Repeat isolated limb perfusion with TNF a and melphalan for recurrent
limb melanoma after failure of previous perfusion
小乔初嫁
E.M.Noorda a,*,B.C.Vrouenraets a,O.E.Nieweg a,A.N.van Geel b,
A.M.M.Eggermont b,
B.B.R.Kroon a
a Department of Surgery,The Netherlands Cancer Institute,Antoni van Leeuwenhoek Hospital,Amsterdam,The Netherlands沈药
b Department of Surgery,Erasmus Medical Centre,Daniel den Hoed Cancer Centre,Rotterdam,The Netherlands
Accepted20October2005
Available online18January2006
Abstract
Aim:To asss the effectiveness of isolated limb perfusion(ILP)with tumour necrosis factor-alpha(TNF a)and melphalan for recurrent or persistent melanoma lesions after previous ILP.
Methods:Between1978and2001,21patients(mean age65,range29–83years)underwent repeat ILP for recurrent or persistent melanoma after a previous ILP.First ILPs had been performed with melphalan alone in13patients and with addition of TNF a in eight,for a median of nine lesions(interquartile(IQ)range2–23lesions).Repeat ILP was performed with TNF a and melphalan in all21patients for a median of nine lesions(IQ range5–25lesions).Median follow-up after repeat ILP was18months(IQ range6–36months).
Results:Thirteen patients attained a complete respon(CR)after repeat ILP compared to11of17with measurable lesions after thefirst ILP. Nine patients relapd after CR.Median limb recurrence-free survival was13months.Fourteen patients had mild acute regional toxicity after repeat ILP compared to18after thefirst ILP(n.s.).One patient underwent amputation for critical limb ischemia10months following repeat ILP.The limb salvage rate was95%.Overall median survival was62months after CR compared to13months for tho without CR(P Z 0.05).
Conclusion:Repeat ILP with TNF a and melphalan is feasible after previous ILP with mild regional to
xicity.The CR rate is relatively high and comparable to thefirst procedure with good limb recurrence-free survival and high limb salvage rate.
q2005Elvier Ltd.All rights rerved.
Keywords:Melanoma;Regional perfusion;Limb salvage
Introduction
Isolated limb perfusion(ILP)with melphalan for locoregionally recurrent limb melanoma generally results in an80%tumour respon rate with54%complete respons.1Two large studies showed46–54%recurrence after a complete respon with ILP.2,3Amputation is afinal option in such patients when recurrences are not amenable to local surgery.There is reluctance to perform repeat ILP since acute regional toxicity has been reported to be more vere and it could be assumed that patients not responding to initial ILP may have little to gain from a repeat procedure.4Nevertheless,a complete respon rate of 74%has been demonstrated after repeat ILP with melphalan applied in single,double,triple or quential ILP-schedules.5The latter elaborate procedures are no longer performed since the introduction of tumour necrosis factor-alpha(TNF a)in the isolated circuit.However,studies with TNF a in repeat ILP procedures are few:patient ries are small with het
erogeneous patient and tumour character-istics.6,7Therefore,we assd the results of repeat ILP with TNF a in patients who relapd or had persistent dia after previous ILP in our institutions.
穷Patients and methods
From1978to2001,505ILPs were performed in451 patients with recurrent and/or locally advanced extremity melanoma in our institutions.All patient,tumour,treatment and follow-up data were maintained on a computer databa.In this databa,21patients were identified
in *Corresponding author.Address:Department of Surgery,Slotervaart-
ziekenhuis,Louwesweg6,1006BK Amsterdam,The Netherlands.Tel.:
C31205124430;fax:C31205124853.
