0 l Y Y 0 tlarwood Academic Publishers GmbH
Grorc~tliFactors, 1YY0, Vol. 3, pp 267-275 Reprints available directly from the publisher Photocopying permitted by license onl)
Printed in the United Kingdom
Recombinant Human Transforming Growth Factor-Beta 1 (rhTGF-P1) Enhances Healing and Strength of Granulation Skin Wounds L. STEVEN BECK!', THERESA L. CHENI, PHILIP MIKALAUSKP and ARTHUR J. AMMANN 'Researdl and Derdopnient, Geiieiitccli liic., 460 Poiiit Sat1 Bruno Roiikrrard, Sorrtli Sari Fraiicisco, California 94080; 2Utali Rioiircdrcal Tcst Laboratory fiir., 520 Wakara Way, Salt Lakt7 City, Utah 84108
(Xeceizied March 3 7 990, AccrpttTd April 30 7 990) A new animal model to study secondary intention wound healing and the effects of topically applied rhTGF-P1 was developed. A time course study was performed of full thickness 6 mm punch wounds placed on the backs of anesthetized pigs and treated once with either 3% methylcellulose or rhTGF-Pl in 3% methylcellulose or left untreated. Wounds receiving rhTGF-Pl had enhanced tensile strength at days 4 and 7 compared to controls. Studies of the response on days 4 and 7 to graded doses of rhTGF-Bl showed that a dose of 250 or 2500 ng rhTGF-P7 gave a similar enhanced wound strength, while 25 ng rhTGF-Dl had no effect. Blood flow to treated granulating wounds as measured by ' T e microspherrs indicate an increase in flow in wounds treated with 250, 500 or 2500ng rhTGF-Pl compared to controls. These results indicate a possible use for rhTGF-PI in enhancing wound healing clinically.
KFYWORDS:
recombinant liiiiiiaii tr~~n40riiiinji grou th factor B1, jiranulation
INTRODUCTION
ounds, ten3ile btrengtli, bIoi)~lt k w
indicate an essential role of these factors on promoting the healing process (Dijkc a n d Iwata, Normal wound healing has been described as a 1989; Rotlie a n d Falanga, 1989). Models of secondary series of cellular events designed to return the intention healing Lvhich invol\,es tilling of a tissue damaged tissue to its original functional state defect through formation of granulation tissue and (Peacock, 1984). These events are partially controlled epithelialization have been performed. I'artial and by growth factors which modulate cell migration full thickness Lvound experiments i.ising biochemical and proliferation, angiogenesis, matrix formation, and histomorphometric methods in swine have and early phase remodeling of wounds. The recent shown enhancement of wound healing in the presavailability of recombinant growth tactors has ence of TGF-P1, EGF, IGF-I, and I'DGF (Brown et allowed investigators to examine the sequence of al., 1986; Lynch, Colvin and Antoniades, 1989). To biological events initiated by these factors and to date, however, no functional tiieasurements of evaluate their potential role in wound healing in strength of full thickness wounds healing by seconanimals. Transforming growth factor /?l (TGF-Pl), dary intention have been described in swine. transforming growth factor a (TGF- a ), epidermal Wound healing studies have used swine a s a n growth factor (ECF), basic fibroblast growth factor animal model due to the similarity of swine skin to (bFGF), insulin - I i ke gro M' t h fact or - I ( 1G F- I), plat el e t human skin (Bustad and McClellan, 1965). Earlier, derived growth factor (PDGF) and others have been Ordman and Cillnian (19hha, b c:) used swinc to studied in various wound healing models which study the effects of surgical materials on healing by examining tensile strength and histology of linear incision wounds. More recentlv, the incision model 267
in pigs was used to compare the strength of wounds created by various scalpel devices (Buell and Schuller, 1983; Millay et al., 1987) or to compare various devices and methods of skin closure (Roth and Windle, 1988). The pig is also used to study the pathophysiology of skin, myocutaneous and fasciocutaneous flap failure (Kerrigan et al., 1986; Pang et al., 1986). Partial or full thickness dermal wounds on pigs have been used to examine the effects of wound dressings (Dyson et al., 1988; Eaglstein, 198.5; Eaglstein et al., 1988; Leipziger et al., 1985), pharmaceutics (Alvarez et al., 1984), or low energy laser therapy (Basford et a]., 1986; Saperia et al., 1986) on healing. A partial thickness skin wound produced by a dermatone permits the evaluation of events occurring in a superficial process primarily by histomorphometric and biochemical methods. The advantage of this method is the ability to chronologically measure cellular events and to determine the rate of re-epithelialization (Chvapil et al., -1988). The wounds, however, are difficult to produce in a reproducible manner. A large total surface area on the pig is often necessary to create enough samples for the number of variables examined to assure statistical significance due to large interan i ma 1 va r i a bi 1i t y . The purpose of this study was to evaluate the effects of recombinant human transforming growth tactor @1 (rhTGF-Pl) on the healing of full thickness punch biopsy wounds in swine. A tensometric method is described which combines the welldefined endpoint of tensile strength of the linear incision model with a full thickness wound that requires granulation tissue for healing. In addition, blood flow to the wound was measured by a radioactive microsphere technique as a measure of angiogenesis. This punch biopsy model in swine provides a reproducible and quantitative assessment of wound healing parameters and may be useful in e\~aluatingthe effects of growth factors in addition to rhTGF-@I on the quality and strength of healing granulation tissue.
