should be used. In investigations of humans with intravascular coagulation and microembolism the biopsy method Llpsala J Med Sci 79 Quantitative determination of pulmonary microembolism in dog 8 3 is not practicable. H e n c e t h e external m e t h o d h a s R e c e i v e d N o v e m h e r 5 , 1973 been e m p l o y e d in t h e s t u d y of p u l m o n a r y fibrin and platelet deposition a s s o c i a t e d with t r a u m a . In t h e s e cases e v i d e n c e of transient (10) or progressive (3) a c c u m u l a t i o n o f isotope h a s b e e n o b s e r v e d . Address for reprints: Department Of Forensic Medicine Rattsmedicinska avd. U n i v e r s i t y of REFERENCES 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. I I . 12. 13. Aas, K. A. & Gardner, F. H.: Survival of blood platelets labelled with Chromium”. J Clin Invest 37: 1257. 1958. Blomback, B. & Blomback, M.: Purifaction of human and bovine fibrinogen. A r k i v Kemi 10: 415, 1966. Busch, C., Dahlgren, S., Jakobson, S., Jung, B., Modig, J . & Saldeen, T.: Determination of fibrin trapping in the lungs of patients developing the microembolism syndrome. Acta Chir Scand 137: 599, 1971. Busch. C., Lindquist, 0. & Saldeen, T.: Respiratory insufficiency in the dog induced by pulmonary microembolism and inhibition of the fibrinolytic system. Effect of defibrinogenation, leucopenia and thrombocytopenia. Acta Chir Scand 1974 (in press). Busch, C., Lundquist, H . & Saldeen, T.: Quantita- tive determination of intravascular coagulation in vivo in various organs in the dog. Thromb Diathes Haemorrh (Stuttg.)26: 315, 1971. Busch, C., Rammer, L . & Saldeen, T.: Quantitation of fibrin deposition and elimination in organs of rats injected with labelled fibrinogen and albumin. Thromb Diathes Haemorrh (Stuttg.) 29: 94, 1973. Busch, C. & Rammer, L.: Quantitation of fibrin deposition and elimination in organs of rats in- jected with labelled fibrinogen by isolation of the fibrin from water soluble tracer. Thromb Diathes Haemorrh (Stuttg.) 29: 108, 1973. Coombey, D. & Tyler, H . M.: Quantitative moni- toring of intra-vascular coagulation and fibrinolysis in the lungs of rats. Thromb Diathes Haemorrh (Stuttg.) 27: 241, 1972. Huggins, R. A,, Smith, E. L. & Deavers, S . Volume distribution of Evans blue dye and iodinated albumin in the dog. Am J Physiol2M: 35 I , 1963. Modig. J., Busch, C., Olerud, S. & Saldeen, T.: Evidence of platelet and fibrin deposition in the lungs during intra-medullary orthopaedic trauma. To be published. Rosa. U., Scasselatti, G . A. & Pennisi, F.: Label- ling of human fibrinogen by electrolytic iodination. Biochem Biophys Acta (Amst.) 86: 519, 1964. Saldeen, T.: Experimental investigation on fat em- bolism and intravascular coagulation. Acta Path Microbiol Scand 66: 271, 1966. Saldeen, T. Quantitative determination of intra- vascular coagulation in the lungs of experimental animals. Scand J Haemat6: 205. 1969. S-75 1 23 UppsalH’I Sweden Upsala J Med Sci 79