- Email: [email protected]
Effect of dicumarol on wound healing
EFFECT OF DICUMAROL FREDERIC
W.
TAYLOR,
M.D.
ON WOUND
AND H.
Indianapolis,
A
N experimenta attempt is made to investigate the effect of dicumarol@ on wound heahng. The method of direct observation and histoIogic study is used in wounds made in dogs. Under the experimenta range (prothrombin of IO per cent or less) there is a definite tendency to bleed into the wound. Besides this, no difference between control and experimental wounds could be detected. Wounds under the influence of dicumaroI@ heaIed as rapidIy as those made under similar circumstances without this drug. These Iindings are in keeping with clinical experience. Various methods used in wound-heaIing study are briefIy outlined. Some of the fauIts and shortcomings of each are indicated. Particular attention is leveled at the procedure which bursts a previousIy sutured hollow viscus or body cavity by inside pressure. It is pointed out that stress on such a wound is proportional to the diameter of the cavity as well as to the pressure in that cavity. Much of the work on wound strength has been done by this method without any consideration of the organ size or diameter.
HASKELL
HEALING*
ZIPERMAN,
M.D.
Indiana noted as compared with the contro1 animaIs. Bendix and NecheIes2 using dogs also had diffrcuIty in the contro1 of prothrombin Ievels. Their experiences with dicumaroI@ were too brief for concIusions. They warn of the systemic dangers of this drug as compared with heparin. In a broad sense the effect of dicumaroI@ on wounds is aIready we11 understood. It is known that wounds heal very satisfactorily in patients receiving a therapeutic quantity of the drug. A critica survey on an experimenta basis has been of vaIue to us and should be of interest to others. METHODS
Seventeen mongrel dogs varying in weight from 19 to 52 pounds were used. One of these dogs was used once as an experimental and again later as a control anima1, thus making a tota of eighteen dogs used. Of these, eIeven were given dicumaroI@ and seven were used as controIs. Each dog was given a daiIy diet of Purina Dog Chow suppIemented by standard canned dog food with water as desired. Vitamin C deIiciency as a factor in wound heaIing was eIiminated by giving each anima1 a daiIy dose of IOO mg. of ascorbic acid (cecon). Each experimental dog was given 50 mg. of dicumaroI@ as an initial dose and from 25 to 50 mg. of dicumaroI@ either daily or every other day depending on the prothrombin time. Prothrombin times were determined daiIy or every second day throughout the course of the experimenta1 work. AI1 these animals were maintained at a prothrombin IeveI of IO per cent or Iess. Prothrombin times were determined as foI10~s: 434 cc. of whole bIood were drawn into a test tube containing $5 cc. of .I moIar sodium oxalate. The test was run within two hours after the blood was drawn and usuaIIy within one hour. To .I cc. of the aforementioned plasma was added .I cc. of thrombopIastin derived from
**** The continued use of dicumaroI@ in surgical patients suggests the advisabiIity of inquiring into the effect of this drug on wound heaIing. Numerous studies have been made in regard to heparin and wound healing. Two of these1s2 incIude as a portion of that study a brief inquiry and comment on dicumaroI.@ No other references to the subject couId be found. Part of SandbIom’sl extensive review on the subject of wound heaIing deaIs with the effect of dicumaroI@ on heaIing skin wounds in rabbits. He experienced considerable difficulty in obtaining uniform prothrombin blood Ievels. There is aIso reason to doubt that the prothrombin remained at therapeutic IeveIs throughout the five-day test. The resulting breaking strengths (tensiometer) were quite variabIe but no significant differences were
* From the Department of Surgery, Indiana University MedicaI SchooI, Indianapolis, Ind.