E-mail address:cheno@slz.nl(E.M.Noorda).
whom29repeat ILPs were performed using TNF a and melphalan for persisting or recurrent melanoma lesions after previous ILP.Six of the patients underwent three ILPs and one four.There were three male patients(15%)and the mean age was65years(range29–83years).Recurrent melanoma lesions were located on the lower limb in19and on the upper limb in two patients.Stage of dia was classified according to the MD Anderson staging system (Table1).8
ILP technique
Our ILP technique has been described in detail el-where.9During ILP,the major artery and vein are clamped at the desired level,collateral vesls are ligated and a tourniquet is applied around the limb,proximal to the region of ILP.After inrtion of the catheters,the isolated limb is perfud by an extracorporeal circulation,oxygenated and propelled by a heart–lung machine.A melphalan do of 13mg/l-perfud tissue in the upper limb and10mg/l for the lower limb is added to the perfusate.For
TNF a this is3 and4mg,respectively,irrespective of limb volume.In ven patients,0.2mg interferon-gamma(IFN g)was additionally injected subcutaneously for2days before repeat ILP.The same do was given intraoperatively in the isolated circuit.Adequate tissue temperatures are achieved and maintained by heating the heparinid perfusate and application of a warm water blanket around the limb.Limb temperatures are kept between37and388C(normothermia) in ILP with melphalan alone or38and408C(mild hyperthermia)in ca of ILP with TNF a.ILP with melphalan lasts1h and90min when TNF a is ud.At termination of the ILP,the perfusate is drained out and the limb is rind with an electrolyte solution.The tourniquet is then relead and catheters are removed.In one patient,the first ILP consisted of a quential ILP scheme,in which ILP with hyperthermia(42–438C)was followed1week later by ILP with melphalan alone under normothermic conditions.10
Toxicity and respon
Acute regional tissue toxicity after ILP was graded according to Wieberdink et al.11Tumour respon was measured by WHO criteria.12Post-operatively,patients stayed in bed with the leg elevated until acute toxic limb reactions subsided.They were gradually mobilid with the help of a physiotherapist and discharged when fully ambulant.
Characteristics and results after thefirst ILP
In Table2characteristics and results of thefirst ILP are shown.In four patients,tumour respon of thefirst ILP was not evaluable becau the procedure was performed as adjuvant to the excision of a primary tumour as part of a randomid trial,13or prophylactically after the excision of in-transit metastas.Twelve patients(71%)with in-transit metastas or a local recurrence had a high tumour load according to the combined criteria of Fraker and more than10lesions and/or lesions larger than3cm.14,15 After thefirst ILP,11of17patients(65%,95%CI39–90%) with measurable dia attained a complete respon(CR). Median duration of that CR was7months(IQ range5–25 Table2
Characteristics and results of thefirst ILP and repeat ILP
First ILP Repeat ILP P-
value MD Anderson stage of dia
Stage I primary melanoma20
Stage IIB satellitosis60
Stage IIIA in-transit sin
lymph-node metastas
910n.s. Stage IIIAB in-transit with
方便英语
lymph-node metastas
411!0.05 Number of lesions(median,
IQ range)
9(2–23)9(5–25)
ILP-type
Melphalan120
TNF a C melphalan(IFN g)8(3)21(7)
HT and melphalan
quentially
10
国际象棋怎么下Level of ILP
Axillary/iliac/femoral2/10/92/5/14n.s. Tumour respon(n Z17)a n.s. Complete respon1113
Partial respon22
No change31美味姐姐
Progression15
Acute regional toxicity n.s. Grade I(no reaction)33
Grade II(mild erythema/
edema)
1511
Grade III(mild erythema,
edema and blistering)
36
Grade IV(extensive
epidermiolysis,damage to
deep tissues;threatening
compartment syndrome)
01
Mean(range)hospital stay
(days)
19.5(6–39)n.s.
HT,true hyperthermia(42–438C);IFN g,interferon gamma.
a Four patients had no evaluable lesions since ILP was adjuvant to excision of the primary tumour(n Z2)or in-transit metastas(n Z2).
Table1
MD Anderson stage of dia,adapted for stage II3
Stage I Primary melanoma
Stage IIa Local recurrence
Stage IIb Satellites
Stage IIIa In-transit metastas
Stage IIIb Regional node metastas
Stage IIIab In-transit-and regional node metastas
Stage IV Distant metastas
E.M.Noorda et al./EJSO32(2006)318–324319
months).All acute regional toxicity reactions were between grades I and III,with three(14.5%)grade I,
15(71%)grade II and three(14.5%)grade III reactions.Two patients had neuropathy still prent at the time of the repeat ILP.One of the two had undergone the quential schedule with peroneal nerve dysfunction following thefirst hyperthermic ILP.The other patient had nsory nerve disturbances of the lower leg.Before or at the time of thefirst ILP,15patients had undergone lymph node clearance with two axillary disctions,one inguinal disction,10combined inguinal/ iliac/obturator disctions and two disctions at the iliac/ obturator level.The median time betweenfirst and repeat ILP was17months(IQ range10–30months).Within this interval,16patients had veral recurrences,which were all managed by excision.