Department at Genentech Inc. Concentrations ranging trom 0.1 to 1110 y g rhTGF-@l/ml were prepared by the Formulations Department at Genentech in 3% methycellulose containing 20 mM sodium acetate buffer at pH 5.0 (vehicle). The material W ~ stored at YC until used. Animal Surgery Young, 3-month-old male White Yorkshire pigs (25-30 kg) housed and maintained according to the American Association for the Accreditation of Laboratory Animal Care (AAALAC) guidelines were anesthetized with ketamine hydrochloride (35 mg/ kg) and atropine sulfate (0.05 mg/kg). After removal of hair, 28 circular tattoos 1cm in diameter were placed under aseptic conditions on the back in a 4 across and 7 down pattern; 3-7 days later the pigs were again anesthetized and maintained with a halothane/nitrous oxide mixture. Full thickness wounds were produced aseptically in shaved sites in the middle of the previously placed tattoos using a sterile 6 nim trephine. Each full thickness wound was 3.2-3.5 mm deep including the epidermal pegs and dermis which contained hair follicles. The clotted wounds were rinsed with a dilute solution of betadine and saline and dried with a sterile gauze. A single dose of rhTGF-Pl (2.5-2500 nglwound) o r vehicle in a volume of 0.025 nil was applied. A light layer of Vet-Bond" glue (3M, St Paul, MN) was painted around the outside margins of the tattoos to enhance the adherence of the dressing. All wounds were covered with Opsite" (Smith-Nephew Massillon, OH) followed by 10 cm (4 inch) wide Vetwrap" tape (3M, St Paul, MN) placed over the Opsite" and around the trunk of the pig. The anesthetized pigs were placed into a recovery cage and monitored until fully recovered. Sampling Wounds for Tensometry
Full thickness rectangular sections of skin with the long axis parallel to the spine and with the circular wound centrally located were removed from anesthetized pigs. These samples were trimmed to be 10x30 mm and fixed in 3% neutral buffered MATERIALS AND METHODS formalin in labeled containers for 7 days prior to tensometric measurement. Formalin fixation was Source and Preparation of rhTGF-P7 performed for ease of handling the fragile tissue at rhTGF-pl was a product of transfected human 293 the early time points. Although fixation increases cells of human TGF-PI cDNA (Derynck et al., 1985) cross-linking of collagen and increases the absolute and was purified by the Process and Development tensile strength of a wound compared to unfixed
S
rhl'GE-pl FNHANCES ULCER WOUND TFNSll E STRENGIH
wounds, the increase over time parallels tensile strength from wounds without formalin fixation (Levenson et al., 1965). The pigs were then humanely euthanized using T-61" (Hoechst Roussel, Somerville, NJ). Samples obtained on a specific day of recovery after wounding were derived from a single pig. The tissues were uniformly trimmed in width and thickness to assure that the edges of the scar were exposed in both planes of section (Fig. 1). Dimensions of the wound for tensile strength calculation were determined by the use of a micrometer caliper which measured the width and thickness of each sample. Tensometry was performed on coded samples using a calibrated tensometer (lnstron Universal Testing Instrument Model 1011, Instron Corp., Canton, MA). The values determined were breaking strength (g) which is a measure of force in grams applied to the tissue, and tlie tensile strength (g/mm') which is the load applied per crosssectional area across the wound at the time of peak load (breaking strength:cross-section area). Breaking strength was determined at the point which the scar tissue visually broke and a major deflection in the tracing occurred. Tensilc strength was then calculated by dividing the observed breaking strength by the measured cross-sectional area of the scar. Since breaking and tensile strength curves were similar with time only tensile strength data was reported.
