- Email: [email protected]
Intraarterial chemotherapy as neoadjuvant treatment of oral cancer
Journal of Cranio-Maxillofacial Surgery (1999) 27, 302±307 # 1999 European Association for Cranio-Maxillofacial Surgery
Intraarterial chemotherapy as neoadjuvant treatment of oral cancer AdorjaÂn F. KovaÂcs,1 Bernd Turowski,2 Mostafa T. Ghahremani,1 Matthias Loitz1 1
Department of Maxillofacial Plastic Surgery (Head: Prof. Dr Dr K. Bitter), Johann Wolfgang Goethe-University Medical School, Frankfurt am Main, Germany; 2Institute for Neuroradiology (Head: Prof. Dr F. E. Zanella), Johann Wolfgang Goethe-University Medical School, Frankfurt am Main, Germany SUMMARY. Neoadjuvant chemotherapy in patients with primary squamous cell carcinomas of the oral cavity should lead to high remission rates whilst having low morbidity. Ecacy can also be enhanced by treating small tumour stages. As part of a multi-modality therapy of all stages of primary oral cavity carcinoma, 103 patients were treated with neoadjuvant intraarterial (i.a.) chemotherapy. After regimen A with 100 mg/m2 i.a. cisplatin followed by 5 day continuous intravenous infusion of 5-¯uorouracil (1 g/m2 per day) in 36 patients, an i.a. high pressure chemo-perfusion with a dose of 150 mg/m2 cisplatin was used with simultaneous intravenous infusion of 9 g/m2 sodium thiosulfate (regimen B, 67 patients). Subsequent treatment comprised radical surgery and simultaneous radiochemotherapy with docetaxel. Partial and complete remissions were found in 80.6% (regimen A) and 67.2% (regimen B) of cases, tumour growth was inhibited in 11.1% and 31.3%. Very low toxicity could be shown especially in regimen B. 66.7% and 74.6% of patients could be operated on radically. Survival rate was 61.1% (regimen A, 22.7 months of mean observation time) and 79.1% (regimen B, 8.4 months). Patients with highgrade remissions seemed to have a survival advantage. Neoadjuvant i.a. chemotherapy with cisplatin, especially in its high dose variant, is a practical therapeutic tool for the treatment of all stages of primary oral cavity carcinoma. # 1999 European Association for Cranio-Maxillofacial Surgery
systems successfully prevented complications of intraarterial (i.a.) chemotherapy, such as infections, catheter misplacement, thromboses and haemorrhages. Robbins et al. (1992) started to combine these concepts by using a high dose chemoradiotherapy with cisplatin and a simultaneous application of sodium thiosulfate for systemic, competitive neutralisation of the cytostatic (Goel et al., 1989). Generally, the indication for this treatment was thought to be palliative. However, its eect on the survival rate of all patients with oral squamous cancer can only be investigated by using i.a. induction chemotherapy for smaller tumour stages in addition, as was done with systemic chemotherapy in a few earlier studies (Martin et al., 1990; Volling and SchroÈder, 1995). This paper presents the neoadjuvant i.a. chemotherapy with cisplatin as part of a multi-modality treatment concept of 103 patients with untreated primary squamous cell carcinomas of the oral cavity. All tumour stages were treated using this method. A local remission was intended with reduction both of tumour associated symptoms, and, possibly, aggressiveness of the tumour concerning cell spread. The therapy was planned to have as low systemic toxicity as possible for prevention of prognostically relevant morbidity (due to accumulation of side-eects). The ®rst regimen combined the i.a. application mode with a systemic chemotherapy while the second regimen used high dose cisplatin alone in combination with sodium thiosulfate for simultaneous systemic neutralisation.