Characteristics of repeat ILP
Characteristics of repeat ILP are displayed in Table2. The stage of dia was different from that atfirst ILP,with significantly more regional lymph nodal involvement(stage IIIAB dia)in11out of21patients(52%,P!0.05).The median number of lesions was nine(IQ range5–25).Fifteen patients(71%)had a high tumour load according to the combined Fraker and Rossi criteria.Of the other six patients,one had lesions on the sole of the foot that were considered locally unrectable andfive patients had multiple but less than10small lesions spread over a large area of the extremity.Repeat-ILP was indicated in the patients,becau either new lesions were developing m
ore frequently and progressively,and/or lesions were con-sidered unrectable.All repeat ILPs were performed with TNF a and melphalan,with the addition of IFN g in ven cas.At the time of the repeat procedure,five patients underwent an iliac/obturator lymph node disction.Only one patient did not undergo any lymph node disction at all during the cour of his dia.Median follow-up after repeat ILP was18months(IQ range6–36months). Statistics
Statistical analysis was performed with Student’s t-test when comparing groups with a normal distribution or c2test when proportions were concerned.Non-parametrical tests, mostly the Mann–Whitney test,were ud if the distribution of data was not normal.For limb recurrence-free survival,a Kaplan–Meier analysis was performed and a log-rank test for comparison of differences between independent groups.沙城一小
A P-value of%0.05was regarded as significant.
Results
Tumour respon
In Table2,results of repeat ILP with TNF a and melphalan are shown and compared to tho after the
first ILP.Repeat ILP resulted in a CR in13patients(62%,95% CI39–85%),which was comparable to the65%CR rate after thefirst procedure(95%CI39–90%,P Z0.9).Two patients attained a partial respon,one patient had no change of dia andfive had progressive dia despite repeat ILP.Loco-regional recurrences occurred in nine complete responders(69%)after a median period of9 months(IQ range7–14months)after repeat ILP.The limb recurrence-free survival for tho with a CR was31%(95% CI6–56%)with a median limb recurrence-free interval of 13months(95%CI5–22months).
Seven(64%)of the complete responders to thefirst ILP (n Z11)attained a complete respon after the repeat ILP, whereas two of the six without CR(33%)to thefirst ILP attained a CR after repeat ILP(P Z0.49).
Toxicity and long-term morbidity
Acute regional toxicity after the repeat ILP was not significantly different from thefirst ILP,when mild grade I/II reactions were compared(85and67%,respectively) with vere grade III–IV reactions(15and33%,respect-ively,P Z0.28).No grade V reaction occurred.The patient with a grade IV reaction after repeat ILP,had an impending compartment syndrome following ILP that resolved wit
hout quelae.One patient with a grade II toxicity reaction and a CR,had vere atherosclerotic dia of the perfud extremity with wound healing problems that necessitated amputation of the limb10months after the repeat ILP. There was no neuropathy due to repeat ILP.One patient developed limb malfunction with muscle atrophy and fibrosis after radiation therapy for further limb recurrences. The mean hospital stay was23days(range9–65days), which was not significantly longer than after thefirst procedure(20days,P O0.05).
Management of recurrences or persistent dia after repeat ILP
The17patients with locoregionally recurrent or persistent dia after repeat ILP were treated with excision(s)(n Z10),radiation therapy(n Z3),repeat ILP (n Z7),systemic chemotherapy and/or immunotherapy(n Z 5)and in one patient no treatment was instituted due to progressive distant metastasic dia.
Seven patients underwent a third ILP at a median of19 months after the cond ILP-procedure(IQ range8–55 months).Tumour respon after the third ILP was complete in four patients,partial in one and two patients had progressive dia despite the third ILP.Three patients with CR recurred in the perfud area afterfive,13and19 months of the third ILP.Acute regional toxicity was mild with a gr
ade II reaction infive patients and a grade III reaction in two patients.One patient had wound healing problems after additional radiotherapy and excisions of progressive lesions after ILP and general muscle atrophy
E.M.Noorda et al./EJSO32(2006)318–324 320
and fibrosis of the treated extremity.Another patient underwent a fourth ILP for recurrent dia,42months after the third ILP,which again resulted in a CR with a duration of 3months.