269
15 or 2 5 p m diameter (3M, St Paul, MN) were suspended in 2 ml isotonic saline with 0.01% Tween 80 in a microsphere injecting vial (VWR Scientific, San Francisco, CA). This suspension was mixed vigorously with a vortex mixer and injected into the left ventricle within 6U sec using two 20 ml syringes of saline injected through the microsphere injector. Concurrent to the microsphere injection, blood was withdrawn from the abdominal aorta via the femoral arterial catheter attached to a pump set at 12.0 ml/ min starting 15 sec before, and until 45 sec after injection of microspheres; 10 min later full thickness samples of the wound sites were removed with a n 8 nim punch, dried with a towel and weighed. Normal skin samples were obtained from areas adjacent to each wound as controls. The pigs were humanely euthanized with T-61" at the completion of the study. The tissues were placed into scintillation \rials and radioactivity determined with a gamma counter (Model 1282 Compugamma CS, LKB Wallace, Finland). Total radioactivity was determined from the blood sample removed from the abdominal aorta. Blood flow to each site was calculated by dividing the counts per minute (CI'M) of the tissue by the total CPM of the blood sample multiplied by the rate of blood withdrawal (12 mI/ min), that is (CPM,,,/CPM,,,,,,,,x7 2 ml/min. Values were expressed as ,ul/min.
Statistical Analysis Sampling Wounds for Blood Flow Blood flow to the wound was measured with radioactive microspheres according to the method described by Saxena and Verdouw (1985). Anesthetized pigs were placed in lateral recumbency and the left carotid artery and a femoral artery a n d vein were isolated for cannulation. A Swan-Ganz catheter (Edwards Div., Baxter Healthcare Corp., Santa Ana, CA) was placed into the pulmonary artery via the femoral vein for thermodilution cardiac output; a pigtail aortography catheter was advanced into the left ventricle via the left carotid artery for radioactive microsphere injection; and a pressure catheter was placed into the abdominal aorta via the femoral artery for blood withdrawal and central arterial pressure measurement. Mean aortic blood pressure was maintained at approximately 100 mmHg and cardiac output was determined by thermodilution techniques prior to microsphere injection. Cerium-141 labeled microspheres ( - 4 N O " ) either
The treatment groups were conipared with tlie vehicle treated or control groups of the same pigs. The data were first analyLed by analysis of variance. When the overall F-test showed a signlficant difference among groups, individual group means were compared using the Bonferroni multiple comparison procedure or Student's t-test. Sample size varied according to treatment group but in all cases the N (number of wounds per group) was at least 7.