INTRODUCTION The treatment of squamous cell carcimomas of the oral cavity is not satisfactory in terms of patient survival. This is a reason for the many dierent treatment plans commonly used consisting of varying combinations of surgery, radiation and chemotherapy. Systemic chemotherapy regimens with cisplatin and 5-¯uorouracil appear to be very ecient locally (Rooney et al., 1985) and in the subsumed prevention of metastatic spread (Jaulerry et al., 1992). But the advantage of induction chemotherapy for the patients is still considered controversial (Stell and Rawson, 1990; Munro, 1995), being associated with severe toxicity and morbidity potentially relevant to the survival rate. The concept of `down staging' of a tumour could not be con®rmed and radical surgery is still indispensible (Poulsen et al., 1996). Another problem of primary chemotherapy was the development of resistant cell lines which could be overcome by higher doses (Teicher et al., 1987). The reduction of peripheral toxicity combined with high local ecacy was a further aim (Baker and Wheeler, 1983), especially important in oral cancer patients who often suer from other intercurrent illnesses. For 50 years regional chemotherapy using the arterial route in the head and neck region, tried to ful®l these claims and various cytostatic agents and technical means were used. Most indications were palliative treatment of advanced cancer stages and recurrences (Sullivan et al., 1953; Bitter, 1976; Eckardt and Kelber, 1994). Development of ®ne catheter delivery 302
Intraarterial chemotherapy as neoadjuvant treatment of oral cancer 303
PATIENTS AND METHODS The study started in December 1996. Results of this study date from May 1999. Two unselected groups were created: 36 consecutive patients (77.8% male and 22.2% female, average age 57.3 years) received i.a. chemotherapy with 100 mg m2 cisplatin without systemic neutralisation followed by an intra-venous (i.v.) continuous infusion of 1 g/m2 5-¯uorouracil for functional synergism (regimen A). The second group of 67 consecutively treated patients (83.6% male and 16.4% female, average age 59.9 years) received i.a. high pressure chemotherapy with the high dose of 150 mg/m2 cisplatin over 5±15 min in combination with parallel i.v. application of 9 g/m2 sodium thiosulfate after a delay of 10 s (regimen B). To avoid resistance development in both groups, a maximum of 3 cycles were scheduled, if no grade III or IV remission occurred at the ®rst cycle. Remissions were de®ned: 0 ± tumour progression, I ± stable disease (no tumour growth), II ± reduction of tumour mass 5 50% (lowgrade partial remission), III ± reduction of tumour mass 4 50% (high-grade partial remission), IV ± complete remission (disappearance of the tumour). The assessment was done by inspection, palpation and CT examination after the ®rst cycle. Grades of remission which might have been higher after the second and third cycles were not considered in this study to make results comparable. No biopsies were taken after chemotherapy to avoid interference with the planned operation. Location of the primary tumours were typically in the ¯oor of the mouth (about 40%) and the tongue (about 20%). Performance status of the patients (Table 1) was good in a high percentage, although the majority of them presented advanced stages of disease (Table 1). Nevertheless, between 30% and 40% had T1 and T2 tumours. All tumours were histologically con®rmed squamous cell carcinomata; three quarters were moderately dierentiated. Similarity of the two patient groups was best shown using the therapy-dependent prognostic index TPI which combines data on tumour size, in®ltration, regional lymph nodes, distant metastases and age (Platz et al., 1983) (Fig. 1). After chemotherapy, the multi-modality treatment plan provided radical surgery of the mouth and neck, if operation was feasible depending on the general or local condition. A so-called `down staging' of the tumour was impossible. Surgery was followed by fractionated simultaneous radio-chemotherapy of the tumour area and the neck lymphatics over 5 weeks (51.3 Gy at 1.9 Gy per fraction) with systemic application of docetaxel (25 mg/m2) once a week. This last treatment modality was oered to all patients, with or without surgery. Pretherapeutic staging included the use of palpation, US, CT and MR imaging for examination of the neck lymph nodes and PET for diagnosis of second tumours, neck lymph nodes and distant metastases. On the morning of treatment, patients were given 74 mg
Table 1 ± Characteristics of the patient population according to ECOG (Eastern Cooperative Oncology Group) for performance status and UICC (Sobin and Wittekind, 1997) for tumour staging Grade
0 1 2 3 4
Performance Tumour (T) status (%) (%)
Nodes (N) (%)
Tumour stages (%)
A
B
A
B
A
B
A
B
61.1 22.2 11.1 0 5.6
70.1 15 13.4 1.5 0
/ 11.1 25 0 63.9
/ 17.9 30.8 1.5 19.8
38.9 16.6 38.9 5.6 /
47.8 22.4 26.9 2.9 /
/ 11.1 11.1 8.3 69.5
/ 10.5 23.9 11.9 53.7
dolasetron and 75 mg prednisolone i.v. Afterwards, 1.5 litres of a full electrolyte solution (with 20 mval potassium chloride) were given by i.v. infusion over 2 h. Then catheterisation of the right femoral artery was carried out using a size 4 (french) catheter containing a coaxial micro-catheter. After superselective visualisation of the tumour-feeding vessel using ¯uoroscopy and a contrast medium, either 100 mg/m2 cisplatin dissolved in 500 ml 0.9% saline solution was infused i.a. over 1 h (regimen A), or 150 mg/m2 cisplatin dissolved in the same amount of saline solution was infused with controlled pressure (2 ml/s) (regimen B). For analgesia, 0.1 to 0.3 mg fentanyl was delivered i.v. (and on occasions 5 to 15 mg mepivacain) into the perfused artery. With a delay of 10 s, an i.v. infusion of 9 g/m2 sodium thiosulfate was given in parallel. After the treatment, 1 litre of full electrolyte solution with 20 mval potassium chloride was infused i.v. over 5 h. The next day, the patients were hyperhydrated with 3 L of a two-third electrolyte solution, thrombosis prophylaxis with heparin S.C., also dolasetron i.v. if necessary. Routine laboratory checks were made on alternate days. Ward stay lasted between 4 and 6 days for most patients of regimen B. In regimen A, this time was prolonged by the 5 day continuous infusion of 5-¯uorouracil (1 g/m2). Daily application of allopurinol (300 mg) and anti-emetic drugs was mandatory. The side-eects of the cycles were noted. The sideeects of the second and third cycles were always less pronounced than of the ®rst cycle.