Ju(v, I 930
“3
11-t
TayIor,
Ziperman-Dicumarol
FIG;. 1. Four-day wound heating contro1 on left; experimental skin wound under influence of dicumarol@ on right. Artifact of the contro1 wound gives appearance of non-healing. There is somewhat more maturity in this particular experimental wound. (van G&on’s stain.)
placental extract.* The mixture was incubated for one minute in a water bath followat 37.9~. ing which .I cc. of ~25 moIar calcium chloride at the same temperature was added. The tube containing these ingredients was removed from the water bath and tilted regularly every three to four seconds from the verticaI to the horizontal and back until a cIot formed. A stop watch was used for timing and was started on the addition of the calcium chloride and stopped with the formation of a clot. AII of the experimental dogs had at Ieast one and most had two prothrombin titrations each of plasma diIutions of 50, 30, 20 and IO per cent. Using these values a hyperbolic curve was drawn in which the ordinates represented per cent and the abscissas represented prothrombin time in seconds. Experimental work was started onIy after the anima1 had reached and was maintained at a prothrombin level of I0 per cent or less. *This preparation has been used in the Indiana University IHospitaIs with valid and rather constant results.
in Wound
Healing
FIG. 2. Nine-day wound healing contro1 on Left (hemntoxylin-eosin stain); ideal dicumarol@ wound healing on right (trichrome stain).
In the first six dogs incisions were made through skin, fascia and muscIe of the abdominal wall. This site was discontinued because of the frequency of wound infections and because incisions through fascia and muscIe added variables which seemed unnecessary. For the most part skin incisions were made over the dogs’ shouIder so that the animal could neither Iick nor scratch them. At the start of the experiment three +$inch incisions were made at this site so that one couId be excised on the fourth, sixth and ninth postoperative day. Each incision was cIosed with a singIe silk mattress-on-edge suture. The excised wounds were fixed in Zenker’s solution and stained with hematoxyhn-eosin, van Gieson’s and trichrome stain. RESULTS
GrossIy, there was no difference in the appearance of the experimental and control wounds except for one feature. There was a definite tendency on the part of the experimental wounds to bleed. The resuIt was that if the ooze was not entirely stopped, a hematoma would develop. This, of course, would definiteIy weaken the wound and delay healing. Sandblomi noted the same bIeeding tendency. American
Journal
of Surgery
TayIor,
Ziperman-DicumaroI
U’hen this was controlIed in the experimenta wound, we were unabIe to detect any detrimental effect of dicumaroI@ on heaIing rate or production of any of the elements necessary to wound healing. Forty-seven wounds were excised at various intervaIs of heaIing (four, six and nine days) and studied histoIogicaIIy. Twenty were from control animaIs and twenty-seven from experimental. No attempt was made to estimate or tabulate the amount of coIlagen or reticulum found in the sections as others13 have done. The photomicrographs shown in Figures I and 2 are representative and typica of the series. With the three stains used no rea1 difference was noted between the contro1 and experimental wounds. For wounds of the same age the amount of recticuIum, coIIagen and the number of fibrocytes seemed to be identical. COMMENTS
This method of studying wound healing may seem intangibIe and open to question. However, so are other methods. For this reason a brief discussion of the probIem seems indicated. There are in genera1 three different types of procedures which have been used in the study of experimental wound healing. They are the foIlowing: Perhaps the most popular means of comparing wound healing and wound strength has been that used by Harvey and Howes3 and others.