RESULTS
Strength of Granulation Wounds Treatment on each row of wounds were equally divided between rhTGF-P1 and vehicle in order to eliminate bias d u e to anatomic variability from site to site First, the strength ot wounds were determined 4, 7, 10 and 14 days after surgery and
BECK r t nl
270
compared untreated or veliicle treated wounds with wounds that had received one application of rhTGFbl (250 ng/wound) at the time of wounding. The cross-sectional area of the trimmed wounds (see Fig. l), measured just before tensometry, decreased from day 4 to 14 in all groups but remained similar among the untreated, vehicle and rhTGF-PI treated groups at any given time. The mean values for tensile strength increased with tiine in all groups with higher values in rliTGF-pl than untreated or vehicle control at the early time points (Fig. 2). rIiTCF-Pl increased tensile strength ( P < 0.01) on d a y 4 when compared to wounds treated with vehicle alone or left untreated and when compared to the untreated wounds on day 7 (Pcar ( A ) 111 tlie middle 5urby tattoo ( R ) ( 2 ) Trimmed th margin of scar ( A ) at both ) Sidt. v i t v c)t ( 2 ) showing t i o n ot scar ( A ) and tattoo
1
days 4 and 7 atter wounding and treatment w t h 2 5, 25, 231 and 2500 ng rhTGF-pllwound. Values from each rhTGF-pl treated group were expressed as a percentage of the mean vehicle control of the same pig (Table 1) Tensile strength at day 7 was greater ( P < O 01) for wounds treated with 250 and 2500 ng rhTGF-P1 than vehicle TABLE 1 Fftect5 of rhTGb-BI on Strength of Granulation Wounds I r n d r strength was measured from wounds treated with vehicle or rhTGF-BI once at the time of surgery and allowed to heal 1 or 7 d a \ \ Values from each dose group and time point represent data from one pig for a total of eight pigs in this study Tensile strength ("10 of control)" Dav 4 Dav 7
Dose of rhTGF-Bl (ngiwound)
104 (12 X) 97 (11 4) 160 (33 2) 114 (13 Y)
2.5
25 231
2500
115 (4 6 ) 89 ( 5 1 ) 149 (9 7)' 730 (11 3 ) '
'Values from each rhTGF-Dl treated wound were adjusted for the mean all the vehicle control measurements bv expressing the response a 5 a percent of the mean vehicle control which wa3 asslgned t o be 10(1% Data are expressed as mean ( f SEM). Number ot wounds for each dose group and time point was at least 12 " P < (1.01 rhTGF-Pl vs vehicle. 01
2
3
4
5
6
271
rhTGF-Pl ENHANCES ULCER WOUND TENSILE STRENGTH
" .
I
4 DAYS
-.-
7 DAYS
10 d A Y d
14 DAYS
DAYS AFTER WOUNDING
FIGURE 2. Time response changes of tensile strength (g/mm') in pig dermal ulcer wounds. Wounds were untreated or received vehicle alone or 250 ng of TGF-Bl/wound at the time of surgical wounding. Data is expressed as mean k SEM. -0- Untreated, rhTGF-BI. *P
Grorc~tliFactors, 1YY0, Vol. 3, pp 267-275 Reprints available directly from the publisher Photocopying permitted by license onl)
Printed in the United Kingdom
Recombinant Human Transforming Growth Factor-Beta 1 (rhTGF-P1) Enhances Healing and Strength of Granulation Skin Wounds L. STEVEN BECK!', THERESA L. CHENI, PHILIP MIKALAUSKP and ARTHUR J. AMMANN 'Researdl and Derdopnient, Geiieiitccli liic., 460 Poiiit Sat1 Bruno Roiikrrard, Sorrtli Sari Fraiicisco, California 94080; 2Utali Rioiircdrcal Tcst Laboratory fiir., 520 Wakara Way, Salt Lakt7 City, Utah 84108
(Xeceizied March 3 7 990, AccrpttTd April 30 7 990) A new animal model to study secondary intention wound healing and the effects of topically applied rhTGF-P1 was developed. A time course study was performed of full thickness 6 mm punch wounds placed on the backs of anesthetized pigs and treated once with either 3% methylcellulose or rhTGF-Pl in 3% methylcellulose or left untreated. Wounds receiving rhTGF-Pl had enhanced tensile strength at days 4 and 7 compared to controls. Studies of the response on days 4 and 7 to graded doses of rhTGF-Bl showed that a dose of 250 or 2500 ng rhTGF-P7 gave a similar enhanced wound strength, while 25 ng rhTGF-Dl had no effect. Blood flow to treated granulating wounds as measured by ' T e microspherrs indicate an increase in flow in wounds treated with 250, 500 or 2500ng rhTGF-Pl compared to controls. These results indicate a possible use for rhTGF-PI in enhancing wound healing clinically.
