Hobson KG, DeWing M, Ho HS, Wolfe BM, Cho K, Greenhalgh DG. Expression of Transforming Growth Factor β1 in Patients With and Without Previous Abdominal Surgery. Arch Surg. 2003;138(11):1249-1252. doi:10.1001/archsurg.138.11.1249
Transforming growth factor β1 (TGF-β1) plays an important role in the formation of adhesions after abdominal operations.
Prospective, observational study.
University-based, tertiary referral center.
Patients undergoing elective open abdominal operations were recruited and divided into 2 groups. Twenty-two patients with a history of abdominal surgery were designated as study patients, and 10 patients with no history of abdominal surgery served as controls.
Samples of normal peritoneum, peritoneal scar tissues, and serum were obtained from all patients at the time of surgery.
Main Outcome Measures
Samples were assayed for total TGF-β1 expression using an enzyme-linked immunosorbent assay.
Scar tissues expressed significantly greater amounts of TGF-β1 (0.47 pg/µL) compared with normal peritoneal tissue from both study patients (0.29 pg/µL; P = .03) and controls (0.17 pg/µL; P = .002). Serum TGF-β1 levels were also higher in study patients (1.71 pg/µL) compared with controls (1.22 pg/µL; P = .02). Neither adhesion nor serum TGF-β1 expression correlated with time since last operation, total number of previous operations, or severity of intra-abdominal adhesions.
These results suggest that TGF-β1 may play an important role in human peritoneal adhesion formation.
PERITONEAL ADHESIONS are a major cause of morbidity after laparotomy. A recent review estimated that the treatment of adhesion-related complications in the United States in 1994 costs $1.3 billion.1 Complications related to postsurgical adhesions include pain, infertility, and bowel obstruction, a complication that led to the hospitalization of more than 436 000 female patients in the United States in 1990.2 While most patients with abdominal adhesions develop symptoms years after the initial operation, as many as 38% develop acute intestinal obstruction within the first year.3 Even after surgical lysis of adhesions, adhesion-related complications are estimated to recur in 11% to 21% of patients.4 In addition, adhesions make subsequent intra-abdominal operations more difficult, and put the patient at higher risk for complications, such as enterotomy, fistula, or injury to other intra-abdominal organs.
Transforming growth factor β1 (TGF-β1) is a growth factor with a wide variety of tissue-specific effects. It has been shown to stimulate the production of extracellular matrix by 2 mechanisms. First, TGF-β1 stimulates fibroblasts to produce many proteins, including collagen, fibronectin, and integrins. It also decreases the production of proteins whose function is to degrade the extracellular matrix, such as collagenase and heparinase.5 It is because of these effects that TGF-β1 has been implicated in the pathogenesis of several diseases involving abnormal fibrin and collagen deposition, including hypertrophic scarring, pulmonary fibrosis, and nephrosclerosis.6- 8 Intraperitoneal TGF-β1 injection leads to increased adhesion formation, and high levels of TGF-β1 have been found in the fibroblasts, inflammatory cells, and endothelial smooth muscle cells of peritoneal adhesions.9,10 Studies of peritoneal wounding in rats have demonstrated decreased adhesion formation after intraperitoneal injection of blocking antibodies specific for TGF-β1.11,12 Nonetheless, the effects of TGF-β1 are not all deleterious. Studies have demonstrated decreased healing in mice that have a deletion in the TGF-β1 gene compared with their homozygous littermates.13 While TGF-β1 has been described as an important mediator of peritoneal adhesion formation in rodents, its role in the formation of human intra-abdominal adhesions remains undefined.
Patients with hypertrophic cutaneous scars express higher systemic levels of TGF-β1 than patients with normally healing wounds.14 This suggests that the TGF-β1 response may occur on more than just a local level. An increase in systemic levels in patients with abdominal adhesions could indicate a propensity for more severe scarring in patients with a history of abdominal operations.
Because of the evidence of the importance of TGF-β1 in the pathogenesis of scars and peritoneal adhesions in animal models, we hypothesized that TGF-β1 levels would be higher in the tissue of patients with a history of abdominal surgery than in patients with no prior abdominal surgery. We also hypothesized that some of the inflammatory cells that express high amounts of TGF-β1 would likely escape to the systemic circulation rather than be confined to the irritated mesothelial lining. This could lead to elevated systemic levels of TGF-β1 in the plasma of patients with previous abdominal operations.
Subjects were recruited from an outpatient surgical clinic at the University of California–Davis Medical Center (Sacramento, Calif). The study protocol was approved by the human subjects review committee of the University of California–Davis. All subjects provided informed consent prior to enrolling in the study. Patients scheduled for elective open abdominal surgery were considered eligible for enrollment. Exclusion criteria included known active cancer or planned laparoscopic procedure. Study patients were defined as patients with a history of at least 1 abdominal operation. Previous operations ranged in complexity from appendectomies and cholecystectomies to partial gastrectomies and multiple ventral hernia repairs. Control patients were defined as patients undergoing elective open abdominal operation with no history of any invasive abdominal procedure.