“-s This consists of making an incision in the abdomina1 waI1 or the stomach or other hoIIow viscus of an experimenta anima1. The wound is then sutured under the conditions of the experiment. HeaIing progresses to the termination of the tests at which time water or air is injected into the hoIIow viscus or abdomina1 cavity until the wound or organ ruptures. On the surface this wouId seem to be a very exact means of comparing wound heaIing and wound strength. Perhaps it is the most trustworthy method but it is not fauItIess. First of a11 it has been assumed by al1 the workers using this method that the manometer reading at which the sutured wound ruptures can be directly transIated into the tensile strength of that wound. This is false and quite misleading since an added factor is the particular diameter of the viscus in question. In other words with equa1 pressures the tension on
Ju(v, I ygo
in Wound
Healing
115
a sutured wound varies directly with the diameter of the viscus and is not whoIIy dependent on the manometer reading. Wound tension is only indirectIy related to the cross sectiona area of the hoIIow viscus as suggested by Price. y There is nothing mysterious about figuring the tensiIe stress on the wall of any tank or bIadder-Iike structure. It is a common engineering computation. lo,‘1 It is equal to the unit pressure times the projected area (diameter) over which that pressure acts divided by two (two sides). Since the viscus is circular, the diameter is usually given as radius (diameter divided by two). Thus S = pr when S is the tensile stress, p is the unit pressure and r is the radius. The significance of this may not be clear unti1 we examine an actua1 case. Let us say we have incised and sutured wounds in the stomachs of rabbits A and B. After a week the animals are anesthetized, their stomachs tied at both ends and connected to a manometer. Both ruptured at the suture Iine as the pressure reached IOO mm. Hg. In the past it was argued and reported that both wounds had the same tensiTe strength without considering the size of the particuIar viscus. In our example Iet us say that stomach A was 2 cm. in diameter at its incised area when it burst and stomach B measured 4 cm. Then using our formuIa (S = pr) it is found that aIthough the interna pressure (p) was the same, the tensiIe strength (S), withstood by wound B was two times that of A since r was twice as great. By the same reasoning wound B was twice as strong although the pressure which broke both wounds was the same. This is appIicabIe onIy to the freely distended viscus or abdomina1 waI1 which is blown open by an equaIIy distributed gas (or Iiquid). It is a Iaw of physics and must not be confused with experimental resuIts. Other objections to this method of woundheaIing study are those which cannot be eIiminated in any such study. There is a Iarge human factor of different operators and the same operator under different times and circumstances. It is likewise not vaIid to compare the strength of wounds sutured with absorbabIe and non-absorbabIe material. The nonabsorbabIe materia1 obviousIy may add to the tested strength of the wound and, therefore, not give a true reflection of the situation at hand. Catgut in itseIf may be the uncontrolled
116
TayIor,
Ziperman-DicumaroI
factor through its variation in absorption time or reaction in the tissues. There is aIso the variation in response or sensitivity of the individual (or animal) to the catgut. The second method of testing wound-heaIing strength is actualIy to find out how much force is required to pul1 a given wound apart. SeveraI types of tensiometers have been used to measure this force. One variation consists of testing strips of tissue supposedly of equal width through like portions of the wound.12 The variabIes of such tests are obviousIy many. This is especiaIIy true when considering the added factors of interrupted or continuous suture, just where in reIation to the sutures the particuIar tissue was taken, the method of fixing tissue to the tensiometer and the variable shearing action of this instrument on the wound. To these objections may be added many of those mentioned in the frrst method. The third method of wound-healing study consists of (I) directly observing and measuring of the healing of various skin defectsI or (2) histoIogic study of various stages of