KFYWORDS:
recombinant liiiiiiaii tr~~n40riiiinji grou th factor B1, jiranulation
INTRODUCTION
ounds, ten3ile btrengtli, bIoi)~lt k w
indicate an essential role of these factors on promoting the healing process (Dijkc a n d Iwata, Normal wound healing has been described as a 1989; Rotlie a n d Falanga, 1989). Models of secondary series of cellular events designed to return the intention healing Lvhich invol\,es tilling of a tissue damaged tissue to its original functional state defect through formation of granulation tissue and (Peacock, 1984). These events are partially controlled epithelialization have been performed. I'artial and by growth factors which modulate cell migration full thickness Lvound experiments i.ising biochemical and proliferation, angiogenesis, matrix formation, and histomorphometric methods in swine have and early phase remodeling of wounds. The recent shown enhancement of wound healing in the presavailability of recombinant growth tactors has ence of TGF-P1, EGF, IGF-I, and I'DGF (Brown et allowed investigators to examine the sequence of al., 1986; Lynch, Colvin and Antoniades, 1989). To biological events initiated by these factors and to date, however, no functional tiieasurements of evaluate their potential role in wound healing in strength of full thickness wounds healing by seconanimals. Transforming growth factor /?l (TGF-Pl), dary intention have been described in swine. transforming growth factor a (TGF- a ), epidermal Wound healing studies have used swine a s a n growth factor (ECF), basic fibroblast growth factor animal model due to the similarity of swine skin to (bFGF), insulin - I i ke gro M' t h fact or - I ( 1G F- I), plat el e t human skin (Bustad and McClellan, 1965). Earlier, derived growth factor (PDGF) and others have been Ordman and Cillnian (19hha, b c:) used swinc to studied in various wound healing models which study the effects of surgical materials on healing by examining tensile strength and histology of linear incision wounds. More recentlv, the incision model 267
in pigs was used to compare the strength of wounds created by various scalpel devices (Buell and Schuller, 1983; Millay et al., 1987) or to compare various devices and methods of skin closure (Roth and Windle, 1988). The pig is also used to study the pathophysiology of skin, myocutaneous and fasciocutaneous flap failure (Kerrigan et al., 1986; Pang et al., 1986). Partial or full thickness dermal wounds on pigs have been used to examine the effects of wound dressings (Dyson et al., 1988; Eaglstein, 198.5; Eaglstein et al., 1988; Leipziger et al., 1985), pharmaceutics (Alvarez et al., 1984), or low energy laser therapy (Basford et a]., 1986; Saperia et al., 1986) on healing. A partial thickness skin wound produced by a dermatone permits the evaluation of events occurring in a superficial process primarily by histomorphometric and biochemical methods. The advantage of this method is the ability to chronologically measure cellular events and to determine the rate of re-epithelialization (Chvapil et al., -1988). The wounds, however, are difficult to produce in a reproducible manner. A large total surface area on the pig is often necessary to create enough samples for the number of variables examined to assure statistical significance due to large interan i ma 1 va r i a bi 1i t y . The purpose of this study was to evaluate the effects of recombinant human transforming growth tactor @1 (rhTGF-Pl) on the healing of full thickness punch biopsy wounds in swine. A tensometric method is described which combines the welldefined endpoint of tensile strength of the linear incision model with a full thickness wound that requires granulation tissue for healing. In addition, blood flow to the wound was measured by a radioactive microsphere technique as a measure of angiogenesis. This punch biopsy model in swine provides a reproducible and quantitative assessment of wound healing parameters and may be useful in e\~aluatingthe effects of growth factors in addition to rhTGF-@I on the quality and strength of healing granulation tissue.