RESULTS There was a high overall response rate of 91.7% in regimen A and of 98.5% in regimen B (Fig. 2). Tumour progression was observed in only 8.3% and 1.5% of cases respectively. The synergy of cisplatin and 5-¯uorouracil had a higher impact on partial and complete remissions (55.6% and 25%) when compared with the high dose cisplatin therapy (64.2% and 3%). Therefore, the rate of stable disease was much higher in scheme B (31.3%). More than 75% of these patients had only a single cycle. Repeated courses were used for cases with very large tumours
304 Journal of Cranio-Maxillofacial Surgery
Fig. 1 ± Distribution of therapy-dependent prognosis index TPI for the examined patient population. White columns: regimen with i.a. cisplatin and i.v. 5-¯uorouracil. Black columns: regimen with i.a. cisplatin and i.v. sodium thiosulfate.
Fig. 2 ± Grades of remission after ®rst cycle of i.a. chemotherapy. White columns: regimen with i.a. cisplatin and i.v. 5-¯uorouracil. Black columns: regimen with i.a. cisplatin and i.v. sodium thiosulfate.
which were considered to be inoperable. There was no case of a change of remission grade after repetition of a cycle. Therefore, repeated i.a. chemotherapy provided local control. The planned operation could not be carried out in 29 cases (12 in the ®rst and 17 in the second group). Five patients died due to advanced malignancy despite temporary local control. During pre-therapeutic examination, the tumours of 9 patients were considered to be inoperable because of involvement of the skull base and/or vertebrae, or in®ltration of
the whole tongue and ¯oor of the mouth including the common carotid artery. Eight patients had very poor general condition with multiple systemic diseases. One patient suered from alcoholic myocardial insuciency, one had inoperable bronchial malignancy and three had pulmonary metastatic spread. For each patient, the possibility of a simultaneous radio-chemotherapy was considered. If no further therapy could be oered, the i.a. chemotherapy was de®ned as palliative. Acceptance of the i.a. chemotherapy was high. This could be explained by the
Intraarterial chemotherapy as neoadjuvant treatment of oral cancer 305
After an observation period for the group with regimen A of 17.5+8.1 months on average (maximum 30 months), 14 patients were dead (after 9.4+6.7 months) and 22 (61.1%) alive (after (22.7+ 2.9 months). For regimen B, observation time was 7.8+5.2 months (maximum 20 months). Fourteen patients were dead (after 5.4+2.8 months) and 53 (79.1%) alive (after 8.4+5.5 months). Looking at the relationship of grades of remission and survival (Table 3), there was a tendency for more patients with low-grade remissions to die prematurely while more patients with high-grade remissions were still alive.
fact that 26.9% of all patients of regimen B had no clinically detectable acute side-eects (compared with 14.7% in regimen A). Therefore, only 2 patients refused the recommended continuation of the therapy. Surgical resection was undertaken within 3 weeks of the last dose of chemotherapy. Of the patients in regimen A, 66.7% could be operated on (resection with histologically con®rmed healthy margins). In regimen B, the rate was 74.6%. Compared to the stage-grouping (Table 1), this meant that many patients with advanced tumours were operated on. Twelve suprahyoid lymph node dissections were carried out in the group with regimen A and 40 in the group with regimen B, while the rate of functional neck dissections was lower (11 times/18 times). There were no unusual postoperative complications except a higher rate of tracheotomies in the group with regimen B (24 cases 48%) when compared to regimen A (3 cases 12.5%), which may be due to enhanced swelling of the area perfused with high-dose cisplatin. The side-eects of both i.a. chemotherapy regimens are presented in Table 2. They are divided into gastrointestinal side-eects (nausea/vomiting/diarrhoea), haematological side-eects (anaemia/leukopaenia/thrombopaenia), disturbances of electrolytes (potassium, ferritin), renal and hepatic side-eects (according to WHO: Miller et al., 1981). The regimen B was clearly superior in nearly every case. Only the serum potassium showed a more frequent decrease and the renal side-eects were equal in spite of the much larger dose in regimen B. Creatinine never exceeded a serum value of 1.9 and showed normalisation of values after 5 days on average. Only in regimen A was it necessary to interrupt therapy due to hyperemesis and it had to be supplemented by erythrocyte transfusions or colony stimulating factor (CSF) (twice). Other side-eects were partial temporary alopecia (15% in regimen A and 19% in regimen B) and high tone hearing loss (assessed subjectively in 10% in regimen A and 15% in regimen B). One patient with alcoholic cardiomyopathy developed cardiac decompensation after i.a. chemotherapy. In 212 catheter interventions, only one side-eect was noted (apoplexy possibly due to an arteriosclerotic plaque or thrombus which could have been dislodged by the micro-catheter).
