Susceptibility of Fibromatosis Cells in Short-Term Culture to Ifosfamide: A Possible Experimental Treatment in Clinically Aggressive Cases

Purpose. Deep fibromatoses are large, often rapidly growing but benign soft tissue tumours. Although surgery is the mainstay of treatment, in unremitting and aggressive cases the use of cytotoxic chemotherapy may produce objective tumour responses. Fresh tumour samples from four patients with fibromatosis were investigated as part of a study of drug resistance in soft tissue tumours. Methods. Following short-term culture of fibromatosis cells in vitro , chemosensitivity to 4-hydroperoxy-ifosfamide, the active form of ifosfamide and doxorubicin was tested. Following 72-h continuous exposure to each drug, surviving cell fraction was assessed using the lactate dehydrogenase assay. Results. Mean IC50 values for ifosfamide and doxorubicin were 6.2 and 0.35 µmol/l, respectively. In samples of soft tissue sarcoma (STS) from the same study the mean IC50 values for ifosfamide and doxorubicin were 14.8 and 1.69 µmol. The difference in mean ifosfamide IC50 values for fibromatosis and STS samples was statistically significant. Discussion. We are not aware of any other report suggesting the use of ifosfamide in this condition. These observations suggest that, for patients with inoperable or progressive lesions of fibromatosis causing significant morbidity, it may be valuable to include ifosfamide in experimental treatment regimens.


Introduction
T he deep ® b rom atoses, also know n as desm oid tum ours, are large, often rapidly grow ing but benign soft tissue tum ours. They are rare with a reported incidence of 0.2± 0.5 per 100 000 population per year, m ost com m only affe cting young adults, especially wom en. Based on their anatom ical location, the deep ® bromatoses can be divided into three subtypes, extraabdominal, abdom inal and intra-ab dom inal. Extraabdominal ® brom atosis arises principally from the connective tissue of m uscle and the overlying fasc ia or aponeurosis and chie¯y affects the m uscles of the shoulder and pelvic girdles and the thigh. Abdom inal ® brom atosis arises from the m usculo-apo neurotic structures of the abdom inal wall and tends to occur in wo m en o f child bear in g age d u rin g o r after pregnancy, leading to the hyp othesis that horm onal factors are im portant in the aetiology of the condition. Intra-abd om inal ® brom atosis arises inside the abdominal cavity and includes pelvic and m esenteric ® b ro m ato sis an d ® b ro m ato sis in G ar d n er ' s syn d ro m e. 1 H isto lo g ica lly th ey are a ll p o o rly circumscribed in® ltrating tumours comprising spindle cells regularly dispersed in collagen, without atypical or hyperchrom atic nuclei but with ver y occasional m itotic ® gures. T he presence of num erous or atypical mitotic ® gures should arouse suspicion of a m alignant lesion. Following the typical presentation with a soft tissue m ass, diagnosis should be con® rm ed, preferably by needle core biopsy, before treatment is undertaken. Radical surgical excision is the treatm ent of choice du e to the in® ltrative natu re o f th e tu m o u r. 2± 4 However, because the tum ours are non-m alignant with potential for prolonged sur vival, even in the presence of bulky disease, it is desirable to avoid m utilating su rgery. 5,6 L ocal recurrence occurs in approxim ately 40% of cases rising to 90% after m a rg in a l exc isio n . 5,7± 10 R a d io th era py h a s a n established role both as an adjunct to surgery and for inoperable cases. 11± 14 Tumour regression following radiotherapy m ay be prolonged. Spontaneous regression of growth and a`plateau phase' have been observed in cases of ® brom atosis with residual or inoperable disease after surgery. 5,6 This observation contributes to the reluctance of clinicians to use cytotoxic treatments routinely in the management of an essentially benign disease. M any clinicians feel that the risk of treatment-related side effects, particularly the long-term risk of induction of malignant disease m akes this form of treatment unacceptable. However, in unremitting and aggressive cases the use of horm one or cytotoxic chem otherapy can more easily be justi® ed and there are several published repor ts of the successfu l use of anti-o estrogens, including toremifene and tam oxifen, 15± 18 progesterone, 19 non-steroidal anti-in¯am matory prostaglan d in in hibitin g d r u gs 16 an d var io u s c ytotoxic chem o th erapy re gim ens 20± 23 to con trol cases o f ® brom atosis. Although m any different chem otherapy regim ens have been used, there is no clear ag reement between investigators on which regim en is best. In this report, we present in vitro chem osensitivity data for tum our cells growing in short-term culture d erived from sam ples of excised tum our in four cases of ® brom atosis treated by surgery. Tum our cells were exposed to doxorubicin and ifosfam ide, the two ® rst line drugs most com monly used in the treatment of adult soft tissue sarcom a (ST S) and the results are com pared with typical m alignant ST S sam ples.

