- Email: [email protected]
Regional toxicity after isolated limb perfusion with melphalan and tumour necrosis factor- α versus toxicity after melphalan alone
EJSO 2001; 27: 390–395 doi:10.1053/ejso.2001.1124, available online at http://www.idealibrary.com on
Regional toxicity after isolated limb perfusion with melphalan and tumour necrosis factor- versus toxicity after melphalan alone B. C. Vrouenraets∗, A. M. M. Eggermont†, A. A. M. Hart‡, J. M. Klaase∗, A. N. van Geel†, O. E. Nieweg∗ and B. B. R. Kroon∗ Departments of ∗Surgery and ‡Radiotherapy, The Netherlands Cancer Institute/Antoni van Leeuwenhoek ziekenhuis, Amsterdam, and †Department of Surgery, University Hospital Rotterdam/Dr Daniel den Hoed Cancer Centre, Rotterdam, The Netherlands
Aims: To determine whether the addition of high-dose tumour necrosis factor- (TNF) to isolated limb perfusion (ILP) with melphalan increases acute regional tissue toxicity compared to ILP with melphalan alone. Methods: A retrospective, multivariate analysis of toxicity after normothermic (37–38°C) and ‘mild’ hyperthermic (38–40°C) ILPs for melanoma was undertaken. Normothermic ILP with melphalan was performed in 294 patients (70.8%), ‘mild’ hyperthermic ILP with melphalan in 71 patients (17.1%) and ‘mild’ hyperthermic ILP with melphalan combined with TNF in 50 patients (12.0%). Toxicity was nil or mild (grades I–II according to Wieberdink et al.) in 339 patients (81.7%), and more severe acute regional toxicity (grades III–V) developed in 76 patients (18.3%). A stepwise logistic regression procedure was performed for the multivariate analysis of prognostic factors for more severe toxicity. Results: On univariate analysis, ‘mild’ hyperthermic ILP with melphalan plus TNF significantly increased the incidence of more severe acute regional toxicity compared to normothermic and ‘mild’ hyperthermic ILP with melphalan alone (36% vs 16% and 17%; P=0.0038). However, after ILP using TNF no grade IV (compartment compression syndrome) or grade V (toxicity necessitating amputation) reactions were seen. Significantly more severe toxicity was seen after ILPs performed between 1991 and 1994 compared with earlier ILPs (33% vs 14%; P=0.0001). Also, women had a higher risk of more severe toxicity than men (22% vs 7%; P=0.0007). After multivariate analysis, prognostic factors which remained significant were: sex (P=0.0013) and either ILP schedule (P=0.013) or treatment period (P= 0.0003). Conclusions: Regional toxicity after ‘mild’ hyperthermic ILP with melphalan and TNF was significantly increased compared to ILP with melphalan alone. This may be caused by increased thermal enhancement of melphalan due to the higher tissue temperatures (39–40°C) at which the melphalan in the TNF-ILPs was administered or by an interaction between high-dose TNF and melphalan. 2001 Harcourt Publishers Ltd Key words: isolated limb perfusion; melphalan; tumour necrosis factor-alpha; melanoma; toxicity.
INTRODUCTION Isolated limb perfusion (ILP) is the treatment of choice in locally inoperable melanoma of the limbs, with complete responses in up to 55% of patients after ILP with melphalan.1 The addition of tumour necrosis factor-alpha (TNF), at a dosage up to 10 times that which can be tolerated after systemic administration, increases the
Correspondence to: Professor dr. B. B. R. Kroon, Department of Surgery, The Netherlands Cancer Institute/Antoni van Leeuwenhoek ziekenhuis, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands. Tel: +31 20 512 2552; Fax: +31 20 512 2554; E-mail: [email protected] 0748–7983/01/040390+06 $35.00/0
complete response rate to 70–90%.2–4 Moreover, the application of TNF in ILP has now been demonstrated to yield high response rates (over 80%) in locally advanced soft tissue sarcomas.5,6 Systemic toxicity after ILP using TNF, at first a major problem in some centres, is directly correlated to leakage of TNF into the systemic circulation during the procedure.6–8 When leakage is prevented and a thorough washout of the limb vasculature is performed at the end of ILP, only a short-lived (less than 1 hour) systemically measurable TNF peak is observed. In those patients, systemic toxicity is restricted to fever on the day of surgery and otherwise does not differ significantly from toxicity after ILP with melphalan alone,6,9 leaving regional toxicity as the dose limitating factor. 2001 Harcourt Publishers Ltd
REGIONAL TOXICITY AFTER ISOLATED LIMB PERFUSION Initial reports from Lie´nard et al.2,7 stated that highdose TNF administered by ILP does not add any acute regional toxicity compared to ILP with melphalan alone. This is important since a strong correlation has been demonstrated between the severity of the acute regional toxic reactions and the incidence of long-term morbidity, especially tissue fibrosis, muscle atrophy and limb malfunction.10 In other reports,11,12 however, the rapidity of onset, the overall duration and the severity of most regional side-effects appeared to be more prominent with the addition of TNF to melphalan. A retrospective, multivariate analysis was undertaken of all single normothermic and ‘mild’ hyperthermic ILPs performed at The Netherlands Cancer Institute, Amsterdam, and the Dr Daniel den Hoed Cancer Centre, Rotterdam, The Netherlands, to determine whether the addition of high-dose TNF to ILP with melphalan increases acute regional tissue toxicity compared to ILP with melphalan alone.