DISCUSSION Chemotherapy cannot be the sole therapy for oral cancer (despite initially high remission rates) if curative treatment is planned. Radical surgery is indispensable (Poulsen et al., 1996) and is not compromised by preceding systemic chemotherapy (Lore et al., 1989). These assertions are valid for i.a. chemotherapy, too, which has enjoyed a renaissance recently. Its advantage is a higher possible local dose of the cytostatic agent compared to the systemic route if a competitive peripheral neutralisation is used (Robbins et al., 1992; Robbins et al., 1994a). Animal experiments and clinical investigation show a better eect from the i.a. route compared to the i.v. route (Bitter, 1976; Harker and Stephens, 1992). The results con®rm the high initial local eectiveness of i.a. cisplatin (Mortimer et al., 1988), especially when in high dose (Robbins et al., 1994a, b). The comparison of two unselected groups regarding staging and number of non-operable patients seemed possible (as TPI and the percentage of patients actually operated on showed) but the main task of the present study was ecacy and practicability of i.a. chemotherapy. Overall, response was better in regimen B. To date, the higher complete remission grades in the regimen A do not seem to be decisive, because they have to be seen in the context of a multimodality therapy and are countered by much greater side-eects when compared with regimen B (Table 2), although less than conventional systemic chemotherapy (Bachaud et al., 1993). It has to be stressed that, in this study, remissions were assessed clinically and not histologically. Nevertheless, the results of this study demonstrate a connection between high-grade
Table 2 ± Side-eects of intraarterial chemotherapy. Group A: regimen with i.a. cisplatin and i.v. 5-¯uorouracil. Group B: regimen with i.a. cisplatin and i.v. sodium thiosulfate. Grades according to WHO (Miller et al., 1981) Grade Nausea (%)
A I II III IV
B
Diarrhaea (%)
Anaemia (%)
Leukopaenia (%)
Thrombocytopaenia (%)
Sideroemia (%)
Hypokalaemia (%)
Serum creatinine (%)
Hepatic enzymes (GOT, GPT, GGT) (%)
A
B
A
B
A
B
A
B
A
B
A
B
A
A
1.5 0 0 0
13.9 8.3 0 0
4.5 13.9 1.5 8.3 1.5 2.8 0 0
4.5 0 0 0
8.3 5.6 0 0
0 1.5 1.5 0
22.2 0 0 0
17.9 0 0 0
16.7 0 0 0
22.4 0 0 0
33.3 34.3 13.9 0 1.5 0 0 0 0 0 0 0
38.9 17.9 2.9 11.1 11.9 0 5.6 0 0 0 0 0
B
B 10.5 0 0 0
306 Journal of Cranio-Maxillofacial Surgery Table 3 ± Relation of grades of remission and survival. Observation times are mentioned in the text Grade of remission
Regimen A Dead
Living
Dead
Living
0 1 2 3 4
3 2 7 1 1
0 2 6 7 7
1 (1.5%) 6 (8.9%) 7 (10.4%) 0 0
1 15 20 13 4
(8.3%) (5.6%) (19.4%) (2.8%) (2.8%)
Regimen B
(5.6%) (16.7%) (19.4%) (19.4%)
(1.5%) (22.4%) (29.9%) (19.4%) (6%)
remissions and survival similar to those reported by Ervin et al. (1987) who stated a correlation of response to systemic induction chemotherapy with failure-free survival. Dierent locations of primary tumours, dierent mixtures of clinical stages of the treated populations and dierent drugs and administration routes are the reasons for diculties in comparison. The fact that remissions mainly occurred after the ®rst cycle could allow the early distinction of responders from non-responders. The assessment of remission, prognostically, could be another future task for i.a. chemotherapy, as well as a starting point for dierential therapy. The actual aim of i.a. chemotherapy is not necessarily complete remission, but to restrain the tumour in its local and possibly metastatic aggressiveness before radical surgery without the very high side-eects of systemic chemotherapy. Therefore, stabilization of disease was also de®ned as a response. The eradication of known micro-metastases is still a problem. Although other side-eects show that free cisplatin reaches the peripheral organs and it is known that i.a. chemotherapy without a neutralising agent can reach the same peripheral cytostatic levels (Sileni et al., 1992), doubts exist about i.a. chemotherapy in this regard (Forastiere et al., 1987). Therefore, subsequent treatment with this multimodality therapy consists of a simultaneous radiochemotherapy with weekly systemic application of docetaxel. Postoperative radio-chemotherapy was reported to show improvement of survival (Bachaud et al., 1991). Nevertheless, locoregional recurrence during the ®rst 2 years after treatment remains the main problem in oral cancer patients (Brady and Davis, 1988). The present study focuses on the feasibility and acceptance of i.a. chemotherapy. The data requirements regarding side-eects compared well with other studies (Eckardt and Kelber, 1994; Robbins et al., 1994a; Kerber et al., 1998) and serve to demonstrate that i.a. chemotherapy, especially the high dose variant, was particularly well tolerated. It is also worthy of note that sodium thiosulfate was administered only during the chemotherapy perfusion in contrast to other investigations (Robbins et al., 1994a, b; Kerber et al., 1998). The catheter interventions also carried a very low risk. Observation times were too short to state de®nitive advantages regarding patient survival, and for subgroup analysis of the patient population according to tumour stages and the actual treatment. Small
