| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
NEW APPROACHES TO THE TREATMENT OF HEPATIC MALIGNANCIES |
and Melphalan
From the Division of Surgery and Surgical Oncology, Robert Rössle Hospital and Tumor Institute, Max Delbrück Center for Molecular Medicine (PH, EL, CK, PMS); and Institute of Microbiology, Charité (A-HR, RS), The Humboldt University at Berlin, Germany.
Correspondence: Address correspondence and reprint requests to: Peter Hohenberger, MD, Division of Surgery and Surgical Oncology, Robert Rössle Hospital, Charité, Campus Berlin-Buch, Lindenberger Weg 80, D-13125 Berlin, Germany; Fax: 49-30-9417-1439; E-mail: hohenberger{at}rrk-berlin.de
ABSTRACT
Background: Isolated limb perfusion (ILP) with recombinant human tumor necrosis factor-
(rhTNF-) and melphalan harbors the risk of septic shock–like syndrome. Pentoxifyllin (PTX) produced a beneficial effect on cytokine response and survival in animal experiments of septic shock, and we were interested to explore its effect during TNF-ILP in humans.
Methods: Eighteen consecutive patients underwent TNF-ILP and received PTX (30 mg/kg/day), whereas another 13 consecutive patients did not. PTX was given systemically after the limb extracorporeal circulation was started. Cardiac index, systemic vascular resistance (SVR), and pulmonary vascular resistance were recorded via a Swan-Ganz catheter. Blood levels of TNF-, interleukin-6, procalcitonin, and lipopolysaccharide-binding protein were determined before, during, and after ILP.
Results: After reperfusion, systemic levels of TNF- were significantly less increased in the PTX group (peak, 2.8 vs. 1.3 ng/mL; P < .05), as were interleukin-6 values (peak, 68 vs. 22 pg/mL; P < .02) and lipopolysaccharide-binding protein plasma levels (peak, 215 vs. 105 µg/mL; P < .03). Differences in cardiac index, SVR, and mean arterial blood pressure were not significantly different. Norepinephrine or dobutamine to maintain SVR was less required in the PTX group.
Conclusions: PTX attenuates systemic cytokine production and influences components of the systemic inflammatory response after TNF-ILP. PTX may play a beneficial role in the management of septic shock–like syndrome, particularly in patients with leakage from the ILP circuit.
Key Words: Isolated limb perfusion • TNF- • Septic shock • Pentoxifyllin • SIRS • Cytokine
Isolated limb perfusion (ILP) with recombinant human tumor necrosis factor- (rhTNF-; tasonermin) and melphalan is a highly effective treatment for soft tissue sarcoma of the extremities. In a neoadjuvant setting, ILP contributes to radical resection of locally advanced sarcomas and long-term limb salvage.1,2 In the treatment of malignant melanoma, TNF-ILP resulted in higher complete response rates and longer median disease-free survival in comparison to melphalan alone in a randomized study.3 Despite the fact that overall survival was not improved, TNF-ILP clearly remains the option for second-line treatment of extremity melanoma if perfusion with melphalan alone fails.4
Beyond the technical aspects of establishing an extracorporeal circuit and leakage control, a more widespread application of TNF perfusion is hampered by the risk of systemic toxicity. The septic shock–like syndrome yields episodes of severe hypotension that can be counterbalanced only by the application of catecholamines. The incidence of these episodes has been reported from 12% to 29%.3,5–8 The incidence and severity of the septic shock–like syndrome correlates with the occurrence of leakage from the limb to the systemic circulation and subsequent activation of the cytokine network.9 One of the main denominators is the systemic level of circulating TNF- being the major stimulus of interleukin (IL)-1, IL-6, procalcitonin (PCT), and other components of the systemic inflammatory response (SIRS). From animal experiments, it is known that pentoxifyllin (PTX) reduces the concentration of circulating TNF- and produces a beneficial effect on survival in septic shock models by using inoculation of bacteria or bacterial compounds known to trigger TNF- release.11,12 Thus, we explored the effects of PTX on SIRS and subsequent cardiovascular alterations in TNF-ILP in humans to discover the following: (1) influence of PTX on plasma levels of TNF-; (2) influence on the response of IL-6, PCT, and lipopolysaccharide (LPS)-binding protein (LBP; a class 1 acute phase reactant known to bind and transfer bacterial LPS13) to TNF-ILP; (3) whether changes of cardiovascular parameters (incidence of hypotensive episodes) after surgery are influenced by PTX; and (4) the requirement to administer norepinephrine or dobutamine to counterbalance hypotensive episodes.