heaIing wounds. The first type of study is, of course, Iimited to wounds healing by secondary intention. In the method of study by histoIogic section we have no concrete knowledge of just how strong the wound may be. We have onIy histologic evidence which we compare to a similar wound in a control anima1. This may be quite convincing but nevertheIess the results are a picture and not a measure in pounds strength per Iinear inch of healing wound. When evaluating such a wound various features are taken into consideration. The wound may be graded as to content of reticulum, and fibrocytes. coIlagen, fibrin, hemorrhage These are a11 important features for consideration in any healing wound but they do not necessariIy indicate whether or not idea1 healFor instance it has been ing has occurred. inferred that the greater the production of reticuIum and coIIagen the stronger has been wound healing. I2 This may be so but in our
in Wound
Healing
experience infected wounds produced excessive amounts of collagen and reticulum. These excesses are far from the ideal goal for wound healing. In fact the most idea1 primary healing occurred when these substances were at a minimum and were represented histoIogicaII3 by only a thin line. We have used this method to study wound heaIing under the inffuence of dicumaroI.@ The fauIts and shortcomings of the method are fully admitted. In the same breath we emphasize the fauIts of other methods. REFERENCES
P. The tensile strength of healing wounds. Actn cbir. Scandinm., go: 93, 1944. 2. BENDIX, R. M. and NECHELES, H. Wound healing and heparin, using heparin deposits. Surgery, 26: I. SANDBLOM,
7999 1949. 2. HARVEY. S. C. and HOWES. E. L. Effect of hich protein diet on the veIoci& of growth of fibriblasts in the healing wound. Ann. Surg., 91: 641, ‘930. 4. HOWES, E. L. and HARVEY, S. C. The clinical significance of experimentat studies in wound heaIing. Ann. Surg., 102: 941, 1935. 5. HOWES, E. L. The immediate strength of the sutured wound. Surgerv, 7: 24, 1940. 6. LAXMAX, J. H. and INGALS, J. H. Vitamin C deficiency and wound healing. Ann. Surg., 105: 616, 1933. 7. KOBAK, M. W., BENDITT, E. P., WISSLEK, R. W. and STEFFEE, C. H. The reIation of protein deficiency to experimenta wound healing. Surg., Gynec. f?* Obst., 85 : 75 I, 1947. 8. ALRICH, E. M. and LEHMAN, E. P. The effect of heparin on wound heaIing. Surg., Qnec. CTObst., 87: 26, 1948. Q. PRICE, P. B. Stress, strain and sutures. Ann. Surg., 128: 4o8, 1948. IO. MARKS, L. S. Mechanical Engineer’s Handbook. 4th ed., p. 449. New York, 1941. McGraw-Hi11 Book Co., Inc. I I. KENT, ROBERT THURSION. Kent’s Mechanica Engineers Handbook. New York, 1938. John Wiley and Sons, Inc. 12. WOLFER, J. A., FARMER, C. J., CARROLL, W. W. and MANSHARDT, D. 0. Experimental study in wound heaIing in vitamin C depleted human subjects. Surg., Gynec. @T Obst., 84: I, 1947. 13. SCHEINBERG,S., BRALOW, S. P. and NECHELAS, H. Experimental study of wound healing. SurgeTv, 24: 972, 1948.
American
Journal of Surgeq
W.
TAYLOR,
M.D.
ON WOUND
AND H.
Indianapolis,
A
N experimenta attempt is made to investigate the effect of dicumarol@ on wound heahng. The method of direct observation and histoIogic study is used in wounds made in dogs. Under the experimenta range (prothrombin of IO per cent or less) there is a definite tendency to bleed into the wound. Besides this, no difference between control and experimental wounds could be detected. Wounds under the influence of dicumaroI@ heaIed as rapidIy as those made under similar circumstances without this drug. These Iindings are in keeping with clinical experience. Various methods used in wound-heaIing study are briefIy outlined. Some of the fauIts and shortcomings of each are indicated. Particular attention is leveled at the procedure which bursts a previousIy sutured hollow viscus or body cavity by inside pressure. It is pointed out that stress on such a wound is proportional to the diameter of the cavity as well as to the pressure in that cavity. Much of the work on wound strength has been done by this method without any consideration of the organ size or diameter.
HASKELL
HEALING*
ZIPERMAN,
M.D.