The limb salvage rate for all 21patients was 95%.Seven patients (33%)were rendered locoregionally dia-free until the end of their follow-up with a median duration of 51months (IQ range 23–103months).Five of the patients were dia-free after the repeat ILP and two patients after a third procedure,with additional excisions in four patients.Survival
Overall 5-year survival after repeat ILP was 46%(95%CI 24–68%)with a median survival of 51months (95%CI 19–83months).Median overall survival for patients with CR was 62months (95%CI 36–89months).All patients who had persistent dia after the repeat ILP died after a median of 13months (95%CI 0.4–26months,P Z 0.05).Survival curves are provided in Fig.1.Discussion
After successful treatment of locoregionally recurrent limb melanoma with ILP using melphalan,further recurrences develop in 46–54%of the patients.2,3Persistent dia after unsuccessful ILP with melphalan is prent in approximately 46%of the perfud patients.1Treatment options
Treatment options for recurring or persistent dia after ILP vary depending on the extent of the dia.Further local excisions of recurring lesions,when not too extensive,can provide local control in a substantial number of patients.
If cutaneous or superficial subcutaneous lesions are smaller than 2cm,CO 2lar ablation is also a valuable treatment option.16,17Radiotherapy with or without hyperthermia is an option if the number of lesions is less than ven and they are smaller than 4cm but difficult to rect.18However,if large or numerous (sub)cutaneous lesions are involved,ILP ems the only alternative to avoid amputation.In the prent ries,repeat ILP was performed in 71%of the cas for high tumour load lesions,according to recently defined criteria,15,19or unrectable lesions becau of their location.
The advantage of ILP compared to the other local treatments is that the whole area at risk of recurrence is treated,with the possibility of eradicating microscopic dia that may be prent in th
e affected limb.The longer dia-free interval after ILP compared to excision alone for locoregionally recurrent melanoma,as obrved in veral studies,7,20,21ems to be the result of this additional loco-regional tumour-cell-killing effect.Respon and limb recurrence after repeat ILP as compared to other studies
In the prent study,the CR rate of repeat ILP with TNF a was 62%which ems comparable to the 74%CR rate reported from our previous repeat ILP ries and in the lower range of the 58–90%CR rate that has generally been reported after TNF a -ILP.4,14,22–25Compared to melphalan alone ILP,with its 54%CR rate,62%is a good result in view of the extent of the dia in the prent ries,and is probably the result of the addition TNF a .The addition of IFN g in some patients has probably had no significant effect on the respon rates since a randomid pha II trial demonstrated no improvement in respon rates after its addition to TNF a .25Our limb recurrence rate after CR of 69%is high compared to the limb recurrence rate of 38–58%after TNF a -ILP in other ries.14,22–24This can be due to the longer duration of follow-up of the prent ries and reflects the strong tendency for the dia to recur in our specific group of patients.In a retrospective analysis performed by Feldman et al.,7the management of limb recurrences after ILP was described in 28patients.Repeat ILP was performed in six patients,resulting in five CRs with recurrences occurri
ng in three of them.Bartlett et al.showed a 50%CR rate after repeat TNF a -ILP in a ries of 10patients with persistent dia after the initial ILP with melphalan alone.6
The median duration of CR of 13months in the prent study is comparable to the 12–17months reported after initial ILP with TNF a .7,14,15,20In our patient population with a high tendency for locoregionally recurrent melanoma and failure after earlier ILP,a third of the patients were rendered locoregionally dia-free by repeat ILP(s),with or without additional excisional surgery,for a median period of more than 4years.All
84
7260483624120100
80
60
40
20
Figure 1.Overall survival after repeat ILP for all patients,and parately for tho with CR and no CR by repeat ILP.X -axis,time after repeat ILP (months);Y -axis,cumulative survival (%);numbers of patients at risk are provided at each point in time.