PATIENTS AND METHODS
Patients
Eighteen consecutive patients (10 women and 8 men; median age, 59 years; range, 29–80 years) underwent ILP for locally advanced soft tissue sarcoma (n = 9) or malignant melanoma (n = 9). All patients underwent leg perfusion, except one patient, who underwent arm perfusion. The study protocol had been approved by the local ethics committee (AA2, Humboldt University of Berlin), and the patients had given their written, informed consent.
Isolation Perfusion and Application of Drugs
The perfused limb volume was measured with the water displacement method.14 The detailed technique of ILP has been described previously.8 Shortly after exposition and cannulation of the major artery and vein of the limb, extracorporeal circulation was established with a roller pump and heat exchanger (Stöckert Corp., Munich, Germany). Gas exchange was achieved with a bubble oxygenator (Baxter Corp., Utrecht, The Netherlands). The perfusate temperature was 39°C to 41°C, and the volume of the perfusate was kept constant at approximately 700 mL. During ILP, tissue temperature, measured by needle probes, was ≥38°C and was kept <40.5°C. Perfusion time was 90 minutes. Leakage control was performed by injection of indium-111–labeled autologous erythrocytes and 99mTc-labeled albumin to the limb circuit and continuous monitoring of the systemic circulation.15 After perfusion, the limb was rinsed with 2 to 3 L of hydroxyethyl starch until no further reduction of the radiopharmaceutical activity in the limb was achievable. The rhTNF (supplied by Boehringer Ingelheim Corp., Ingelheim, Germany) was given at a dose of 3 mg (upper limb) or 4 mg (lower limb). The melphalan dosage was 10 mg/L of perfused limb volume, as described previously.16
The total dose of PTX was 30 mg/kg/day. Half of the dose was given as a 4-hour short-term infusion immediately after the ILP circuit was started. The second half of the drug followed as a continuous infusion for another 20 hours (Fig. 1).
|
Levels of IL-6 were analyzed with enzyme-linked immunosorbent assay (ELISA) kits (DPC Biermann, Bad Nauheim, Germany) with reference values of <25 pg/mL. To determine PCT, an immunoluminometric assay was used (LUMItest PCT; Brahms Diagnostica, Berlin, Germany; reference range, 0 to .5 ng/mL). TNF- levels were measured with Quantikine ELISA (DPC Biermann). The data of one patient with leakage of 5.4% from the ILP circuit to the systemic circulation was excluded from analysis of serum TNF levels. Serum LBP was assessed by an ELISA using the monoclonal antibodies 1E8 and 2B5, as described,17 with a detection level threshold of 1 ng/mL and a reference range of up to 17.5 mg/L.
Cardiocirculatory and Clinical Evaluation
After the perfusion, the patients were transferred to the intensive care unit with cardiopulmonary monitoring for at least 24 hours. Heart rate (beats/min) and systolic, diastolic, and mean arterial pressure (MAP; mm Hg) were continuously recorded. A Swan-Ganz pulmonary artery catheter was placed for invasive perioperative monitoring of cardiac index (CI; L/minute/m2), pulmonary arterial pressure, and pulmonary capillary wedge pressure. Pulmonary and systemic vascular resistance (SVR; dynes/sec/cm5) were recorded from the onset of the operation until reperfusion of the limb (Thermodilution catheter; Arrows, Reading, PA).