PATIENTS AND METHODS From 1978 to 1995, 530 patients underwent 668 ILPs for melanoma in our institutions. A computer-assisted database, containing all patient, tumour and treatment variables, allowed the selection of patients who underwent a single normothermic (37–38°C) or ‘mild’ hyperthermic (38–40°C) ILP with melphalan with or without TNF. Of patients having more than one ILP treatment, we considered only the first. During the study period different single ILP schedules were used for various indications. Until 1985, normothermic ILP with melphalan was performed both for ‘high-risk’ primary melanoma and for recurrent melanoma.13,14 From 1986 to 1994, all patients with ‘highrisk’ primary melanoma underwent ‘mild’ hyperthermic ILP with melphalan according to an EORTC trial protocol;15 whilst patients with recurrent melanoma were treated using our routine normothermic ILP schedule. After 1994, patients with ‘high-risk’ primary melanoma were no longer treated by ILP. In 1991, TNF and gamma-interferon (IFN) were introduced under ‘mild’ hyperthermic conditions in our ILP methodology for the treatment of recurrent disease in patients who were suitable candidates for this experimental protocol.3 ILP was conducted in the lower limb at either the external iliac or the femoropopliteal level, and in the upper limb at the axillary or brachial level. For iliac ILPs, the iliac and obturator lymph nodes were removed, and for the axillary procedure, an axillary lymph node dissection was performed. All ILPs were carried out with a routine dose of 10 mg melphalan/l perfused tissue for lower limbs and 13 mg/l for upper limbs.16,17 Melphalan was injected gradually (in 2–3 min) into the arterial line. The duration of all ILPs was 1 hour. Four temperature probes were inserted into subcutaneous tissue and muscle. In the ‘mild’
391
Table 1 Grading of acute limb toxicity according to Wieberdink et al.16 Grade I Grade II Grade III
Grade IV
Grade V
No subjective or objective evidence of reaction Slight erythema and/or oedema Considerable erythema and/or oedema with some blistering; slightly disturbed motility Extensive epidermolysis and/or obvious damage to the deep tissues, causing definite functional disturbances; threatening or manifest compartment syndromes Severe reaction which may necessitate amputation
hyperthermic EORTC trial ILPs,15 the melphalan was injected when the subcutaneous tissues of the distal part of the extremity had reached a temperature of 38°C. After 10 min of melphalan perfusion, the temperature was gradually elevated to 39–40°C and was kept constant for the rest of the 60 min of treatment. The TNF ILP protocol involved use of 4 mg of TNF for iliac and femoral ILPs and 3 mg for axillary and brachial ILPs.3 For 2 days before surgery, 0.2 mg IFN was injected subcutaneously; the same dose was given during the ILP. The whole ‘mild’ hyperthermic ILP with TNF lasted 90 min, during which the temperature of the perfusate was maintained at 40°C. IFN and TNF were injected successively as boluses into the arterial line when distal subcutaneous temperatures reached 38°C after which temperatures were gradually elevated. All temperatures were kept at 39–40°C during the last 60 min of ILP. Melphalan was administered in the routine dose 30 min after TNF injection. In some patients IFN was omitted as part of a multicentre trial.4 Specific details of our ILP techniques have been described elsewhere.14 No prophylactic fasciotomies were performed. Acute regional toxicity after ILP was graded according to Wieberdink et al.16 (Table 1). With all these ILP schedules, we strived to limit acute regional toxicity to Wieberdink grades II or III, meaning completely reversible reactions without long-term sequelae. The main aim of the present study was to examine the relation between ILP schedule and acute regional toxicity after ILP. A stepwise logistic regression procedure was performed for the multivariate analysis of prognostic factors for more severe acute regional toxicity (Wieberdink grades III–V). Ordinal variables were used linearly, but linearity was tested for. In testing a particular variable only those patients who had a missing value for that variable or for any variable already included in the model were excluded from the analysis. The inclusion and exclusion limit was set at PΖ0.15.
392
RESULTS A total of 415 patients were selected for this study: 309 women (74.5%) and 106 men (25.5%) with a mean age of 53 years (range: 17–89 years). A normothermic ILP with melphalan was performed in 294 patients (70.8%). Seventy-one patients (17.1%) underwent a ‘mild’ hyperthermic ILP with melphalan alone as part of the EORTC prophylactic ILP trial. Fifty patients (12.0%) were treated under ‘mild’ hyperthermic conditions with TNF and melphalan. Fourteen patients (3.4%) had grade I acute regional toxicity, 325 (78.3%) grade II, 71 (17.1%) grade III, three (0.7%) grade IV and two patients (0.5%) had a grade V reaction. Grade IV reactions occurred after two normothermic ILPs and after one ‘mild’ hyperthermic ILP with melphalan alone. Amputations were needed after one normothermic and one ‘mild’ hyperthermic ILP with melphalan alone. In Table 2, the patient and treatment variables tested for their association with acute regional toxicity are shown. On univariate analysis, the most significant factor predicting more severe acute regional toxicity was the period in which the ILP was performed, with more severe toxicity after ILPs performed between 1991 and 1994 compared with earlier ILPs (33% vs 14%; P= 0.0001). Women had an increased risk of more severe toxicity compared to men (22% vs 7%; P=0.0007). The ILP schedule was also found to be related to the severity of the acute regional toxic reactions after ILP (P= 0.0038), with a greater probability of grade III–V toxicity after ILP with TNF than after normothermic or ‘mild’ hyperthermic ILP with melphalan alone (36% vs 16% and 17%, respectively). Treatment period, sex and melphalan dose satisfied the inclusion criteria of PΖ0.15 in the multivariate stepwise logistic regression analysis (Table 3). Adjusted for these variables, the evidence for a relation between ILP schedule and acute regional toxicity disappeared (P= 0.89). Only treatment period (P=0.0003) and sex (P= 0.0009) retained their predictive value for toxicity. However, as treatment period and ILP schedule were strongly linked (Table 4), the power to detect differences between ILP schedules was markedly decreased. We therefore adjusted for sex, ILP schedule and melphalan dose instead of treatment period (Table 3). This decreased the significance of the treatment period as a prognostic factor for acute regional toxicity considerably (P=0.022), suggesting that the increased incidence of grade III–V toxicity in the period 1991–1994 could partly be explained by the use of TNF and higher melphalan doses during that period. In this model, the estimated odds ratio of the TNF schedule versus the other two schedules was 2.7 with a 95% confidence interval of 1.4–5.1. With more severe toxicity in 16% of ILPs without TNF, this would mean more severe toxicity in 34%
B. C. VROUENRAETS ET AL. Table 2 Variables used in the analysis of acute regional toxicity after ILP-univariate analysis of prognostic factors for more severe acute regional toxicity after ILP (grades III–V according to Wieberdink et al.)16 Variable
Sex Male Female Age (years) <40 40–49 50–59 >60 Dose of melphalan (mg) Unknown <75 75–99 100–124 >125 Venous pressure (cm) at 30 min Unknown <10 >10 Leakage to the body (%) Unknown 0 0.1–1 >1 Level of isolation Axillary/brachial Iliac Femoropopliteal ILP period 1978–1984 1985–1990 1991–1994 ILP schedule Normothermic ‘Mild’ hyperthermic TNF
Number Number of of patients (%) patients with grade III–V toxicity
P-value
0.0007 106 309
7 (6.6) 69 (22.3)
91 62 102 160
15 13 21 27
0.95 (16.5) (21.0) (20.6) (16.9) 0.38 1 99 113 147 55
17 17 31 11
(17.2) (15.0) (21.1) (20.0) 0.80
61 310 44
47 (15.2) 6 (13.6) 0.29
32 222 106 55
45 (20.3) 15 (14.2) 9 (16.4)
68 293 53
13 (18.8) 54 (18.4) 9 (17.0)
179 140 96
25 (14.0) 19 (13.6) 32 (33.3)
294 71 50
46 (15.7) 12 (16.9) 18 (36.0)
0.97
0.0001
0.0038
(95% CI: 21%–49%) of ILPs with TNF, adjusted for sex and melphalan dose.
DISCUSSION The selection of possible prognostic factors for acute regional toxicity in the present series was based on recent reports on this issue.18–21 The present study confirmed only the prognostic significance of sex, with women at increased risk for developing toxicity.10,18 This female risk factor may be attributed to the fact that
REGIONAL TOXICITY AFTER ISOLATED LIMB PERFUSION Table 3 Prognostic factors for acute regional toxicity after ILP in the stepwise logistic regression analysis Variable
Step 3∗ P-value
Step 3a† P-value
Sex Age Dose of melphalan Venous pressure Leakage to the body Level of isolation ILP period ILP schedule
0.0009 0.42 0.13 0.70 0.23 0.17 0.0003 0.89
0.0013 0.50 0.14 0.77 0.39 0.17 0.022 0.013
∗ Adjusted for in the following order: ILP period, sex, and dose of melphalan. † Adjusted for in the following order: sex, ILP schedule, and dose of melphalan.