numbers only allow preliminary conclusions. Further studies must include control groups. Nevertheless, it shows that all T1 and T2 tumours can be treated because of the signi®cantly reduced side-eects. Earlier studies all reported on treatment of more advanced cancer stages and hence the aim could only be palliative in most cases. Logically, general survival can only be assessed for improvement when smaller tumour stages are also included in which a curative approach is adopted. There are hints in prospective studies that this may be the correct conclusion (Volling and SchroÈder, 1995). A multi-modality therapy which is supposed to be superior to a single therapy can be carried out only when the individual modalities have reduced side-eects which are not cumulative. The reduction of morbidity may also have prognostic importance. At the present time there is no clear evidence from controlled studies that chemotherapy has any bene®t in long-term control of oral squamous cell carcinoma, and the high level of adverse side-eects makes its use hard to justify. Using this technique as described, it is possible to reduce the side-eects to allow further studies into the role of dierent chemotherapy regimens. References Bachaud JM, David JM, Boussin G, Daly N: Combined postoperative radiotherapy and weekly cisplatin infusion for locally advanced squamous cell carcinoma of the head and neck: preliminary report of a randomized trial. Int J Radiat Oncol Biol Phys 20: 243±246, 1991 Bachaud JM, David JM, Shubinski RE et al.: Predictive factors of a complete response to and adverse eects of a CDDP-5FU combination as primary therapy for head and neck squamous carcinomas. J Laryngol Otol 107: 924±930, 1993 Baker SR, Wheeler RH: Intraarterial chemotherapy for head and neck cancer. Part I. Theoretical considerations and drug delivery systems. Head Neck Surg 6: 664±682, 1983 Bitter K: Pharmacokinetic behaviour of bleomycin-cobalt-57 with special regard to intraarterial perfusion of the maxillo-facial region. J Maxillofac Surg 4: 226±231, 1976 Brady LW, Davis LW: Treatment of the head and neck by radiation therapy. Semin Oncol 15: 29±38, 1988 Eckardt A, Kelber A: Palliative, intraarterial chemotherapy for advanced head and neck cancer using an implantable port system. J Oral Maxillofac Surg 52: 1243±1246, 1994 Ervin TJ, Clark JR, Weichselbaum RR et al.: An analysis of induction and adjuvant chemotherapy in the multidisciplinary treatment of squamous-cell carcinoma of the head and neck. J Clin Oncol 5: 10±20, 1987 Forastiere AA, Baker SR, Wheeler R, Medvec BR: Intra-arterial cisplatin and FUDR in advanced malignancies con®ned to the head and neck. J Clin Oncol 5: 1601±1606, 1987 Goel R, Cleary SM, Horton C et al.: Eect of sodium thiosulfate on the pharmacokinetics and toxicity of cisplatin. J Natl Cancer Inst 81: 1552±1560, 1989 Harker GJ, Stephens FO: Comparison of intra-arterial versus intravenous 5-¯uorouracil administration on epidermal squamous cell carcinoma in sheep. Eur J Cancer 28: 1437±1441, 1992 Jaulerry C, Rodriguez J, Brunin F et al.: Induction chemotherapy in advanced head and neck tumors: results of two randomized trials. Int J Radiat Oncol Biol Phys 23: 483±489, 1992 Kerber CW, Wong WH, Howell SB, Hanchett K, Robbins KT: An organ-preserving selective arterial chemotherapy strategy for head and neck cancer. AJNR Am J Neuroradiol 19: 935±941, 1998
Intraarterial chemotherapy as neoadjuvant treatment of oral cancer 307 Lore JM, Bonilla JA, Spaulding M, Lee KY, Kaufman S, Sundquist N, Smeeding D: Preoperative adjuvant chemotherapy for advanced head and neck cancer: a surgical evaluation. J Surg Oncol Suppl. 1: 2±6, 1989 Martin M, Hazan A, Vergnes L et al.: Randomized study of 5¯uorouracil and cisplatin as neoadjuvant therapy in head and neck cancer: a preliminary report. Int J Radiat Oncol Biol Phys 19: 973±975, 1990 Miller AB, Hoogstraaten B, Staquet M, Winkler A: Reporting results of cancer treatment. Cancer 47: 207±214, 1981 Mortimer JE, Taylor ME, Schulman S, Cummings C, Weymuller E, Laramore G: Feasibility and ecacy of weekly intraarterial cisplatin in locally advanced (stage III and IV) head and neck cancers. J Clin Oncol 6: 969±975, 1988 Munro AJ: An overview of randomized controlled trails of adjuvant chemotherapy in head and neck cancer. Br J Cancer 71: 83±91, 1995 Platz H, Fries R, Hudec H: Retrospektive DOÈSAK Studie uÈber Karzinome der MundhoÈhle. TherapieabhaÈngiger Prognoseindex TPI. Dtsch Z Mund-Kiefer-GesichtsChir 7: 5±12, 1983 Poulsen M, Aldren C, Tripcony L, Walker Q: Is surgery necessary in stage III and stage IV cancer of the head and neck that responds to induction chemotherapy? Arch Otolaryngol Head Neck Surg 122: 467±471, 1996 Robbins KT, Storniolo AM, Kerber C, Seagren S, Berson A, Howell SB: Rapid superselective high-dose cisplatin infusion for advanced head and neck malignancies. Head Neck 14: 364±371, 1992 Robbins KT, Storniolo AM, Kerber C et al.: Phase I study of highly selective supradose cisplatin infusions for advanced head and neck cancer. J Clin Oncol 12: 2113±2120, 1994a Robbins KT, Vicario D, Seagren S et al.: A targeted supradose cisplatin chemoradiation protocol for advanced head and neck cancer. Am J Surg 168: 419±422, 1994b