Indiana noted as compared with the contro1 animaIs. Bendix and NecheIes2 using dogs also had diffrcuIty in the contro1 of prothrombin Ievels. Their experiences with dicumaroI@ were too brief for concIusions. They warn of the systemic dangers of this drug as compared with heparin. In a broad sense the effect of dicumaroI@ on wounds is aIready we11 understood. It is known that wounds heal very satisfactorily in patients receiving a therapeutic quantity of the drug. A critica survey on an experimenta basis has been of vaIue to us and should be of interest to others. METHODS
Seventeen mongrel dogs varying in weight from 19 to 52 pounds were used. One of these dogs was used once as an experimental and again later as a control anima1, thus making a tota of eighteen dogs used. Of these, eIeven were given dicumaroI@ and seven were used as controIs. Each dog was given a daiIy diet of Purina Dog Chow suppIemented by standard canned dog food with water as desired. Vitamin C deIiciency as a factor in wound heaIing was eIiminated by giving each anima1 a daiIy dose of IOO mg. of ascorbic acid (cecon). Each experimental dog was given 50 mg. of dicumaroI@ as an initial dose and from 25 to 50 mg. of dicumaroI@ either daily or every other day depending on the prothrombin time. Prothrombin times were determined daiIy or every second day throughout the course of the experimenta1 work. AI1 these animals were maintained at a prothrombin IeveI of IO per cent or Iess. Prothrombin times were determined as foI10~s: 434 cc. of whole bIood were drawn into a test tube containing $5 cc. of .I moIar sodium oxalate. The test was run within two hours after the blood was drawn and usuaIIy within one hour. To .I cc. of the aforementioned plasma was added .I cc. of thrombopIastin derived from
**** The continued use of dicumaroI@ in surgical patients suggests the advisabiIity of inquiring into the effect of this drug on wound heaIing. Numerous studies have been made in regard to heparin and wound healing. Two of these1s2 incIude as a portion of that study a brief inquiry and comment on dicumaroI.@ No other references to the subject couId be found. Part of SandbIom’sl extensive review on the subject of wound heaIing deaIs with the effect of dicumaroI@ on heaIing skin wounds in rabbits. He experienced considerable difficulty in obtaining uniform prothrombin blood Ievels. There is aIso reason to doubt that the prothrombin remained at therapeutic IeveIs throughout the five-day test. The resulting breaking strengths (tensiometer) were quite variabIe but no significant differences were
* From the Department of Surgery, Indiana University MedicaI SchooI, Indianapolis, Ind.
Ju(v, I 930
“3
11-t
TayIor,
Ziperman-Dicumarol
FIG;. 1. Four-day wound heating contro1 on left; experimental skin wound under influence of dicumarol@ on right. Artifact of the contro1 wound gives appearance of non-healing. There is somewhat more maturity in this particular experimental wound. (van G&on’s stain.)
placental extract.* The mixture was incubated for one minute in a water bath followat 37.9~. ing which .I cc. of ~25 moIar calcium chloride at the same temperature was added. The tube containing these ingredients was removed from the water bath and tilted regularly every three to four seconds from the verticaI to the horizontal and back until a cIot formed. A stop watch was used for timing and was started on the addition of the calcium chloride and stopped with the formation of a clot. AII of the experimental dogs had at Ieast one and most had two prothrombin titrations each of plasma diIutions of 50, 30, 20 and IO per cent. Using these values a hyperbolic curve was drawn in which the ordinates represented per cent and the abscissas represented prothrombin time in seconds. Experimental work was started onIy after the anima1 had reached and was maintained at a prothrombin level of I0 per cent or less. *This preparation has been used in the Indiana University IHospitaIs with valid and rather constant results.
in Wound
Healing
FIG. 2. Nine-day wound healing contro1 on Left (hemntoxylin-eosin stain); ideal dicumarol@ wound healing on right (trichrome stain).