Table 4 ILP period and ILP schedule ILP schedule Normothermic
‘Mild’ TNF Total hyperthermic ILP period
1978–1984 1985–1990 1991–1994
179 96 19
0 44 27
0 0 50
179 140 96
Total
294
71
50
415
women have a lower muscle/fat ratio than men. Since there is a higher melphalan uptake in muscle than fat,22 the muscle tissue of female patients is exposed to a relatively higher dose of melphalan. The major aim of the present study was to determine whether the addition of TNF to melphalan influenced the severity of acute regional toxic reactions after ILP. In this respect, multivariate analysis incorporating the use of TNF as one of the tested variables showed a significantly increased incidence of grade III–V toxicity after ‘mild’ hyperthermic ILP with TNF and melphalan (36%) compared to normothermic (16%) or ‘mild’ hyperthermic (17%) ILP with melphalan alone (P= 0.0038). However, no grade IV or V reactions were seen after ILP with TNF and some of the grade III reactions were in fact ‘mild’ grade III reactions with blistering limited to the sole of the foot or palm of the hand and without any long-term sequelae.24 In other large series,5,6,23 more severe acute regional toxicity occurred in over 30% of patients after ILP with TNF and melphalan, with grade IV toxicity in 8–11% and a reaction necessitating amputation in one patient in each series.6, 23 This is considerably higher than the reported 14% incidence of grade III–V reactions after 420 ‘mild’ hyperthermic ILPs with melphalan alone performed in the EORTC trial of ILP for high-risk primary melanoma.15
393
This high incidence of grade III reactions after ILP with TNF and melphalan is of concern since in one series10 these more severe reactions were strongly linked to muscle atrophy/fibrosis and limb malfunction, occurring in 11% and 15% respectively of 367 patients at 1 year after ILP. In patients with more severe grade III toxic reactions, permanent long-term regional morbidity has been reported to occur in approximately 40–50% of cases.24 Moreover, grade III reactions with blistering are very uncomfortable, can be extremely painful, are not well accepted by patients, and can prolong hospital stay considerably.19,20,24 That these reactions often indicate damage to the deep tissues is reflected by high creatine kinase levels post-ILP.24,25 The discrepancy between regional toxicity after ‘mild’ hyperthermic ILP using TNF and ‘mild’ hyperthermic ILP with melphalan alone, may be explained by differences in scheduling of the hyperthermia during melphalan circulation. In the EORTC trial protocol,15 melphalan was administered when tissue temperatures reached 38°C, temperatures were kept at 38°C during the first 10 min of melphalan circulation and were only gradually elevated to 39–40°C for the last 50 min of treatment. In the TNF-ILPs in the present study, however, melphalan was administered 30 min after TNF injection when tissue temperatures were already in the range of 39–40°C. It appears that a linear correlation exists between tissue temperatures and thermal enhancement of melphalan (i.e. the potentiating effect of hyperthermia on melphalan) without a threshold temperature,26 although conclusive data on the potential enhancement by only modestly elevated temperatures (<41°C) are lacking. Moreover, it has been shown that the melphalan perfusate concentration drops rapidly, about 40–45%, during the initial 10 min after administration.17 These findings indicate that the tissues of the patients treated in the TNF-ILP protocol (Fig. 1) were exposed to a higher area under the concentration–time curve for melphalan at temperatures of 39–40°C and to a higher total exposure to hyperthermia (temperature level × duration) than patients in the EORTC trial (Fig. 2). It may be assumed, therefore, that the increased toxicity in our patients treated with TNF and melphalan was caused, at least partly, by thermal enhancement of the melphalan. We recommend that tissue temperatures not be raised to levels above 38°C during the first 10 min after melphalan administration. The dose levels of TNF in ILP were an empirical choice by the original investigators,2,7 with essentially no pre-clinical testing reported and no phase I information to define dose-limiting toxicity. The dose was arbitrarily made 10-fold higher than the maximal tolerated systemic TNF dose, imitating dose scheduling of melphalan in ILP.12 Moreover, the 3–4 mg TNF dose approaches the optimal systemic tumouricidal dose found in animal studies.27 A dose-escalation study for TNF in ILP with melphalan has been performed by Fraker et al.12 Skin and
B. C. VROUENRAETS ET AL.
120
40
39 60
38
Tissue temperatures (°C)
Concentration of melphalan in perfusate (µgr/ml)
394
Melphalan concentration in perfusate Tissue temperature
0
30
60 Time (min)
90
Melphalan administration
120
40
39 60
38
Tissue temperatures (°C)
Concentration of melphalan in perfusate (µgr/ml)
Figure 1 Melphalan administration at relatively high temperatures during ILP with TNF.
Melphalan concentration in perfusate Tissue temperature
0
30
60 Time (min)
Melphalan administration
Figure 2 Temperature increase 10 min after melphalan administration in EORTC ‘mild’ hyperthermic ILP.
subcutaneous toxicity was similar in the 4 and 6 mg TNF subgroups. Muscle and peripheral nerve toxicity, however, was significantly worse in the 6 mg TNF subgroup, with 25% of the patients having prolonged pain and muscle atrophy and a 67% incidence of peripheral neuropathy.12 It appears that with escalation to 6 mg of TNF, the dose-limiting regional toxicity of the combination had been reached. Why the addition of more TNF leads to increased regional toxicity is unclear. Limited experience with ILP using TNF as a single agent showed virtually no regional side-effects at all with doses of up to 6 mg.12 This suggests
that interaction between high-dose TNF and melphalan may lead to increased regional toxicity. Augmented delivery of melphalan to the perfused tissues, due to a TNF-induced increased vascular permeability, is one possibility but is only speculative.12,28 Second, direct TNF cytotoxicity to normal cells may also be possible, since TNF show signs of cytotoxicity in around 30% of tumour cell lines tested so far.27 Since increasing the TNF dose to 6 mg did not lead to improved response rates or a longer duration of response,12 it may be that lower TNF doses are equally effective and less toxic.5,29–31 The addition of IFN during ILP with TNF and melphalan did not increase regional toxicity compared to the two-drug regimen in a randomized phase II trial.4 In conclusion, the addition of TNF during ILP with melphalan was associated with increased regional toxicity in present series. However, severe reactions necessitating fasciotomies or amputation were not seen after TNF-ILP and some of the grade III reactions were mild. The increased risk of more severe toxicity may be due to the relatively high tissue temperatures at which the melphalan in these ILPs was administered or to an interaction between high-dose TNF and melphalan. Further studies are required to clarify whether dose reduction of TNF is possible without reducing antitumour effects, and whether this decreases regional toxicity. At present, we recommend that tissue temperatures not be raised above 38°C within the first 10 min after melphalan administration.