Department at Genentech Inc. Concentrations ranging trom 0.1 to 1110 y g rhTGF-@l/ml were prepared by the Formulations Department at Genentech in 3% methycellulose containing 20 mM sodium acetate buffer at pH 5.0 (vehicle). The material W ~ stored at YC until used. Animal Surgery Young, 3-month-old male White Yorkshire pigs (25-30 kg) housed and maintained according to the American Association for the Accreditation of Laboratory Animal Care (AAALAC) guidelines were anesthetized with ketamine hydrochloride (35 mg/ kg) and atropine sulfate (0.05 mg/kg). After removal of hair, 28 circular tattoos 1cm in diameter were placed under aseptic conditions on the back in a 4 across and 7 down pattern; 3-7 days later the pigs were again anesthetized and maintained with a halothane/nitrous oxide mixture. Full thickness wounds were produced aseptically in shaved sites in the middle of the previously placed tattoos using a sterile 6 nim trephine. Each full thickness wound was 3.2-3.5 mm deep including the epidermal pegs and dermis which contained hair follicles. The clotted wounds were rinsed with a dilute solution of betadine and saline and dried with a sterile gauze. A single dose of rhTGF-Pl (2.5-2500 nglwound) o r vehicle in a volume of 0.025 nil was applied. A light layer of Vet-Bond" glue (3M, St Paul, MN) was painted around the outside margins of the tattoos to enhance the adherence of the dressing. All wounds were covered with Opsite" (Smith-Nephew Massillon, OH) followed by 10 cm (4 inch) wide Vetwrap" tape (3M, St Paul, MN) placed over the Opsite" and around the trunk of the pig. The anesthetized pigs were placed into a recovery cage and monitored until fully recovered. Sampling Wounds for Tensometry
Full thickness rectangular sections of skin with the long axis parallel to the spine and with the circular wound centrally located were removed from anesthetized pigs. These samples were trimmed to be 10x30 mm and fixed in 3% neutral buffered MATERIALS AND METHODS formalin in labeled containers for 7 days prior to tensometric measurement. Formalin fixation was Source and Preparation of rhTGF-P7 performed for ease of handling the fragile tissue at rhTGF-pl was a product of transfected human 293 the early time points. Although fixation increases cells of human TGF-PI cDNA (Derynck et al., 1985) cross-linking of collagen and increases the absolute and was purified by the Process and Development tensile strength of a wound compared to unfixed
S
rhl'GE-pl FNHANCES ULCER WOUND TFNSll E STRENGIH
wounds, the increase over time parallels tensile strength from wounds without formalin fixation (Levenson et al., 1965). The pigs were then humanely euthanized using T-61" (Hoechst Roussel, Somerville, NJ). Samples obtained on a specific day of recovery after wounding were derived from a single pig. The tissues were uniformly trimmed in width and thickness to assure that the edges of the scar were exposed in both planes of section (Fig. 1). Dimensions of the wound for tensile strength calculation were determined by the use of a micrometer caliper which measured the width and thickness of each sample. Tensometry was performed on coded samples using a calibrated tensometer (lnstron Universal Testing Instrument Model 1011, Instron Corp., Canton, MA). The values determined were breaking strength (g) which is a measure of force in grams applied to the tissue, and tlie tensile strength (g/mm') which is the load applied per crosssectional area across the wound at the time of peak load (breaking strength:cross-section area). Breaking strength was determined at the point which the scar tissue visually broke and a major deflection in the tracing occurred. Tensilc strength was then calculated by dividing the observed breaking strength by the measured cross-sectional area of the scar. Since breaking and tensile strength curves were similar with time only tensile strength data was reported.
269
15 or 2 5 p m diameter (3M, St Paul, MN) were suspended in 2 ml isotonic saline with 0.01% Tween 80 in a microsphere injecting vial (VWR Scientific, San Francisco, CA). This suspension was mixed vigorously with a vortex mixer and injected into the left ventricle within 6U sec using two 20 ml syringes of saline injected through the microsphere injector. Concurrent to the microsphere injection, blood was withdrawn from the abdominal aorta via the femoral arterial catheter attached to a pump set at 12.0 ml/ min starting 15 sec before, and until 45 sec after injection of microspheres; 10 min later full thickness samples of the wound sites were removed with a n 8 nim punch, dried with a towel and weighed. Normal skin samples were obtained from areas adjacent to each wound as controls. The pigs were humanely euthanized with T-61" at the completion of the study. The tissues were placed into scintillation \rials and radioactivity determined with a gamma counter (Model 1282 Compugamma CS, LKB Wallace, Finland). Total radioactivity was determined from the blood sample removed from the abdominal aorta. Blood flow to each site was calculated by dividing the counts per minute (CI'M) of the tissue by the total CPM of the blood sample multiplied by the rate of blood withdrawal (12 mI/ min), that is (CPM,,,/CPM,,,,,,,,x7 2 ml/min. Values were expressed as ,ul/min.