Rooney M, Kish J, Jacobs J et al.: Improved complete response rate and survival in advanced head and neck cancer after threecourse induction therapy with 120-hour 5-FU infusion and cisplatin. Cancer 55: 1123±1128, 1985 Sileni VC, Fosser V, Maggian P et al.: Pharmacokinetics and tumour concentration of intraarterial and intravenous cisplatin in patients with head and neck squamous cancer. Cancer Chemother Pharmacol 30: 221±225, 1992 Sobin LH, Wittekind C: UICC: TNM classi®cation of malignant tumours. New York: John Wiley & Sons Inc., 1997 Stell P, Rawson NSB: Adjuvant chemotherapy in head and neck cancer. Br J Cancer 61: 779±787, 1990 Sullivan RD, Jones R, Schnabel TG, Shorey J: The treatment of human cancer with intraarterial nitrogen mustard utilising a simpli®ed catheter technique. Cancer 6: 121±134, 1953 Teicher BA, Holden SA, Kelley MJ et al.: Characterization of a human squamous carcinoma cell line resistant to cisdiamminedichloroplatinum(II). Cancer Res 47: 388±393, 1987 Volling P, SchroÈder M: VorlaÈu®ge Ergebnisse einer prospektiv randomisierten Studie zur primaÈren Chemotherapie bei MundhoÈhlen -und Pharynxkarzinomen. HNO 43: 58±64, 1995 Dr. med. Dr. med. dent. AdorjaÂn KovaÂcs Klinik und Poliklinik fuÈr Kiefer-und Plastische Gesichtschirurgie UniversitaÈtsklinikum Haus 21 Theodor-Stern-Kai 7 D-60590 Frankfurt am Main Germany Tel.: 069-63016610 Fax: 069-63015644 Paper received 23 June 1999 Accepted 16 September 1999
Intraarterial chemotherapy as neoadjuvant treatment of oral cancer AdorjaÂn F. KovaÂcs,1 Bernd Turowski,2 Mostafa T. Ghahremani,1 Matthias Loitz1 1
Department of Maxillofacial Plastic Surgery (Head: Prof. Dr Dr K. Bitter), Johann Wolfgang Goethe-University Medical School, Frankfurt am Main, Germany; 2Institute for Neuroradiology (Head: Prof. Dr F. E. Zanella), Johann Wolfgang Goethe-University Medical School, Frankfurt am Main, Germany SUMMARY. Neoadjuvant chemotherapy in patients with primary squamous cell carcinomas of the oral cavity should lead to high remission rates whilst having low morbidity. Ecacy can also be enhanced by treating small tumour stages. As part of a multi-modality therapy of all stages of primary oral cavity carcinoma, 103 patients were treated with neoadjuvant intraarterial (i.a.) chemotherapy. After regimen A with 100 mg/m2 i.a. cisplatin followed by 5 day continuous intravenous infusion of 5-¯uorouracil (1 g/m2 per day) in 36 patients, an i.a. high pressure chemo-perfusion with a dose of 150 mg/m2 cisplatin was used with simultaneous intravenous infusion of 9 g/m2 sodium thiosulfate (regimen B, 67 patients). Subsequent treatment comprised radical surgery and simultaneous radiochemotherapy with docetaxel. Partial and complete remissions were found in 80.6% (regimen A) and 67.2% (regimen B) of cases, tumour growth was inhibited in 11.1% and 31.3%. Very low toxicity could be shown especially in regimen B. 66.7% and 74.6% of patients could be operated on radically. Survival rate was 61.1% (regimen A, 22.7 months of mean observation time) and 79.1% (regimen B, 8.4 months). Patients with highgrade remissions seemed to have a survival advantage. Neoadjuvant i.a. chemotherapy with cisplatin, especially in its high dose variant, is a practical therapeutic tool for the treatment of all stages of primary oral cavity carcinoma. # 1999 European Association for Cranio-Maxillofacial Surgery
systems successfully prevented complications of intraarterial (i.a.) chemotherapy, such as infections, catheter misplacement, thromboses and haemorrhages. Robbins et al. (1992) started to combine these concepts by using a high dose chemoradiotherapy with cisplatin and a simultaneous application of sodium thiosulfate for systemic, competitive neutralisation of the cytostatic (Goel et al., 1989). Generally, the indication for this treatment was thought to be palliative. However, its eect on the survival rate of all patients with oral squamous cancer can only be investigated by using i.a. induction chemotherapy for smaller tumour stages in addition, as was done with systemic chemotherapy in a few earlier studies (Martin et al., 1990; Volling and SchroÈder, 1995). This paper presents the neoadjuvant i.a. chemotherapy with cisplatin as part of a multi-modality treatment concept of 103 patients with untreated primary squamous cell carcinomas of the oral cavity. All tumour stages were treated using this method. A local remission was intended with reduction both of tumour associated symptoms, and, possibly, aggressiveness of the tumour concerning cell spread. The therapy was planned to have as low systemic toxicity as possible for prevention of prognostically relevant morbidity (due to accumulation of side-eects). The ®rst regimen combined the i.a. application mode with a systemic chemotherapy while the second regimen used high dose cisplatin alone in combination with sodium thiosulfate for simultaneous systemic neutralisation.