REFERENCES 1. Vrouenraets BC, Nieweg OE, Kroon BBR. Thirty-five years of isolated limb perfusion for melanoma: indications and results. Br J Surg 1996; 83: 1319–28. 2. Lie´nard D, Ewalenko P, Delmotte J-J, Renard N, Lejeune FJ. Highdose recombinant tumor necrosis factor alpha in combination with interferon gamma and melphalan in isolation perfusion of the limbs for melanoma and sarcoma. J Clin Oncol 1992; 10: 52–60. 3. Lie´nard D, Eggermont AMM, Schraffordt Koops H, et al. Isolated perfusion of the limb with high-dose tumour necrosis factor-alpha (TNF), interferon-gamma (IFN-) and melphalan for stage III melanoma: results of a multicentre pilot study. Melanoma Res 1994; 4 (Suppl. 1): 21–6. 4. Lie´nard D, Eggermont AMM, Schraffordt Koops H, et al. Isolated limb perfusion with tumour necrosis factor-alpha and melphalan with or without interferon-gamma for the treatment of in-transit melanoma metastases: a multicentre randomized phase II study. Melanoma Res 1999; 9: 491–502. 5. Eggermont AMM, Schraffordt Koops H, Lie´nard D, et al. Isolated limb perfusion with high-dose tumor necrosis factor- in combination with interferon- and melphalan for nonresectable extremity soft tissue sarcomas: a multicenter trial. J Clin Oncol 1996; 14: 2653–65. 6. Eggermont AMM, Schraffordt Koops H, Klausner JM, et al. Isolated limb perfusion with tumor necrosis factor and melphalan for limb salvage in 186 patients with locally advanced soft tissue extremity sarcomas: the multicenter European experience. Ann Surg 1996; 224: 756–65. 7. Lie´nard D, Lejeune F, Ewalenko P. In transit metastases of malignant melanoma treated by high dose rTNF in combination with interferon- and melphalan in isolation perfusion. World J Surg 1992; 16: 234–40.
REGIONAL TOXICITY AFTER ISOLATED LIMB PERFUSION 8. Thom AK, Alexander HR, Andrich MP, et al. Cytokine levels and systemic toxicity in patients undergoing isolated limb perfusion with high-dose tumor necrosis factor, interferon gamma, and melphalan. J Clin Oncol 1995; 13: 264–73. 9. Vrouenraets BC, Kroon BBR, Ogilvie AC, et al. Absence of severe systemic toxicity after leakage-controlled isolated limb perfusion with tumor necrosis factor-alpha and melphalan. Ann Surg Oncol 1999; 6: 405–12. 10. Vrouenraets BC, Klaase JM, Kroon BBR, et al. Long-term morbidity after regional isolated perfusion with melphalan for melanoma of the limbs: the influence of acute regional toxic reactions. Arch Surg 1995; 130: 43–7. 11. Fraker DL, Alexander HR. Isolated limb perfusion with high-dose tumor necrosis factor for extremity melanoma and sarcoma. In: DeVita VT, Hellman S, Rosenberg SA (eds) Important Advances in Oncology 1994. Philadelphia: JB Lippincott, 1994: 179–92. 12. Fraker DL, Alexander HR, Andrich M, Rosenberg SA. Treatment of patients with melanoma of the extremity using hyperthermic isolated limb perfusion with melphalan, tumor necrosis factor, and interferon gamma: results of a tumor necrosis factor doseescalating study. J Clin Oncol 1996; 14: 479–89. 13. Klaase JM, Kroon BBR, van Geel AN, et al. Limb recurrence-free interval and survival in patients with recurrent melanoma of the extremities treated with normothermic isolated perfusion. J Am Coll Surg 1994; 178: 564–72. 14. Kroon BBR. Regional isolation perfusion in melanoma of the limbs; accomplishments, unsolved problems, future. Eur J Surg Oncol 1988; 14: 101–10. 15. Schraffordt Koops H, Vaglini M, Suciu S, et al. Prophylactic isolated limb perfusion for localized, high-risk limb melanoma: results of a multicenter randomized phase III trial. J Clin Oncol 1998; 16: 2906–12. 16. Wieberdink J, Benckhuijsen C, Braat RP, van Slooten EA, Olthuis GAA. Dosimetry in isolation perfusion of the limbs by assessment of perfused tissue volume and grading of toxic tissue reactions. Eur J Cancer Clin Oncol 1982; 18: 905–10. 17. Benckhuijsen C, Kroon BBR, van Geel AN, Wieberdink J. Regional perfusion treatment with melphalan for melanoma in a limb: an evaluation of drug kinetics. Eur J Surg Oncol 1988; 14: 157–63. 18. Klaase JM, Kroon BBR, van Geel AN, Eggermont AMM, Franklin HR, Hart AAM. Patient and treatment related factors associated with acute regional toxicity after isolated perfusion for melanoma of the extremities. Am J Surg 1994; 167: 618–20. 19. Lie´nard D, Lejeune F, Autier Ph, Sales F, Ghanem G. Physiological and pharmacokinetics parameters in isolaton perfusion of the limbs. In: Lejeune FJ, Chaudhuri PK, Das Gupta T (eds) Malignant
20. 21. 22.
23.
24.
25. 26. 27.
28.
29. 30.
31.
395 Melanoma: Medical and Surgical Management. New York: McGrawHill, 1994: 241–8. Thompson JF, Eksborg S, Kam PCA, et al. Determinants of acute regional toxicity following isolated limb perfusion for melanoma. Melanoma Res 1996; 6: 267–71. Klaase JM, Kroon BBR, van Slooten GW, et al. Relation between calculated melphalan peak concentrations and toxicity in regional isolated perfusion for melanoma. Reg Cancer Treat 1992; 4: 309–12. Scott RN, Blackie R, Kerr DJ, et al. Melphalan in isolated limb perfusion for malignant melanoma, bolus or divided dose, tissue levels, the pH effect. In: Jakesz R, Rainer H (eds) Progress in Regional Cancer Therapy. Berlin: Springer, 1990: 195–200. Gutman M, Inbar M, Lev-Shlush D, et al. High dose tumor necrosis factor- and melphalan administered via isolated limb perfusion for advanced limb soft tissue sarcoma results in a >90% response rate and limb preservation. Cancer 1997; 79: 1129–37. Vrouenraets BC, Kroon BBR, Klaase JM, et al. Severe acute regional toxicity after normothermic or ‘mild’ hyperthermic isolated limb perfusion with melphalan for melanoma. Melanoma Res 1995; 5: 425–31. Vrouenraets BC, Kroon BBR, Klaase JM, et al. Value of laboratory tests in monitoring acute regional toxicity after isolated limb perfusion. Ann Surg Oncol 1997; 4: 88–94. Dahl O, Mella O. Hyperthermia and chemotherapeutic agents. In: Field SB, Hand JW (eds) Introduction to the Practical Aspects of Clinical Hyperthermia. London: Taylor and Francis, 1990: 108–42. Lejeune FJ. High dose recombinant tumour necrosis factor (rTNF) administered by isolation perfusion for advanced tumours of the limbs: a model for biochemotherapy of cancer. Eur J Cancer 1995; 31A: 1009–16. Folli S, Pe`legrin A, Chalandon Y, et al. Tumor-necrosis factor can enhance radio-antibody uptake in human colon carcinoma xenografts by increasing vascular permeability. Int J Cancer 1993; 53: 829–36. Hill S, Fawcett WJ, Sheldon J, et al. Low-dose tumour necrosis factor and melphalan in hyperthermic isolated limb perfusion. Br J Surg 1993; 80: 995–7. Vaglini M, Santinami M, Manzi R, et al. Treatment of in-transit metastases from cutaneous melanoma by isolation perfusion with tumour necrosis factor-alpha (TNF), melphalan and interferongamma (IFN): dose-finding experience at National Cancer Institute of Milan. Melanoma Res 1994; 4 (Suppl. 1): 35–8. Di Filippo F, Anza` M, Rossi CR, et al. Hyperthermic antiblastic perfusion with TNF-Alpha and L.Pam in patients with stage II, IIIA and IIIAB limb melanoma. Melanoma Res 1997; 7 (Suppl. 1): S32 (Abstract).