Patients
Patients presented here were investigated as part of a research program m e within the Institute of Cancer Research and the Royal M arsden Hospital to study mechanism s of drug resistance in patients with soft tissue tumours. T his project has been app roved by the local ethics comm ittee. Sterile sam ples of fresh tum our were collected from the operative specim ens of patients undergoing prim ary surgery for soft tissue tum ours. These sam ples were studied in the laboratory and the results were correlated w ith the clinical outcom e of the patients. T he four patients in the series foun d to have a pathological diagno sis of ® brom atosis and evaluable for in vitro cytotoxicity are described.

Processing of tumour material
Sam ples of fresh sarcom a tissue were collected under sterile conditions into L 15 Leibovitz culture m edium (G ib co L ife Tec hn o lo g ies, Pa isley, S co tlan d ) containing the antibiotics gentam icin and am photericin B. Sam ples were processed within 48 h of collection as follows: tissue was subjected to m incing using sterile crossed scalpels, followed by digestion at 37Ê C using collagenase (Sigm a C hem icals, Poole) at a working concentration of 500± 1000 IU/m l, diluted in L15 m edium . T he duration of digestion varied according to the rigidity and texture of the sam ple and was between 30 m in and 2 h. Erythrocytes were rem oved using a density gradient (H istopaq ue-1077; Sigm a C hem icals), and further m echanical disaggregation and cell washing was carried out in order to achieve a suspension w ith m inim al cell clumping and debris.
Fibrom atosis cells were cultured in tissue culturē asks and m aintained in an environment of 5% carbon d io xid e at 37Ê C , in H am ' s F 12 m ed iu m w ith G lutam ax (G ibco L ife Technologies), supplem ented with 20% heat-inactivated foetal calf serum (FC S) and the antibiotics gentam icin and am photeracin B. C u ltu res w ere a llow ed to eq u ilibrate a n d w ere subjected to several changes of m edium in order to rem ove further debris over a period of at least 48 h. All cultures grew as attached m onolayers. Sam ples of the fresh tum our cell suspensions initially seeded into tissue culture¯ask s were used to m ake cytospin preparation s u sing a S han do n C yto spin 3 (L ife Sciences International U K Ltd., Basingstoke). G iemsa stained cytospins were used for veri® cation of tum our cell content w hich was greater than 80% in all cases.

Chemosensitivity testing
Fresh tumour cells growing as m onolayer cultures w er e tr y p sin ised ( tr y p sin ± E D TA , G ib co L ife Technologies) to produce a single cell suspension in H am's F12 m edium containing 10% FCS. Cells were dispensed into 96-well tissue culture plates in 200-m l aliquots at a density of 2 3 10 5 /m l± 1 3 10 6 /m l. Cultures were allowed to equilibrate for 24 h in an hum idifying gassin g incubator. D oxorubicin (Sigm a C hem icals), obtained as the hydrochloride salt, was m ade up in sterile distilled water and stored as frozen aliq u o ts. Ifo sf am id e (A S TA M ed ica, F ran k fu r t, G erm any), obtained as the activated 4-hyd roperoxy fo r m , w a s d isso lve d in ster ile d istilled w ater im m ediately before use. Freshly prepared drug solutions diluted in tissue culture m edium containing 10% FC S were dispensed in aliquots to give the desired ® nal concentration, over a dose range of 0.1± 2 m M for doxorubicin and 0.5± 100 m M for ifosfam ide. The cultures were then incubated at 37Ê C in a hum idifying gassin g incubator for 72 h, with the cells continuously exposed to the drugs.
Trays of drugs-treated cells were assessed for cell viab ility using an LD H-release assay kit (Cytotox96, Prom ega Corporation, Southam pton, U K) which we have adapted for use in cytotoxicity testing. 24 Spent tissue culture m edium was aspirated and replaced with 200 m l of serum -fre e phenol red-free D ulbecco's m ed iu m ( G ib co L ife Techn o lo gies). C ells w ere ruptured by freezing to ± 70Ê C and then allow ing to thaw at room temperature. 100 m l of cell lysate was then rem oved from each well and dispensed into a duplicate tray. F ifty m icrolitres of LD H substrate m ixture were added to each well and the colour reaction allowed to take place for 30± 60 m in at room temperature in the dark. F ifty m icrolitres of`Stop solution' from the kit were added to each well and colour abso rbance at 492 nm (red) was m easured using a plate reading spectrophotom eter. Absorbance readings of control untreated cell wells were com pared to drug-treated wells. D ata were calculated as absorbance of test wells relative to control wells (m ean of at least three replicates), with IC 50 denoting the concentration of drug required to cause a 50% loss of cell viab ility relative to control. These values were calculated from plots of log 10 drug concentration (x axis) vs fraction of control absorbance (y axis). C ytotoxicity testing results for ® bromatosis sam ples were com pared with IC 50 values from 18 evaluable cases of soft tissue sarcom a collected in the sam e study. T he rem ainder of the STS specim ens and the one other ® brom atosis specim en collected failed to grow or becam e contam inated m aking cytotoxicity testing im possib le. N one of the patients with STS had previously received chemotherapy. Several of the soft tissue sarcom a specimens had IC 50 values for ifosfam ide and/or doxorubicin above the lim it of detection of the cytotoxicity assay. By convention, IC 50 values for these cases were set at the m axim um detectable value, typically 50 m m ol/l for ifosfam ide and 5 m m ol/l for doxorubicin. T he IC 50 values for sam ples of ® brom atosis and soft tissue sarcom a were com pared using the statistical com puter software package SPSS (SPSS Inc., Chicago, IL) by the (nonparam etric) M ann± W hitney U-test w hich uses variable rank and so is not affected by the distortion of the STS sample m ean which was produced by lim iting m axim um IC 50 values.