INTRODUCTION The treatment of squamous cell carcimomas of the oral cavity is not satisfactory in terms of patient survival. This is a reason for the many dierent treatment plans commonly used consisting of varying combinations of surgery, radiation and chemotherapy. Systemic chemotherapy regimens with cisplatin and 5-¯uorouracil appear to be very ecient locally (Rooney et al., 1985) and in the subsumed prevention of metastatic spread (Jaulerry et al., 1992). But the advantage of induction chemotherapy for the patients is still considered controversial (Stell and Rawson, 1990; Munro, 1995), being associated with severe toxicity and morbidity potentially relevant to the survival rate. The concept of `down staging' of a tumour could not be con®rmed and radical surgery is still indispensible (Poulsen et al., 1996). Another problem of primary chemotherapy was the development of resistant cell lines which could be overcome by higher doses (Teicher et al., 1987). The reduction of peripheral toxicity combined with high local ecacy was a further aim (Baker and Wheeler, 1983), especially important in oral cancer patients who often suer from other intercurrent illnesses. For 50 years regional chemotherapy using the arterial route in the head and neck region, tried to ful®l these claims and various cytostatic agents and technical means were used. Most indications were palliative treatment of advanced cancer stages and recurrences (Sullivan et al., 1953; Bitter, 1976; Eckardt and Kelber, 1994). Development of ®ne catheter delivery 302
Intraarterial chemotherapy as neoadjuvant treatment of oral cancer 303
PATIENTS AND METHODS The study started in December 1996. Results of this study date from May 1999. Two unselected groups were created: 36 consecutive patients (77.8% male and 22.2% female, average age 57.3 years) received i.a. chemotherapy with 100 mg m2 cisplatin without systemic neutralisation followed by an intra-venous (i.v.) continuous infusion of 1 g/m2 5-¯uorouracil for functional synergism (regimen A). The second group of 67 consecutively treated patients (83.6% male and 16.4% female, average age 59.9 years) received i.a. high pressure chemotherapy with the high dose of 150 mg/m2 cisplatin over 5±15 min in combination with parallel i.v. application of 9 g/m2 sodium thiosulfate after a delay of 10 s (regimen B). To avoid resistance development in both groups, a maximum of 3 cycles were scheduled, if no grade III or IV remission occurred at the ®rst cycle. Remissions were de®ned: 0 ± tumour progression, I ± stable disease (no tumour growth), II ± reduction of tumour mass 5 50% (lowgrade partial remission), III ± reduction of tumour mass 4 50% (high-grade partial remission), IV ± complete remission (disappearance of the tumour). The assessment was done by inspection, palpation and CT examination after the ®rst cycle. Grades of remission which might have been higher after the second and third cycles were not considered in this study to make results comparable. No biopsies were taken after chemotherapy to avoid interference with the planned operation. Location of the primary tumours were typically in the ¯oor of the mouth (about 40%) and the tongue (about 20%). Performance status of the patients (Table 1) was good in a high percentage, although the majority of them presented advanced stages of disease (Table 1). Nevertheless, between 30% and 40% had T1 and T2 tumours. All tumours were histologically con®rmed squamous cell carcinomata; three quarters were moderately dierentiated. Similarity of the two patient groups was best shown using the therapy-dependent prognostic index TPI which combines data on tumour size, in®ltration, regional lymph nodes, distant metastases and age (Platz et al., 1983) (Fig. 1). After chemotherapy, the multi-modality treatment plan provided radical surgery of the mouth and neck, if operation was feasible depending on the general or local condition. A so-called `down staging' of the tumour was impossible. Surgery was followed by fractionated simultaneous radio-chemotherapy of the tumour area and the neck lymphatics over 5 weeks (51.3 Gy at 1.9 Gy per fraction) with systemic application of docetaxel (25 mg/m2) once a week. This last treatment modality was oered to all patients, with or without surgery. Pretherapeutic staging included the use of palpation, US, CT and MR imaging for examination of the neck lymph nodes and PET for diagnosis of second tumours, neck lymph nodes and distant metastases. On the morning of treatment, patients were given 74 mg
Table 1 ± Characteristics of the patient population according to ECOG (Eastern Cooperative Oncology Group) for performance status and UICC (Sobin and Wittekind, 1997) for tumour staging Grade
0 1 2 3 4
Performance Tumour (T) status (%) (%)
Nodes (N) (%)
Tumour stages (%)
A
B
A
B
A
B
A
B
61.1 22.2 11.1 0 5.6
70.1 15 13.4 1.5 0
/ 11.1 25 0 63.9
/ 17.9 30.8 1.5 19.8
38.9 16.6 38.9 5.6 /
47.8 22.4 26.9 2.9 /
/ 11.1 11.1 8.3 69.5
/ 10.5 23.9 11.9 53.7