Accepted for publication 28 February 2001
Regional toxicity after isolated limb perfusion with melphalan and tumour necrosis factor- versus toxicity after melphalan alone B. C. Vrouenraets∗, A. M. M. Eggermont†, A. A. M. Hart‡, J. M. Klaase∗, A. N. van Geel†, O. E. Nieweg∗ and B. B. R. Kroon∗ Departments of ∗Surgery and ‡Radiotherapy, The Netherlands Cancer Institute/Antoni van Leeuwenhoek ziekenhuis, Amsterdam, and †Department of Surgery, University Hospital Rotterdam/Dr Daniel den Hoed Cancer Centre, Rotterdam, The Netherlands
Aims: To determine whether the addition of high-dose tumour necrosis factor- (TNF) to isolated limb perfusion (ILP) with melphalan increases acute regional tissue toxicity compared to ILP with melphalan alone. Methods: A retrospective, multivariate analysis of toxicity after normothermic (37–38°C) and ‘mild’ hyperthermic (38–40°C) ILPs for melanoma was undertaken. Normothermic ILP with melphalan was performed in 294 patients (70.8%), ‘mild’ hyperthermic ILP with melphalan in 71 patients (17.1%) and ‘mild’ hyperthermic ILP with melphalan combined with TNF in 50 patients (12.0%). Toxicity was nil or mild (grades I–II according to Wieberdink et al.) in 339 patients (81.7%), and more severe acute regional toxicity (grades III–V) developed in 76 patients (18.3%). A stepwise logistic regression procedure was performed for the multivariate analysis of prognostic factors for more severe toxicity. Results: On univariate analysis, ‘mild’ hyperthermic ILP with melphalan plus TNF significantly increased the incidence of more severe acute regional toxicity compared to normothermic and ‘mild’ hyperthermic ILP with melphalan alone (36% vs 16% and 17%; P=0.0038). However, after ILP using TNF no grade IV (compartment compression syndrome) or grade V (toxicity necessitating amputation) reactions were seen. Significantly more severe toxicity was seen after ILPs performed between 1991 and 1994 compared with earlier ILPs (33% vs 14%; P=0.0001). Also, women had a higher risk of more severe toxicity than men (22% vs 7%; P=0.0007). After multivariate analysis, prognostic factors which remained significant were: sex (P=0.0013) and either ILP schedule (P=0.013) or treatment period (P= 0.0003). Conclusions: Regional toxicity after ‘mild’ hyperthermic ILP with melphalan and TNF was significantly increased compared to ILP with melphalan alone. This may be caused by increased thermal enhancement of melphalan due to the higher tissue temperatures (39–40°C) at which the melphalan in the TNF-ILPs was administered or by an interaction between high-dose TNF and melphalan. 2001 Harcourt Publishers Ltd Key words: isolated limb perfusion; melphalan; tumour necrosis factor-alpha; melanoma; toxicity.
INTRODUCTION Isolated limb perfusion (ILP) is the treatment of choice in locally inoperable melanoma of the limbs, with complete responses in up to 55% of patients after ILP with melphalan.1 The addition of tumour necrosis factor-alpha (TNF), at a dosage up to 10 times that which can be tolerated after systemic administration, increases the
Correspondence to: Professor dr. B. B. R. Kroon, Department of Surgery, The Netherlands Cancer Institute/Antoni van Leeuwenhoek ziekenhuis, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands. Tel: +31 20 512 2552; Fax: +31 20 512 2554; E-mail: [email protected] 0748–7983/01/040390+06 $35.00/0
complete response rate to 70–90%.2–4 Moreover, the application of TNF in ILP has now been demonstrated to yield high response rates (over 80%) in locally advanced soft tissue sarcomas.5,6 Systemic toxicity after ILP using TNF, at first a major problem in some centres, is directly correlated to leakage of TNF into the systemic circulation during the procedure.6–8 When leakage is prevented and a thorough washout of the limb vasculature is performed at the end of ILP, only a short-lived (less than 1 hour) systemically measurable TNF peak is observed. In those patients, systemic toxicity is restricted to fever on the day of surgery and otherwise does not differ significantly from toxicity after ILP with melphalan alone,6,9 leaving regional toxicity as the dose limitating factor. 2001 Harcourt Publishers Ltd
REGIONAL TOXICITY AFTER ISOLATED LIMB PERFUSION Initial reports from Lie´nard et al.2,7 stated that highdose TNF administered by ILP does not add any acute regional toxicity compared to ILP with melphalan alone. This is important since a strong correlation has been demonstrated between the severity of the acute regional toxic reactions and the incidence of long-term morbidity, especially tissue fibrosis, muscle atrophy and limb malfunction.10 In other reports,11,12 however, the rapidity of onset, the overall duration and the severity of most regional side-effects appeared to be more prominent with the addition of TNF to melphalan. A retrospective, multivariate analysis was undertaken of all single normothermic and ‘mild’ hyperthermic ILPs performed at The Netherlands Cancer Institute, Amsterdam, and the Dr Daniel den Hoed Cancer Centre, Rotterdam, The Netherlands, to determine whether the addition of high-dose TNF to ILP with melphalan increases acute regional tissue toxicity compared to ILP with melphalan alone.
PATIENTS AND METHODS From 1978 to 1995, 530 patients underwent 668 ILPs for melanoma in our institutions. A computer-assisted database, containing all patient, tumour and treatment variables, allowed the selection of patients who underwent a single normothermic (37–38°C) or ‘mild’ hyperthermic (38–40°C) ILP with melphalan with or without TNF. Of patients having more than one ILP treatment, we considered only the first. During the study period different single ILP schedules were used for various indications. Until 1985, normothermic ILP with melphalan was performed both for ‘high-risk’ primary melanoma and for recurrent melanoma.13,14 From 1986 to 1994, all patients with ‘highrisk’ primary melanoma underwent ‘mild’ hyperthermic ILP with melphalan according to an EORTC trial protocol;15 whilst patients with recurrent melanoma were treated using our routine normothermic ILP schedule. After 1994, patients with ‘high-risk’ primary melanoma were no longer treated by ILP. In 1991, TNF and gamma-interferon (IFN) were introduced under ‘mild’ hyperthermic conditions in our ILP methodology for the treatment of recurrent disease in patients who were suitable candidates for this experimental protocol.3 ILP was conducted in the lower limb at either the external iliac or the femoropopliteal level, and in the upper limb at the axillary or brachial level. For iliac ILPs, the iliac and obturator lymph nodes were removed, and for the axillary procedure, an axillary lymph node dissection was performed. All ILPs were carried out with a routine dose of 10 mg melphalan/l perfused tissue for lower limbs and 13 mg/l for upper limbs.16,17 Melphalan was injected gradually (in 2–3 min) into the arterial line. The duration of all ILPs was 1 hour. Four temperature probes were inserted into subcutaneous tissue and muscle. In the ‘mild’
391
Table 1 Grading of acute limb toxicity according to Wieberdink et al.16 Grade I Grade II Grade III
Grade IV
Grade V
No subjective or objective evidence of reaction Slight erythema and/or oedema Considerable erythema and/or oedema with some blistering; slightly disturbed motility Extensive epidermolysis and/or obvious damage to the deep tissues, causing definite functional disturbances; threatening or manifest compartment syndromes Severe reaction which may necessitate amputation
hyperthermic EORTC trial ILPs,15 the melphalan was injected when the subcutaneous tissues of the distal part of the extremity had reached a temperature of 38°C. After 10 min of melphalan perfusion, the temperature was gradually elevated to 39–40°C and was kept constant for the rest of the 60 min of treatment. The TNF ILP protocol involved use of 4 mg of TNF for iliac and femoral ILPs and 3 mg for axillary and brachial ILPs.3 For 2 days before surgery, 0.2 mg IFN was injected subcutaneously; the same dose was given during the ILP. The whole ‘mild’ hyperthermic ILP with TNF lasted 90 min, during which the temperature of the perfusate was maintained at 40°C. IFN and TNF were injected successively as boluses into the arterial line when distal subcutaneous temperatures reached 38°C after which temperatures were gradually elevated. All temperatures were kept at 39–40°C during the last 60 min of ILP. Melphalan was administered in the routine dose 30 min after TNF injection. In some patients IFN was omitted as part of a multicentre trial.4 Specific details of our ILP techniques have been described elsewhere.14 No prophylactic fasciotomies were performed. Acute regional toxicity after ILP was graded according to Wieberdink et al.16 (Table 1). With all these ILP schedules, we strived to limit acute regional toxicity to Wieberdink grades II or III, meaning completely reversible reactions without long-term sequelae. The main aim of the present study was to examine the relation between ILP schedule and acute regional toxicity after ILP. A stepwise logistic regression procedure was performed for the multivariate analysis of prognostic factors for more severe acute regional toxicity (Wieberdink grades III–V). Ordinal variables were used linearly, but linearity was tested for. In testing a particular variable only those patients who had a missing value for that variable or for any variable already included in the model were excluded from the analysis. The inclusion and exclusion limit was set at PΖ0.15.