Results
O ut of 97 fresh tissue sam ples collected, the diagnosis of ® brom atosis was m ade in four cases. A photograph of ® brom atosis cell growth in culture is shown in Fig.  1(a) and the corresponding histological section in F ig. 1(b). T he characteristics of the patien ts are detailed in Table 1 and the corresponding chem osensitivity testing results are presented in Table 2. The m ean IC 50 for doxorubicin in cases of ® brom atosis was 0.35 m m ol/l. T he m ean IC 50 for ifosfam ide was 6.2 m m ol/l.T he corresponding values for ST S samples were 1.69 and 14.8 m m ol/l, respectively. U sing the M ann± W hitney U-test there was a signi® cant difference between the IC 50 values in STS and ® brom atosis for ifosfam ide (p < 0.01) but not for doxorubicin (p = 0.1). C om parative values for the two drugs are presented in Fig. 2(a) and (b).

Discussion
Com pared to a panel of soft tissue sarcom a cells in culture, the IC 50 values for ® brom atosis cells exposed to ifosfam ide and doxorubicin were lower and, for ifosfam ide, this difference was statistically significant. In STS we have previously shown that a high value IC 50 for the drugs is asso ciated with clinical drug resistance. 25,26 We have been unable to m ake a sim ilar com pariso n for cases of ® brom atosis, no patient has required system ic therapy. Previou s re por ts have dem onstrated the clinical activity of chemotherapy in ad van ced ca se s o f ® b ro m ato sis. A w id ely u sed chem otherapy strategy is the`Philadelphia' regim en of low-dose weekly m ethotrexate (50 m g/week) and vinblastine (10 m g/week). 20,22 The activity of m ore aggressive regim ens based on doxorubicin and dacarba z in e h a s a lso b een d em o n strated 21,23,27 an d additionally, in children, the com bination of dacarbazine with vincristine and cyclophospham ide (VAC) has been shown to be effective. 28 We did not have sufficient clinical m aterial availab le to test ® brom atosis cells against these other cytotoxic drugs.  O ur in vitro observations support the inclusion of doxorubicin in regim ens for the treatment of ® brom atosis. We have also demonstrated the sensitivity of ® brom atosis cells, at least in tissue culture, to ifosfamide.We are not aware of any other reports suggesting that ifosfam ide is a useful drug in this condition, and therefore this introduces the possib ility of a new app roach to treatment using ifosfam ide in ® rst line drug com binations. For sm all, well-circum scribed lesions of ® brom atosis, the treatment of choice should be su rger y. C hem otherapy and radiation therapy shou ld be avoid ed becau se of treatm en t-related to xicity. H ow eve r, fo r patien ts w ith in o p erab le prog ressive lesions, causing signi® cant sym ptom s or posing a threat to norm al organ function, it m ay be valuable to include ifosfam ide in experim ental treatm ent regim ens.