Statistical Analysis Sampling Wounds for Blood Flow Blood flow to the wound was measured with radioactive microspheres according to the method described by Saxena and Verdouw (1985). Anesthetized pigs were placed in lateral recumbency and the left carotid artery and a femoral artery a n d vein were isolated for cannulation. A Swan-Ganz catheter (Edwards Div., Baxter Healthcare Corp., Santa Ana, CA) was placed into the pulmonary artery via the femoral vein for thermodilution cardiac output; a pigtail aortography catheter was advanced into the left ventricle via the left carotid artery for radioactive microsphere injection; and a pressure catheter was placed into the abdominal aorta via the femoral artery for blood withdrawal and central arterial pressure measurement. Mean aortic blood pressure was maintained at approximately 100 mmHg and cardiac output was determined by thermodilution techniques prior to microsphere injection. Cerium-141 labeled microspheres ( - 4 N O " ) either
The treatment groups were conipared with tlie vehicle treated or control groups of the same pigs. The data were first analyLed by analysis of variance. When the overall F-test showed a signlficant difference among groups, individual group means were compared using the Bonferroni multiple comparison procedure or Student's t-test. Sample size varied according to treatment group but in all cases the N (number of wounds per group) was at least 7.
RESULTS
Strength of Granulation Wounds Treatment on each row of wounds were equally divided between rhTGF-P1 and vehicle in order to eliminate bias d u e to anatomic variability from site to site First, the strength ot wounds were determined 4, 7, 10 and 14 days after surgery and
BECK r t nl
270
compared untreated or veliicle treated wounds with wounds that had received one application of rhTGFbl (250 ng/wound) at the time of wounding. The cross-sectional area of the trimmed wounds (see Fig. l), measured just before tensometry, decreased from day 4 to 14 in all groups but remained similar among the untreated, vehicle and rhTGF-PI treated groups at any given time. The mean values for tensile strength increased with tiine in all groups with higher values in rliTGF-pl than untreated or vehicle control at the early time points (Fig. 2). rIiTCF-Pl increased tensile strength ( P < 0.01) on d a y 4 when compared to wounds treated with vehicle alone or left untreated and when compared to the untreated wounds on day 7 (Pcar ( A ) 111 tlie middle 5urby tattoo ( R ) ( 2 ) Trimmed th margin of scar ( A ) at both ) Sidt. v i t v c)t ( 2 ) showing t i o n ot scar ( A ) and tattoo
1
days 4 and 7 atter wounding and treatment w t h 2 5, 25, 231 and 2500 ng rhTGF-pllwound. Values from each rhTGF-pl treated group were expressed as a percentage of the mean vehicle control of the same pig (Table 1) Tensile strength at day 7 was greater ( P < O 01) for wounds treated with 250 and 2500 ng rhTGF-P1 than vehicle TABLE 1 Fftect5 of rhTGb-BI on Strength of Granulation Wounds I r n d r strength was measured from wounds treated with vehicle or rhTGF-BI once at the time of surgery and allowed to heal 1 or 7 d a \ \ Values from each dose group and time point represent data from one pig for a total of eight pigs in this study Tensile strength ("10 of control)" Dav 4 Dav 7
Dose of rhTGF-Bl (ngiwound)
104 (12 X) 97 (11 4) 160 (33 2) 114 (13 Y)
2.5
25 231
2500
115 (4 6 ) 89 ( 5 1 ) 149 (9 7)' 730 (11 3 ) '
'Values from each rhTGF-Dl treated wound were adjusted for the mean all the vehicle control measurements bv expressing the response a 5 a percent of the mean vehicle control which wa3 asslgned t o be 10(1% Data are expressed as mean ( f SEM). Number ot wounds for each dose group and time point was at least 12 " P < (1.01 rhTGF-Pl vs vehicle. 01
2
3
4
5
6
271
rhTGF-Pl ENHANCES ULCER WOUND TENSILE STRENGTH
" .
I
4 DAYS
-.-
7 DAYS
10 d A Y d
14 DAYS
DAYS AFTER WOUNDING
FIGURE 2. Time response changes of tensile strength (g/mm') in pig dermal ulcer wounds. Wounds were untreated or received vehicle alone or 250 ng of TGF-Bl/wound at the time of surgical wounding. Data is expressed as mean k SEM. -0- Untreated, rhTGF-BI. *P