During and after ILP, the patients received crystalloid fluids to stabilize blood pressure and heart rate. Dopamine (3 µg/kg/minute) was administered to all patients for the first 48 hours after surgery to avoid renal failure.18 Few patients required additional application of sodium bicarbonate for urinary alkalization to handle excessive myoglobinemia and myoglobinuria. The body temperature was recorded by a rectal probe for 24 to 48 hours after ILP.
Controls
For comparison of blood values, serum probes were used that were drawn from 13 consecutive patients (melanoma, n = 6; sarcoma, n = 7) who all had undergone TNF-ILP of the leg according to the same protocol except for the application of PTX. For comparison of cardiovascular parameters, we used the data of 83 consecutive patients perfused earlier, as well as those of 26 consecutive patients perfused later who all had been monitored after surgery with the same standards.8
Statistical Analysis
Data are presented as median values or mean ± SD. Nonparametric tests according to the Mann-Whitney U-test, 
2 test, and Wilcoxon’s signed rank test were performed to compare independent and related samples. For correlation analysis between variables, Spearman’s correlation coefficient was calculated. Statistical analysis was performed with SPSS for Windows, release 7.5 (SPSS Inc., Chicago, IL).
RESULTS
Leakage Versus No Leakage
There were two patients with leakage >1% from the ILP circuit in the PTX group and in the control group. The mean systemic leakage rate was 1.4% ± 3.2% of the radiotracer applied to the extracorporeal circulation.
Systemic TNF Levels
Systemic levels of TNF- were increased only during the immediate postoperative phase after tourniquet release and reperfusion. Levels were significantly less increased in the PTX group at 30 minutes after reperfusion (peak level, 2.8 vs. 1.3 ng/mL; P < .05). During ILP and 4 hours after ILP, the values were at the detection threshold (Fig. 2).
|
|
The differences in LBP were most prominent and long-lasting, and there were lower values in the PTX group in comparison to controls. Peak serum concentrations increased 12.2-fold in the control group and only 5.5-fold in the PTX group. Levels were significantly lower at 24 hours (peak levels, 215 vs. 105 µg/mL) and 48 hours (peak level, 177 vs. 84 µg/mL; P < .01) after ILP. Even at 96 hours after ILP, there were still significantly increased values of >100 µg/mL when mean concentrations in the PTX group had already returned to normal (Fig. 3).
Clinical Course and Cardiovascular Parameters
After surgery, all patients developed increased body temperature. There was no difference in the maximum body temperature during the first 24 hours (38.5°C vs. 38.6°C) between the PTX group and the control group.
When ILP started and during its course, there was no difference in MAP between patients with PTX and those without. There were no significant differences between the PTX group and controls in the heart rate, CI, or SVR at any time point assessed after surgery. Immediately after reperfusion, patients in both groups developed a severe decrease of MAP; however, there was a significant increase in the control group in comparison to the PTX group for the first four to six postoperative hours (P = .003). Eight hours after surgery and thereafter, similar values were measured in both groups (Fig. 4).