392
RESULTS A total of 415 patients were selected for this study: 309 women (74.5%) and 106 men (25.5%) with a mean age of 53 years (range: 17–89 years). A normothermic ILP with melphalan was performed in 294 patients (70.8%). Seventy-one patients (17.1%) underwent a ‘mild’ hyperthermic ILP with melphalan alone as part of the EORTC prophylactic ILP trial. Fifty patients (12.0%) were treated under ‘mild’ hyperthermic conditions with TNF and melphalan. Fourteen patients (3.4%) had grade I acute regional toxicity, 325 (78.3%) grade II, 71 (17.1%) grade III, three (0.7%) grade IV and two patients (0.5%) had a grade V reaction. Grade IV reactions occurred after two normothermic ILPs and after one ‘mild’ hyperthermic ILP with melphalan alone. Amputations were needed after one normothermic and one ‘mild’ hyperthermic ILP with melphalan alone. In Table 2, the patient and treatment variables tested for their association with acute regional toxicity are shown. On univariate analysis, the most significant factor predicting more severe acute regional toxicity was the period in which the ILP was performed, with more severe toxicity after ILPs performed between 1991 and 1994 compared with earlier ILPs (33% vs 14%; P= 0.0001). Women had an increased risk of more severe toxicity compared to men (22% vs 7%; P=0.0007). The ILP schedule was also found to be related to the severity of the acute regional toxic reactions after ILP (P= 0.0038), with a greater probability of grade III–V toxicity after ILP with TNF than after normothermic or ‘mild’ hyperthermic ILP with melphalan alone (36% vs 16% and 17%, respectively). Treatment period, sex and melphalan dose satisfied the inclusion criteria of PΖ0.15 in the multivariate stepwise logistic regression analysis (Table 3). Adjusted for these variables, the evidence for a relation between ILP schedule and acute regional toxicity disappeared (P= 0.89). Only treatment period (P=0.0003) and sex (P= 0.0009) retained their predictive value for toxicity. However, as treatment period and ILP schedule were strongly linked (Table 4), the power to detect differences between ILP schedules was markedly decreased. We therefore adjusted for sex, ILP schedule and melphalan dose instead of treatment period (Table 3). This decreased the significance of the treatment period as a prognostic factor for acute regional toxicity considerably (P=0.022), suggesting that the increased incidence of grade III–V toxicity in the period 1991–1994 could partly be explained by the use of TNF and higher melphalan doses during that period. In this model, the estimated odds ratio of the TNF schedule versus the other two schedules was 2.7 with a 95% confidence interval of 1.4–5.1. With more severe toxicity in 16% of ILPs without TNF, this would mean more severe toxicity in 34%
B. C. VROUENRAETS ET AL. Table 2 Variables used in the analysis of acute regional toxicity after ILP-univariate analysis of prognostic factors for more severe acute regional toxicity after ILP (grades III–V according to Wieberdink et al.)16 Variable
Sex Male Female Age (years) <40 40–49 50–59 >60 Dose of melphalan (mg) Unknown <75 75–99 100–124 >125 Venous pressure (cm) at 30 min Unknown <10 >10 Leakage to the body (%) Unknown 0 0.1–1 >1 Level of isolation Axillary/brachial Iliac Femoropopliteal ILP period 1978–1984 1985–1990 1991–1994 ILP schedule Normothermic ‘Mild’ hyperthermic TNF
Number Number of of patients (%) patients with grade III–V toxicity
P-value
0.0007 106 309
7 (6.6) 69 (22.3)
91 62 102 160
15 13 21 27
0.95 (16.5) (21.0) (20.6) (16.9) 0.38 1 99 113 147 55
17 17 31 11
(17.2) (15.0) (21.1) (20.0) 0.80
61 310 44
47 (15.2) 6 (13.6) 0.29
32 222 106 55
45 (20.3) 15 (14.2) 9 (16.4)
68 293 53
13 (18.8) 54 (18.4) 9 (17.0)
179 140 96
25 (14.0) 19 (13.6) 32 (33.3)
294 71 50
46 (15.7) 12 (16.9) 18 (36.0)
0.97
0.0001
0.0038
(95% CI: 21%–49%) of ILPs with TNF, adjusted for sex and melphalan dose.
DISCUSSION The selection of possible prognostic factors for acute regional toxicity in the present series was based on recent reports on this issue.18–21 The present study confirmed only the prognostic significance of sex, with women at increased risk for developing toxicity.10,18 This female risk factor may be attributed to the fact that
REGIONAL TOXICITY AFTER ISOLATED LIMB PERFUSION Table 3 Prognostic factors for acute regional toxicity after ILP in the stepwise logistic regression analysis Variable
Step 3∗ P-value
Step 3a† P-value
Sex Age Dose of melphalan Venous pressure Leakage to the body Level of isolation ILP period ILP schedule
0.0009 0.42 0.13 0.70 0.23 0.17 0.0003 0.89
0.0013 0.50 0.14 0.77 0.39 0.17 0.022 0.013
∗ Adjusted for in the following order: ILP period, sex, and dose of melphalan. † Adjusted for in the following order: sex, ILP schedule, and dose of melphalan.