Blood samples were collected from each patient via an intravenous line at the time of surgery, after the abdominal fascia had been opened. Samples were collected and immediately centrifuged at 30 000 rpm for 5 minutes at 4°C. The serum was then aspirated and stored at −80°C until analyzed. The clot was disposed of. A small sample of normal peritoneum was collected from all patients. For control patients, this sample was harvested from an area adjacent to the midline incision to minimize the amount of exposed, nonperitonealized surface postoperatively. For study patients, normal peritoneal tissue was harvested from any point distant from the site of adhesions. Peritoneal adhesive scar tissue was also collected from each study patient at the time of surgery from the area of most dense adhesions. Specimens were snap frozen in liquid nitrogen and stored at −80°C until later analysis.
Patient medical records were reviewed at the time of surgery. Patient sex, age, and nature of surgery were all recorded. The number and nature of all previous operations was recorded for all study patients, as well as the time elapsed since the most recent surgery. Adhesion severity was assessed by the operating surgeons and assigned a number from 1 to 5 according to a predetermined scale (Table 1). Platelet counts were determined from either intraoperative blood samples or as part of routine postoperative laboratory studies.
Serum and peritoneal TGF-β1 protein levels were then assessed by enzyme-linked immunosorbent assay (ELISA) using a kit specific for TGF-β1. For peritoneum and peritoneal scar tissues, previously frozen tissues were placed in a mild lysis buffer, EDTA, and aprotinin, and homogenized with a Tissue-Tearor homogenizer (Biospec Products Inc, Bartlesville, Okla) to produce a soluble cell extract. Total protein concentrations were determined by spectrophotometry, and specimens were subsequently brought to equal concentrations by dilution with the appropriate amount of mild lysis buffer. All specimens were acidified for 1 hour to facilitate quantification of total TGF-β1 level, followed by neutralization immediately prior to conducting the ELISA. Samples were then analyzed by ELISA with serial dilutions of known quantities of recombinant TGF-β1, used as a positive control and standard to quantify the absolute TGF-β1 protein content of each sample. All specimens were analyzed in duplicate, and the assay was repeated at least 3 times.
Values are expressed as mean ± SEM. All TGF-β1 results are given as picograms per microliter. These results represent relative levels of TGF-β1, since the amount of protein isolated from each tissue was standardized based on spectrophotometry. In other words, TGF-β1 levels are expressed as picograms per equal amounts of protein. Results were compared using paired and unpaired t tests where appropriate. An analysis of variance and linear regression were used to investigate associations. P<.05 was considered statistically significant.
From April through December 1999, tissues were collected from 22 patients with a history of abdominal surgery, and from 10 control patients. Group characteristics were generally similar except for the surgical history (Table 2). All patients underwent elective open abdominal procedures, ranging from ventral hernia repair to gastric bypass.
Levels of TGF-β1 were significantly elevated in scar tissue (0.47 ± 0.06 pg/µL) compared with healthy tissue from either study (0.29 ± 0.05 pg/µL; P = .03) or control (0.17 ± 0.05 pg/µL; P = .002) patients. While the TGF-β1 protein expression in healthy tissue from study patients (0.29 pg/µL) was more than 65% higher than in healthy tissue from controls (0.17 pg/µL), this difference did not achieve statistical significance (P = .11). The levels expressed in scar tissue did not correlate with the severity of adhesions (r2 = 0.10; P = .40).
Serum TGF-β1 levels were significantly higher in patients with a history of abdominal surgery (1.71 ± 0.13 pg/µL) than in control patients (1.22 ± 0.19 pg/µL; P = .02). However, the serum TGF-β1 expression did not correlate with the amount of time since the last surgery (r2 = 0.02; P = .61) or with the total number of previous abdominal operations (r2 = 0.001; P = .85). Serum TGF-β1 levels also did not correlate with the severity of adhesions as graded by the operating surgeon (r2 = 0.15; P = .09). In the scarred tissue itself, we again found no correlation between TGF-β1 expression in scar tissue and time since previous operation (r2 = 0.03; P = .66), the number of previous abdominal operations (r2 = 0.07; P = .43), or even the severity of adhesions (r2 = 0.10; P = .40).
Other investigators have suggested that the use of plasma for determination of circulating TGF-β1 levels may be subject to interference from degranulation of platelets rich in TGF-β1, leading to falsely high levels of TGF-β1. This results in high TGF-β1 levels that correlate more consistently with the platelet count than with the total amount of circulating TGF-β1.15 However, platelet counts did not vary significantly between the 2 groups studied (platelet count, 297 ×103/µL) in patients with previous surgery compared with 242 ×103/µL in patients undergoing surgery for the first time (P = .33), nor was there any correlation between absolute serum TGF-β1 levels and platelet count (r2 = 0.073; P = .33).