|
2(3,75); P = .11] and 6 (23%) of 26 patients undergoing TNF-ILP after the PTX study. DISCUSSION
PTX, a methylxanthine derivative and nonspecific phosphodiesterase inhibitor, reduces TNF release from macrophages and T lymphocytes by suppressing TNF gene transcription.19–22 Furthermore, it reduces the production of free oxygen radicals23 and the expression of surface adhesion molecules in neutrophils, as well as the release of thromboxane.24 PTX might improve the host immune response and hemodynamics in SIRS.25 PTX has been highly protective in different animal models of endotoxin shock, sepsis, ischemia, and reperfusion if administered before inoculation or shortly afterward.11,12 PTX given intraperitoneally improved survival and reduced the concentration of circulating TNF in taurocholate-induced pancreatitis in rats.26 PTX modulated the activity of the reticuloendothelial system in rabbits with controlled blood loss and infusion of nonlethal doses of endotoxin. Pretreatment with PTX improved blood clearance of Escherichia coli and reduced microbial seeding (colonization) in critical organs.27
Several studies tried to transfer these experimental data to the treatment of septic shock in humans. Twelve patients were treated with PTX at 1 mg/kg/hour for 24 hours, and the SIRS was measured by cytokine levels and cardiovascular parameters and compared with matched controls.28 At 24 hours, when the only laboratory controls were performed, systemic TNF- levels, but not levels of IL-6, were significantly less in the PTX group, whereas SVR was significantly higher in the PTX group than in controls. In another 19 critically ill patients, PTX (5 mg/kg for 3 hours) significantly improved CI despite unchanged pulmonary artery occlusion pressure; however, this was not an outcome study.29 In a placebo-controlled, double-blind, randomized trial of 16 patients with septic shock, Zeni et al.30 used PTX (initially 1 mg/kg intravenously followed by 1.5 mg/kg/hour for 24 hours) and measured laboratory values repeatedly. Systemic serum TNF- levels were significantly lower in the PTX group than in controls. IL-6 and IL-8 values also were less increased. No significant differences in oxygenation or hemodynamic parameters could be observed. A randomized, double-blind, placebo-controlled study administered PTX (maximum dose, 1.8 g/day for 28 days) in 51 patients for severe sepsis. The multiple organ dysfunction score improved after 4 days in the PTX group; however, it reached significance only on days 14 and 17. A significant improvement of the alveolar/arterial difference in the partial pressure of oxygen, the fraction of inspired oxygen, and the pressure-adjusted heart rate was observed. Serum endotoxin levels, as well as peak concentrations of TNF- or IL-6, showed no significant changes.31
The reason why clinical PTX studies failed in patients referred to an intensive care unit might be that treatment began rather late after inoculation and patients were already on their way to manifest sepsis. Even had PTX been effective, it was probably far too late to overcome the adverse effects of TNF- released from neutrophils and macrophages and the already-started cytokine cascade. In the limb perfusion setting of rhTNF-/melphalan, the inoculation time is exactly known. Crossing of TNF- from the ILP circuit to the systemic circulation becomes evident by radioactive leakage control.5 Systemic peak TNF- values can be detected 5 to 10 minutes later and are by a factor of 10 to 30 higher than without leakage.9 If no leakage occurs during ILP, the point of inducing SIRS can be determined exactly at reperfusion after tourniquet release and vessel repair.
Consequently, our rationale behind the split dosage of PTX application was (1) to have a loading dose of PTX given during ILP and (2) to continue with PTX for another 24 hours according to cytokine profiles developing after ILP. TNF- is measurable only during the first postoperative hour and is followed by IL-6, which peaks at 4 to 8 hours. PCT follows next; it is detectable at 4 hours at the earliest, with a peak at 24 hours, and returns to normal at 96 hours.5,32 LBP without attenuation by PTX is still significantly increased at 96 hours, indicating that TNF-ILP induces long-term activation of the cytokine cascade. Also, in studies of clinical application of PTX for sepsis, protracted infusion28,30 for up to 28 days was used.31
It is an important observation that the concentration of LBP is significantly increased in the serum in patients with SIRS.33 LBP is a hepatic secretory protein capable of binding the bacterial cell wall product endotoxin and directing it to its cellular receptor CD14. Induction of LBP expression is transcriptionally regulated and is dependent on stimulation with IL-1ß, IL-6, and dexamethasone.13 LBP is released by both LPS and hepatocytes after stimulation with TNF or IL-6 and seems to be the candidate marker to assess whether long-lasting stimulation is present after isolated organ perfusion with rhTNF-.