E.M.Noorda et al./EJSO 32(2006)318–324
321
patients who were locoregionally dia-free at the end of follow-up,had attained a complete respon on the repeat ILP.This is a result that is generally not attained with excision alone in the
patients groups.In 14patients,Feldman et al.found only two patients who remained locoregionally dia-free until the end of follow-up with excisional surgery alone.7Hill et al.described that local control could be obtained in 56%of 32patients with extensive in-transit metastas after multiple ssions of CO 2lar ablation within a period of up to two and a half years.16Drugs in repeat ILP
In Table 3an overview of studies on repeat ILP with various drugs is shown.Repeat ILP can be effective despite a failure of a previous ILP,becau it may be that recurrent dia after ILP consists of grown lesions that were not yet vascularid at the time of the first procedure,which have to become nsitive to therapeutic agents at the time of the repeat procedure.2,6It remains unclear whether TNF a needs to be added to melphalan in the repeat tting.Bulky,well vascularid lesions have been suggested to be particularly susceptible to TNF a .26Since,it is likely that recurring or persistent lesions after previous ILP are more extensive,TNF a could be especially uful in the repeat procedures.Repeat ILP with melphalan alone showed a 74%CR rate in 19patients with measurable melanoma lesions confined to the limbs,the majority of whom where perfud with a single normothermic ILPs.Double,triple or quential normo-and hyperthermic ILP schedules where also applied in this study but have been abandoned since the introduction of TNF a in ILP.4In arch of other drugs in repeat ILP,Hoekstra ated ven patients with recurrent melanoma after prev
ious perfusion with a repeat ILP with cisplatin.They concluded,however,that a high level of neurotoxicity and limb recurrence rate did not justify the u of cisplatin in ILP.27Vaglini et al.achieved a 44%CR rate in 25patients with various cytostatic drugs and biological respon modifiers.28Bonenkamp et al.have treated 13patients with repeat isolated limb infusion with fotemustine and systemic dacarbazine after failure of
a previous perfusion.Becau four patients attained a CR and six patients had vere lim
b toxicity,they concluded that it could not be recommended as a routine method of treatment.30
Acute regional toxicity
Acute regional toxicity after repeat ILP was not higher than after the first ILP.This is a confirmation of what others have found,7although more vere toxicity has been encountered in our previous report on repeat ILPs with melphalan using various schedules.Probably,the sche-dules are responsible for the incread toxicity.Survival
Our 46%5-year survival rate is considerable,but should be interpreted cautiously becau of the relatively small group of patients.Generally,5-year survival for patients with in-transit metastas with
or without regional lymph-node metastas ranges between 27and 46%.31In a large ries of patients who underwent a first ILP for the stages of dia,Krementz ported 5-year survival rates of 23–36%,depending on the prence of lymph-node metastas.32Lejeune et al.found that among 206patients treated with ILP for locoregionally recurring melanoma,distant metastas occurred in only 24%of them as the first site of recurrence.33This confirms our impression that there is probably a subgroup of patients with a rather good prognosis,with recurrences that remain limited to the loco-regional area without distant spread for a long period of time.Nevertheless,all our patients with residual dia after ILP died and the favourable prognosis applies only to patients who attain a CR to ILP,which has also been found by others.34It is also this group of patients that ems to have the best loco-regional benefit of a repeat ILP.Conclusion
Repeat ILP with TNF a is feasible in patients with recurrent dia after previous ILP.It is indicated for patients not suitable for other,less invasive,treatment
Table 3
Complete respon rates after repeat ILP for melanoma Reference n %ILP-type
Vaglini,1993282544Various drugs and ILP-schedules Hoekstra,199327732a Cisplatinum公司祝福
Klop,19964
1974Melphalan:single/double/triple normothermic (37–388C),single mild hyperthermic (38–408C),quential high-do hyperthermic (41.5–438C)Bartlett,199761765Melphalan and TNF a ,mild hyperthermic (39–408C)Feldman,19997683Melphalan G TNF a ,mild hyperthermic (38.5–408C)Grunhagen,200529
25
76
Melphalan and TNF a ,mild hyperthermic (38–408C)
a
Two of four reached CR,in three patients lesions were excid and respon was not measurable.
E.M.Noorda et al./EJSO 32(2006)318–324
322