Intra-abdominal adhesions contribute to significant morbidity and place a significant burden on the medical system and society. In this study, we confirmed that human peritoneal adhesions contain high amounts of TGF-β1 compared with normal tissue. We also confirmed that the increased TGF-β1 expression in scar tissue persists for several years after the initial injury, as there was an average of 2.9 years and as many as 14 years from the time of the previous operation to the current operation. Even the normal tissue in patients with adhesions tended to express increased levels of TGF-β1 compared with controls, although this trend did not quite reach statistical significance. Perhaps the small number of control patients in our study prevented this seemingly large difference from becoming statistically significant.
Holmdahl et al16 recently reported increased TGF-β1 expression in the peritoneal scar tissue of patients with more dense, extensive adhesive disease than in the tissue of those with more isolated, easily lysed adhesive disease. In contrast, our study did not demonstrate any correlation between severity of adhesive disease and the level of TGF-β1 expressed in the scar tissue. This may be due to differences in technique, as Holmdahl et al found the increase only in the active fraction of TGF-β1 (not previously acidified) but found no difference in total TGF-β1 expression. Our study involved the use of only acidified samples, which measured the total (both active and latent) TGF-β1 concentration.17 The measurement of active TGF-β1 may have revealed a difference, such as that found by Holmdahl et al. Our failure to detect a significant correlation may also have been due to the small sample size. With the recent major advances in laparoscopic surgery, it is becoming less and less common for patients with no history of abdominal operations to undergo elective open surgery. This makes recruitment of this group of control patients very difficult.
Notable is the presence of significantly elevated TGF-β1 levels in the serum of study patients. While this might be expected during the acute phase (and all patients were sampled in the acute phase, while undergoing major abdominal surgery), it is somewhat surprising that TGF-β1 levels stayed elevated in the systemic circulation long after the intra-abdominal scarring had ceased. Bakkum et al18 demonstrated that surgically induced adhesions in rats continued to mature for approximately 8 days postoperatively. However, after the initial inflammatory response resolved, they described remarkable consistency in the characteristics of the peritoneal adhesions over the next year. This suggests that peritoneal scar tissue becomes quiescent after a certain time. Clinical observations have confirmed that adhesions appear to be most severe up to a month after operation. However, after approximately 1 month, the inflammatory process subsides, and the adhesions progress minimally.19 Nonetheless, the TGF-β1 levels in both the scar tissue and in the serum remained elevated in the study patients compared with controls long after the instigating operation, a time when scars have already organized, and active scar formation has long since ceased.
One would suspect, then, that the amount of TGF-β1 expressed in the serum would correlate either with the time since the previous inflammatory stimulus or with the number of previous inflammatory stimuli to which the patient has been subjected. However, this study demonstrated no correlation between serum TGF-β1 levels and the number of previous operations or the number of months since the last operation, nor between serum TGF-β1 levels and severity of adhesions as graded by the operating surgeons. One potential source for error could be the use of serum instead of plasma. As mentioned previously, Kropf et al15 recommend the use of plasma rather than serum to determine systemic TGF-β1 levels, as the process of isolating serum involves clotting, which involves significant platelet degranulation. As platelets are rich in TGF-β1, degranulation during clot formation releases large amounts of TGF-β1, leading to artificially elevated TGF-β1 serum levels that correlate more closely with the absolute platelet count rather than with the amount of TGF-β1 circulating in vivo.15 To determine if this bias existed in our study, we evaluated the platelet counts of all study and control patients on the day of surgery. Platelet counts in the study group that were significantly higher than those in the control group would be expected to lead to artificially higher levels of TGF-β1 in the serum of the group with the higher platelet count. However, there was no significant difference in the total number of platelets between the study and control groups. Thus, the observed increase in TGF-β1 expression in the plasma of study patients relative to controls was unlikely the result of differential platelet degranulation.
The significance of the increased serum TGF-β1 expression is unclear. Increased serum levels of TGF-β1 may predispose patients with previous abdominal operations to form more severe abdominal adhesions. Confirmation of an increased propensity to scar would require large longitudinal studies with serial measurements of serum TGF-β1 and severity of adhesions from patients undergoing multiple surgeries. However, such a study would be logistically difficult, as it would be nearly impossible to recruit a sufficient number of subjects to provide adequate power.
Abdominal adhesions remain a significant problem after surgery, creating a large burden on the medical system and society. The findings of this study suggest that TGF-β1 may play a role in the formation of intraperitoneal adhesion tissue after abdominal operations. While this study does not prove that TGF-β1 is a cause of increased adhesions, there are animal studies that suggest that TGF-β1 antibodies reduce scarring.13 The inhibitory effects of TGF-β1 antibodies, in addition to increased TGF-β1 levels in situations where there is increased scarring, tend to suggest that its role in adhesion formation is important. Therapeutic measures involving control of the TGF-β1 pathway may prove valuable in preventing some of the significant morbidity associated with this common postoperative complication.
Corresponding author and reprints: David G. Greenhalgh, MD, Department of Surgery, Shriner's Hospital for Children, 2426 Stockton Blvd, Sacramento, CA 95817.
Accepted for publication April 27, 2003.