In the first six dogs incisions were made through skin, fascia and muscIe of the abdominal wall. This site was discontinued because of the frequency of wound infections and because incisions through fascia and muscIe added variables which seemed unnecessary. For the most part skin incisions were made over the dogs’ shouIder so that the animal could neither Iick nor scratch them. At the start of the experiment three +$inch incisions were made at this site so that one couId be excised on the fourth, sixth and ninth postoperative day. Each incision was cIosed with a singIe silk mattress-on-edge suture. The excised wounds were fixed in Zenker’s solution and stained with hematoxyhn-eosin, van Gieson’s and trichrome stain. RESULTS
GrossIy, there was no difference in the appearance of the experimental and control wounds except for one feature. There was a definite tendency on the part of the experimental wounds to bleed. The resuIt was that if the ooze was not entirely stopped, a hematoma would develop. This, of course, would definiteIy weaken the wound and delay healing. Sandblomi noted the same bIeeding tendency. American
Journal
of Surgery
TayIor,
Ziperman-DicumaroI
U’hen this was controlIed in the experimenta wound, we were unabIe to detect any detrimental effect of dicumaroI@ on heaIing rate or production of any of the elements necessary to wound healing. Forty-seven wounds were excised at various intervaIs of heaIing (four, six and nine days) and studied histoIogicaIIy. Twenty were from control animaIs and twenty-seven from experimental. No attempt was made to estimate or tabulate the amount of coIlagen or reticulum found in the sections as others13 have done. The photomicrographs shown in Figures I and 2 are representative and typica of the series. With the three stains used no rea1 difference was noted between the contro1 and experimental wounds. For wounds of the same age the amount of recticuIum, coIIagen and the number of fibrocytes seemed to be identical. COMMENTS
This method of studying wound healing may seem intangibIe and open to question. However, so are other methods. For this reason a brief discussion of the probIem seems indicated. There are in genera1 three different types of procedures which have been used in the study of experimental wound healing. They are the foIlowing: Perhaps the most popular means of comparing wound healing and wound strength has been that used by Harvey and Howes3 and others.“-s This consists of making an incision in the abdomina1 waI1 or the stomach or other hoIIow viscus of an experimenta anima1. The wound is then sutured under the conditions of the experiment. HeaIing progresses to the termination of the tests at which time water or air is injected into the hoIIow viscus or abdomina1 cavity until the wound or organ ruptures. On the surface this wouId seem to be a very exact means of comparing wound heaIing and wound strength. Perhaps it is the most trustworthy method but it is not fauItIess. First of a11 it has been assumed by al1 the workers using this method that the manometer reading at which the sutured wound ruptures can be directly transIated into the tensile strength of that wound. This is false and quite misleading since an added factor is the particular diameter of the viscus in question. In other words with equa1 pressures the tension on
Ju(v, I ygo
in Wound
Healing
115
a sutured wound varies directly with the diameter of the viscus and is not whoIIy dependent on the manometer reading. Wound tension is only indirectIy related to the cross sectiona area of the hoIIow viscus as suggested by Price. y There is nothing mysterious about figuring the tensiIe stress on the wall of any tank or bIadder-Iike structure. It is a common engineering computation. lo,‘1 It is equal to the unit pressure times the projected area (diameter) over which that pressure acts divided by two (two sides). Since the viscus is circular, the diameter is usually given as radius (diameter divided by two). Thus S = pr when S is the tensile stress, p is the unit pressure and r is the radius. The significance of this may not be clear unti1 we examine an actua1 case. Let us say we have incised and sutured wounds in the stomachs of rabbits A and B. After a week the animals are anesthetized, their stomachs tied at both ends and connected to a manometer. Both ruptured at the suture Iine as the pressure reached IOO mm. Hg. In the past it was argued and reported that both wounds had the same tensiTe strength without considering the size of the particuIar viscus. In our example Iet us say that stomach A was 2 cm. in diameter at its incised area when it burst and stomach B measured 4 cm. Then using our formuIa (S = pr) it is found that aIthough the interna pressure (p) was the same, the tensiIe strength (S), withstood by wound B was two times that of A since r was twice as great. By the same reasoning wound B was twice as