Regarding cardiovascular reactions after TNF-ILP, we could confirm previous data and those of other investigators.5,7,34 A pronounced increase in heart rate 1 hour after reperfusion of the limb was observed. After 24 hours, the heart rate was still above the baseline level. In parallel, an increase in MAP could be recorded. PTX was also able to significantly limit the increase of MAP and to limit the increase in CI. This should contribute to a less problematic postoperative course often induced by a decrease of SVR. It could be discussed whether PTX itself induces an SVR response—one working mechanism postulated when patients are treated for peripheral vascular occlusive disease. However, two studies failed to show deleterious effects in patients with sepsis.30,35 In patients undergoing ILP for sarcoma or melanoma, peripheral vessel occlusive disease usually is ruled out before perfusion, and direct effects of PTX on SVR should not play a major role.
Patients undergoing TNF-ILP represent a rather homogenous population. This setting provides an excellent model to study the optimal timely onset of intervention in comparison to patients referred with escalating sepsis. Each group worldwide that performs TNF-ILP has experienced postoperative morbidity due to SIRS,36,37 and this was also reported after hepatic TNF perfusion, with a severe increase of IL-6 and IL-8 and an increase in MAP.38,39 PTX could be the co-drug that limits the negative effects of TNF application. Recently, PTX was used to clear TNF- in patients with membranous nephropathy, a glomerular disease caused by immune-mediated activation of the complement cascade and characterized by urinary excretion of TNF- and by proteinuria. PTX was able to significantly decrease plasma and urinary levels of TNF- and the extent of proteinuria, and 90% of the patients were in remission from their nephritic syndrome.40
The handling of patients undergoing TNF-ILP might be different between institutions and might influence the incidence of dobutamine or norepinephrine use. In our setting, both the anesthesiologists and the intensive care unit staff made their decisions without asking whether PTX had been applied. The use of catecholamines increased again after the PTX study phase but did not reach the incidence of the pre-PTX phase.
In conclusion, our results document that PTX attenuates systemic cytokine production and significantly influences short-acting and long-lasting mediators of SIRS after TNF-ILP. A randomized study would require many stratification factors (disease, age, level of perfusion, limb volume, and so on). In the ILP setting, PTX may play a beneficial role in the management of the septic shock–like syndrome, particularly in patients with leakage from the ILP circuit. Further investigations clearly are warranted, particularly if approaches to use TNF- in an isolated liver of lung perfusion setting are believed to be worthwhile.
Footnotes
Adding pentoxifyllin to the systemic circulation during isolated limb perfusion with tumor necrosis factor- (TNF-) and melphalan attenuated the postoperative systemic inflammatory response. A beneficial role for patients with leakage from the ILP circuit and broader application of TNF- in regional tumor therapy could result.
Received for publication October 3, 2002. Accepted for publication February 11, 2003.
REFERENCES
-induced circulatory shock: a study in patients treated for cancer with isolated limb perfusion. Crit Care Med 1996; 24: 1806–10.[CrossRef][Medline]
), interferon-gamma (IFN-
) and melphalan for melanoma stage III. Results of a multi-centre pilot study. Melanoma Res 1994; 4 (Suppl 1): 21–6.
This article has been cited by other articles:
 |
|
 |
|
  P. Hohenberger and W. M. Wysocki Neoadjuvant Treatment of Locally Advanced Soft Tissue Sarcoma of the Limbs: Which Treatment to Choose? Oncologist, February 1, 2008; 13(2): 175 - 186. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and controls (
). Values are given as mean + SD. *P < .05, top; *P < .02, bottom; significant differences between the groups after surgery.
). Significant postoperative differences between the groups are indicated by the symbols at right. Time points of serum examinations are according to 
) pentoxifyllin. Note the sharp increase of MAP in the early postoperative phase (1 to 4 hours) in the untreated group. Baseline indicates the timing of measurements related to the procedure.