Table 4 ILP period and ILP schedule ILP schedule Normothermic
‘Mild’ TNF Total hyperthermic ILP period
1978–1984 1985–1990 1991–1994
179 96 19
0 44 27
0 0 50
179 140 96
Total
294
71
50
415
women have a lower muscle/fat ratio than men. Since there is a higher melphalan uptake in muscle than fat,22 the muscle tissue of female patients is exposed to a relatively higher dose of melphalan. The major aim of the present study was to determine whether the addition of TNF to melphalan influenced the severity of acute regional toxic reactions after ILP. In this respect, multivariate analysis incorporating the use of TNF as one of the tested variables showed a significantly increased incidence of grade III–V toxicity after ‘mild’ hyperthermic ILP with TNF and melphalan (36%) compared to normothermic (16%) or ‘mild’ hyperthermic (17%) ILP with melphalan alone (P= 0.0038). However, no grade IV or V reactions were seen after ILP with TNF and some of the grade III reactions were in fact ‘mild’ grade III reactions with blistering limited to the sole of the foot or palm of the hand and without any long-term sequelae.24 In other large series,5,6,23 more severe acute regional toxicity occurred in over 30% of patients after ILP with TNF and melphalan, with grade IV toxicity in 8–11% and a reaction necessitating amputation in one patient in each series.6, 23 This is considerably higher than the reported 14% incidence of grade III–V reactions after 420 ‘mild’ hyperthermic ILPs with melphalan alone performed in the EORTC trial of ILP for high-risk primary melanoma.15
393
This high incidence of grade III reactions after ILP with TNF and melphalan is of concern since in one series10 these more severe reactions were strongly linked to muscle atrophy/fibrosis and limb malfunction, occurring in 11% and 15% respectively of 367 patients at 1 year after ILP. In patients with more severe grade III toxic reactions, permanent long-term regional morbidity has been reported to occur in approximately 40–50% of cases.24 Moreover, grade III reactions with blistering are very uncomfortable, can be extremely painful, are not well accepted by patients, and can prolong hospital stay considerably.19,20,24 That these reactions often indicate damage to the deep tissues is reflected by high creatine kinase levels post-ILP.24,25 The discrepancy between regional toxicity after ‘mild’ hyperthermic ILP using TNF and ‘mild’ hyperthermic ILP with melphalan alone, may be explained by differences in scheduling of the hyperthermia during melphalan circulation. In the EORTC trial protocol,15 melphalan was administered when tissue temperatures reached 38°C, temperatures were kept at 38°C during the first 10 min of melphalan circulation and were only gradually elevated to 39–40°C for the last 50 min of treatment. In the TNF-ILPs in the present study, however, melphalan was administered 30 min after TNF injection when tissue temperatures were already in the range of 39–40°C. It appears that a linear correlation exists between tissue temperatures and thermal enhancement of melphalan (i.e. the potentiating effect of hyperthermia on melphalan) without a threshold temperature,26 although conclusive data on the potential enhancement by only modestly elevated temperatures (<41°C) are lacking. Moreover, it has been shown that the melphalan perfusate concentration drops rapidly, about 40–45%, during the initial 10 min after administration.17 These findings indicate that the tissues of the patients treated in the TNF-ILP protocol (Fig. 1) were exposed to a higher area under the concentration–time curve for melphalan at temperatures of 39–40°C and to a higher total exposure to hyperthermia (temperature level × duration) than patients in the EORTC trial (Fig. 2). It may be assumed, therefore, that the increased toxicity in our patients treated with TNF and melphalan was caused, at least partly, by thermal enhancement of the melphalan. We recommend that tissue temperatures not be raised to levels above 38°C during the first 10 min after melphalan administration. The dose levels of TNF in ILP were an empirical choice by the original investigators,2,7 with essentially no pre-clinical testing reported and no phase I information to define dose-limiting toxicity. The dose was arbitrarily made 10-fold higher than the maximal tolerated systemic TNF dose, imitating dose scheduling of melphalan in ILP.12 Moreover, the 3–4 mg TNF dose approaches the optimal systemic tumouricidal dose found in animal studies.27 A dose-escalation study for TNF in ILP with melphalan has been performed by Fraker et al.12 Skin and
B. C. VROUENRAETS ET AL.
120
40
39 60
38
Tissue temperatures (°C)
Concentration of melphalan in perfusate (µgr/ml)
394
Melphalan concentration in perfusate Tissue temperature
0
30
60 Time (min)
90
Melphalan administration
120
40
39 60
38
Tissue temperatures (°C)
Concentration of melphalan in perfusate (µgr/ml)
Figure 1 Melphalan administration at relatively high temperatures during ILP with TNF.
Melphalan concentration in perfusate Tissue temperature
0
30
60 Time (min)
Melphalan administration
Figure 2 Temperature increase 10 min after melphalan administration in EORTC ‘mild’ hyperthermic ILP.
subcutaneous toxicity was similar in the 4 and 6 mg TNF subgroups. Muscle and peripheral nerve toxicity, however, was significantly worse in the 6 mg TNF subgroup, with 25% of the patients having prolonged pain and muscle atrophy and a 67% incidence of peripheral neuropathy.12 It appears that with escalation to 6 mg of TNF, the dose-limiting regional toxicity of the combination had been reached. Why the addition of more TNF leads to increased regional toxicity is unclear. Limited experience with ILP using TNF as a single agent showed virtually no regional side-effects at all with doses of up to 6 mg.12 This suggests
that interaction between high-dose TNF and melphalan may lead to increased regional toxicity. Augmented delivery of melphalan to the perfused tissues, due to a TNF-induced increased vascular permeability, is one possibility but is only speculative.12,28 Second, direct TNF cytotoxicity to normal cells may also be possible, since TNF show signs of cytotoxicity in around 30% of tumour cell lines tested so far.27 Since increasing the TNF dose to 6 mg did not lead to improved response rates or a longer duration of response,12 it may be that lower TNF doses are equally effective and less toxic.5,29–31 The addition of IFN during ILP with TNF and melphalan did not increase regional toxicity compared to the two-drug regimen in a randomized phase II trial.4 In conclusion, the addition of TNF during ILP with melphalan was associated with increased regional toxicity in present series. However, severe reactions necessitating fasciotomies or amputation were not seen after TNF-ILP and some of the grade III reactions were mild. The increased risk of more severe toxicity may be due to the relatively high tissue temperatures at which the melphalan in these ILPs was administered or to an interaction between high-dose TNF and melphalan. Further studies are required to clarify whether dose reduction of TNF is possible without reducing antitumour effects, and whether this decreases regional toxicity. At present, we recommend that tissue temperatures not be raised above 38°C within the first 10 min after melphalan administration.