strong although the pressure which broke both wounds was the same. This is appIicabIe onIy to the freely distended viscus or abdomina1 waI1 which is blown open by an equaIIy distributed gas (or Iiquid). It is a Iaw of physics and must not be confused with experimental resuIts. Other objections to this method of woundheaIing study are those which cannot be eIiminated in any such study. There is a Iarge human factor of different operators and the same operator under different times and circumstances. It is likewise not vaIid to compare the strength of wounds sutured with absorbabIe and non-absorbabIe material. The nonabsorbabIe materia1 obviousIy may add to the tested strength of the wound and, therefore, not give a true reflection of the situation at hand. Catgut in itseIf may be the uncontrolled
116
TayIor,
Ziperman-DicumaroI
factor through its variation in absorption time or reaction in the tissues. There is aIso the variation in response or sensitivity of the individual (or animal) to the catgut. The second method of testing wound-heaIing strength is actualIy to find out how much force is required to pul1 a given wound apart. SeveraI types of tensiometers have been used to measure this force. One variation consists of testing strips of tissue supposedly of equal width through like portions of the wound.12 The variabIes of such tests are obviousIy many. This is especiaIIy true when considering the added factors of interrupted or continuous suture, just where in reIation to the sutures the particuIar tissue was taken, the method of fixing tissue to the tensiometer and the variable shearing action of this instrument on the wound. To these objections may be added many of those mentioned in the frrst method. The third method of wound-healing study consists of (I) directly observing and measuring of the healing of various skin defectsI or (2) histoIogic study of various stages of heaIing wounds. The first type of study is, of course, Iimited to wounds healing by secondary intention. In the method of study by histoIogic section we have no concrete knowledge of just how strong the wound may be. We have onIy histologic evidence which we compare to a similar wound in a control anima1. This may be quite convincing but nevertheIess the results are a picture and not a measure in pounds strength per Iinear inch of healing wound. When evaluating such a wound various features are taken into consideration. The wound may be graded as to content of reticulum, and fibrocytes. coIlagen, fibrin, hemorrhage These are a11 important features for consideration in any healing wound but they do not necessariIy indicate whether or not idea1 healFor instance it has been ing has occurred. inferred that the greater the production of reticuIum and coIIagen the stronger has been wound healing. I2 This may be so but in our
in Wound
Healing
experience infected wounds produced excessive amounts of collagen and reticulum. These excesses are far from the ideal goal for wound healing. In fact the most idea1 primary healing occurred when these substances were at a minimum and were represented histoIogicaII3 by only a thin line. We have used this method to study wound heaIing under the inffuence of dicumaroI.@ The fauIts and shortcomings of the method are fully admitted. In the same breath we emphasize the fauIts of other methods. REFERENCES
P. The tensile strength of healing wounds. Actn cbir. Scandinm., go: 93, 1944. 2. BENDIX, R. M. and NECHELES, H. Wound healing and heparin, using heparin deposits. Surgery, 26: I. SANDBLOM,
7999 1949. 2. HARVEY. S. C. and HOWES. E. L. Effect of hich protein diet on the veIoci& of growth of fibriblasts in the healing wound. Ann. Surg., 91: 641, ‘930. 4. HOWES, E. L. and HARVEY, S. C. The clinical significance of experimentat studies in wound heaIing. Ann. Surg., 102: 941, 1935. 5. HOWES, E. L. The immediate strength of the sutured wound. Surgerv, 7: 24, 1940. 6. LAXMAX, J. H. and INGALS, J. H. Vitamin C deficiency and wound healing. Ann. Surg., 105: 616, 1933. 7. KOBAK, M. W., BENDITT, E. P., WISSLEK, R. W. and STEFFEE, C. H. The reIation of protein deficiency to experimenta wound healing. Surg., Gynec. f?* Obst., 85 : 75 I, 1947. 8. ALRICH, E. M. and LEHMAN, E. P. The effect of heparin on wound heaIing. Surg., Qnec. CTObst., 87: 26, 1948. Q. PRICE, P. B. Stress, strain and sutures. Ann. Surg., 128: 4o8, 1948. IO. MARKS, L. S. Mechanical Engineer’s Handbook. 4th ed., p. 449. New York, 1941. McGraw-Hi11 Book Co., Inc. I I. KENT, ROBERT THURSION. Kent’s Mechanica Engineers Handbook. New York, 1938. John Wiley and Sons, Inc. 12. WOLFER, J. A., FARMER, C. J., CARROLL, W. W. and MANSHARDT, D. 0. Experimental study in wound heaIing in vitamin C depleted human subjects. Surg., Gynec. @T Obst., 84: I, 1947. 13. SCHEINBERG,S., BRALOW, S. P. and NECHELAS, H. Experimental study of wound healing. SurgeTv, 24: 972, 1948.
American
Journal of Surgeq