REFERENCES 1. Vrouenraets BC, Nieweg OE, Kroon BBR. Thirty-five years of isolated limb perfusion for melanoma: indications and results. Br J Surg 1996; 83: 1319–28. 2. Lie´nard D, Ewalenko P, Delmotte J-J, Renard N, Lejeune FJ. Highdose recombinant tumor necrosis factor alpha in combination with interferon gamma and melphalan in isolation perfusion of the limbs for melanoma and sarcoma. J Clin Oncol 1992; 10: 52–60. 3. Lie´nard D, Eggermont AMM, Schraffordt Koops H, et al. Isolated perfusion of the limb with high-dose tumour necrosis factor-alpha (TNF), interferon-gamma (IFN-) and melphalan for stage III melanoma: results of a multicentre pilot study. Melanoma Res 1994; 4 (Suppl. 1): 21–6. 4. Lie´nard D, Eggermont AMM, Schraffordt Koops H, et al. Isolated limb perfusion with tumour necrosis factor-alpha and melphalan with or without interferon-gamma for the treatment of in-transit melanoma metastases: a multicentre randomized phase II study. Melanoma Res 1999; 9: 491–502. 5. Eggermont AMM, Schraffordt Koops H, Lie´nard D, et al. Isolated limb perfusion with high-dose tumor necrosis factor- in combination with interferon- and melphalan for nonresectable extremity soft tissue sarcomas: a multicenter trial. J Clin Oncol 1996; 14: 2653–65. 6. Eggermont AMM, Schraffordt Koops H, Klausner JM, et al. Isolated limb perfusion with tumor necrosis factor and melphalan for limb salvage in 186 patients with locally advanced soft tissue extremity sarcomas: the multicenter European experience. Ann Surg 1996; 224: 756–65. 7. Lie´nard D, Lejeune F, Ewalenko P. In transit metastases of malignant melanoma treated by high dose rTNF in combination with interferon- and melphalan in isolation perfusion. World J Surg 1992; 16: 234–40.
REGIONAL TOXICITY AFTER ISOLATED LIMB PERFUSION 8. Thom AK, Alexander HR, Andrich MP, et al. Cytokine levels and systemic toxicity in patients undergoing isolated limb perfusion with high-dose tumor necrosis factor, interferon gamma, and melphalan. J Clin Oncol 1995; 13: 264–73. 9. Vrouenraets BC, Kroon BBR, Ogilvie AC, et al. Absence of severe systemic toxicity after leakage-controlled isolated limb perfusion with tumor necrosis factor-alpha and melphalan. Ann Surg Oncol 1999; 6: 405–12. 10. Vrouenraets BC, Klaase JM, Kroon BBR, et al. Long-term morbidity after regional isolated perfusion with melphalan for melanoma of the limbs: the influence of acute regional toxic reactions. Arch Surg 1995; 130: 43–7. 11. Fraker DL, Alexander HR. Isolated limb perfusion with high-dose tumor necrosis factor for extremity melanoma and sarcoma. In: DeVita VT, Hellman S, Rosenberg SA (eds) Important Advances in Oncology 1994. Philadelphia: JB Lippincott, 1994: 179–92. 12. Fraker DL, Alexander HR, Andrich M, Rosenberg SA. Treatment of patients with melanoma of the extremity using hyperthermic isolated limb perfusion with melphalan, tumor necrosis factor, and interferon gamma: results of a tumor necrosis factor doseescalating study. J Clin Oncol 1996; 14: 479–89. 13. Klaase JM, Kroon BBR, van Geel AN, et al. Limb recurrence-free interval and survival in patients with recurrent melanoma of the extremities treated with normothermic isolated perfusion. J Am Coll Surg 1994; 178: 564–72. 14. Kroon BBR. Regional isolation perfusion in melanoma of the limbs; accomplishments, unsolved problems, future. Eur J Surg Oncol 1988; 14: 101–10. 15. Schraffordt Koops H, Vaglini M, Suciu S, et al. Prophylactic isolated limb perfusion for localized, high-risk limb melanoma: results of a multicenter randomized phase III trial. J Clin Oncol 1998; 16: 2906–12. 16. Wieberdink J, Benckhuijsen C, Braat RP, van Slooten EA, Olthuis GAA. Dosimetry in isolation perfusion of the limbs by assessment of perfused tissue volume and grading of toxic tissue reactions. Eur J Cancer Clin Oncol 1982; 18: 905–10. 17. Benckhuijsen C, Kroon BBR, van Geel AN, Wieberdink J. Regional perfusion treatment with melphalan for melanoma in a limb: an evaluation of drug kinetics. Eur J Surg Oncol 1988; 14: 157–63. 18. Klaase JM, Kroon BBR, van Geel AN, Eggermont AMM, Franklin HR, Hart AAM. Patient and treatment related factors associated with acute regional toxicity after isolated perfusion for melanoma of the extremities. Am J Surg 1994; 167: 618–20. 19. Lie´nard D, Lejeune F, Autier Ph, Sales F, Ghanem G. Physiological and pharmacokinetics parameters in isolaton perfusion of the limbs. In: Lejeune FJ, Chaudhuri PK, Das Gupta T (eds) Malignant
20. 21. 22.
23.
24.
25. 26. 27.
28.
29. 30.
31.
395 Melanoma: Medical and Surgical Management. New York: McGrawHill, 1994: 241–8. Thompson JF, Eksborg S, Kam PCA, et al. Determinants of acute regional toxicity following isolated limb perfusion for melanoma. Melanoma Res 1996; 6: 267–71. Klaase JM, Kroon BBR, van Slooten GW, et al. Relation between calculated melphalan peak concentrations and toxicity in regional isolated perfusion for melanoma. Reg Cancer Treat 1992; 4: 309–12. Scott RN, Blackie R, Kerr DJ, et al. Melphalan in isolated limb perfusion for malignant melanoma, bolus or divided dose, tissue levels, the pH effect. In: Jakesz R, Rainer H (eds) Progress in Regional Cancer Therapy. Berlin: Springer, 1990: 195–200. Gutman M, Inbar M, Lev-Shlush D, et al. High dose tumor necrosis factor- and melphalan administered via isolated limb perfusion for advanced limb soft tissue sarcoma results in a >90% response rate and limb preservation. Cancer 1997; 79: 1129–37. Vrouenraets BC, Kroon BBR, Klaase JM, et al. Severe acute regional toxicity after normothermic or ‘mild’ hyperthermic isolated limb perfusion with melphalan for melanoma. Melanoma Res 1995; 5: 425–31. Vrouenraets BC, Kroon BBR, Klaase JM, et al. Value of laboratory tests in monitoring acute regional toxicity after isolated limb perfusion. Ann Surg Oncol 1997; 4: 88–94. Dahl O, Mella O. Hyperthermia and chemotherapeutic agents. In: Field SB, Hand JW (eds) Introduction to the Practical Aspects of Clinical Hyperthermia. London: Taylor and Francis, 1990: 108–42. Lejeune FJ. High dose recombinant tumour necrosis factor (rTNF) administered by isolation perfusion for advanced tumours of the limbs: a model for biochemotherapy of cancer. Eur J Cancer 1995; 31A: 1009–16. Folli S, Pe`legrin A, Chalandon Y, et al. Tumor-necrosis factor can enhance radio-antibody uptake in human colon carcinoma xenografts by increasing vascular permeability. Int J Cancer 1993; 53: 829–36. Hill S, Fawcett WJ, Sheldon J, et al. Low-dose tumour necrosis factor and melphalan in hyperthermic isolated limb perfusion. Br J Surg 1993; 80: 995–7. Vaglini M, Santinami M, Manzi R, et al. Treatment of in-transit metastases from cutaneous melanoma by isolation perfusion with tumour necrosis factor-alpha (TNF), melphalan and interferongamma (IFN): dose-finding experience at National Cancer Institute of Milan. Melanoma Res 1994; 4 (Suppl. 1): 35–8. Di Filippo F, Anza` M, Rossi CR, et al. Hyperthermic antiblastic perfusion with TNF-Alpha and L.Pam in patients with stage II, IIIA and IIIAB limb melanoma. Melanoma Res 1997; 7 (Suppl. 1): S32 (Abstract).
Accepted for publication 28 February 2001