dolasetron and 75 mg prednisolone i.v. Afterwards, 1.5 litres of a full electrolyte solution (with 20 mval potassium chloride) were given by i.v. infusion over 2 h. Then catheterisation of the right femoral artery was carried out using a size 4 (french) catheter containing a coaxial micro-catheter. After superselective visualisation of the tumour-feeding vessel using ¯uoroscopy and a contrast medium, either 100 mg/m2 cisplatin dissolved in 500 ml 0.9% saline solution was infused i.a. over 1 h (regimen A), or 150 mg/m2 cisplatin dissolved in the same amount of saline solution was infused with controlled pressure (2 ml/s) (regimen B). For analgesia, 0.1 to 0.3 mg fentanyl was delivered i.v. (and on occasions 5 to 15 mg mepivacain) into the perfused artery. With a delay of 10 s, an i.v. infusion of 9 g/m2 sodium thiosulfate was given in parallel. After the treatment, 1 litre of full electrolyte solution with 20 mval potassium chloride was infused i.v. over 5 h. The next day, the patients were hyperhydrated with 3 L of a two-third electrolyte solution, thrombosis prophylaxis with heparin S.C., also dolasetron i.v. if necessary. Routine laboratory checks were made on alternate days. Ward stay lasted between 4 and 6 days for most patients of regimen B. In regimen A, this time was prolonged by the 5 day continuous infusion of 5-¯uorouracil (1 g/m2). Daily application of allopurinol (300 mg) and anti-emetic drugs was mandatory. The side-eects of the cycles were noted. The sideeects of the second and third cycles were always less pronounced than of the ®rst cycle.
RESULTS There was a high overall response rate of 91.7% in regimen A and of 98.5% in regimen B (Fig. 2). Tumour progression was observed in only 8.3% and 1.5% of cases respectively. The synergy of cisplatin and 5-¯uorouracil had a higher impact on partial and complete remissions (55.6% and 25%) when compared with the high dose cisplatin therapy (64.2% and 3%). Therefore, the rate of stable disease was much higher in scheme B (31.3%). More than 75% of these patients had only a single cycle. Repeated courses were used for cases with very large tumours
304 Journal of Cranio-Maxillofacial Surgery
Fig. 1 ± Distribution of therapy-dependent prognosis index TPI for the examined patient population. White columns: regimen with i.a. cisplatin and i.v. 5-¯uorouracil. Black columns: regimen with i.a. cisplatin and i.v. sodium thiosulfate.
Fig. 2 ± Grades of remission after ®rst cycle of i.a. chemotherapy. White columns: regimen with i.a. cisplatin and i.v. 5-¯uorouracil. Black columns: regimen with i.a. cisplatin and i.v. sodium thiosulfate.
which were considered to be inoperable. There was no case of a change of remission grade after repetition of a cycle. Therefore, repeated i.a. chemotherapy provided local control. The planned operation could not be carried out in 29 cases (12 in the ®rst and 17 in the second group). Five patients died due to advanced malignancy despite temporary local control. During pre-therapeutic examination, the tumours of 9 patients were considered to be inoperable because of involvement of the skull base and/or vertebrae, or in®ltration of
the whole tongue and ¯oor of the mouth including the common carotid artery. Eight patients had very poor general condition with multiple systemic diseases. One patient suered from alcoholic myocardial insuciency, one had inoperable bronchial malignancy and three had pulmonary metastatic spread. For each patient, the possibility of a simultaneous radio-chemotherapy was considered. If no further therapy could be oered, the i.a. chemotherapy was de®ned as palliative. Acceptance of the i.a. chemotherapy was high. This could be explained by the
Intraarterial chemotherapy as neoadjuvant treatment of oral cancer 305
After an observation period for the group with regimen A of 17.5+8.1 months on average (maximum 30 months), 14 patients were dead (after 9.4+6.7 months) and 22 (61.1%) alive (after (22.7+ 2.9 months). For regimen B, observation time was 7.8+5.2 months (maximum 20 months). Fourteen patients were dead (after 5.4+2.8 months) and 53 (79.1%) alive (after 8.4+5.5 months). Looking at the relationship of grades of remission and survival (Table 3), there was a tendency for more patients with low-grade remissions to die prematurely while more patients with high-grade remissions were still alive.
fact that 26.9% of all patients of regimen B had no clinically detectable acute side-eects (compared with 14.7% in regimen A). Therefore, only 2 patients refused the recommended continuation of the therapy. Surgical resection was undertaken within 3 weeks of the last dose of chemotherapy. Of the patients in regimen A, 66.7% could be operated on (resection with histologically con®rmed healthy margins). In regimen B, the rate was 74.6%. Compared to the stage-grouping (Table 1), this meant that many patients with advanced tumours were operated on. Twelve suprahyoid lymph node dissections were carried out in the group with regimen A and 40 in the group with regimen B, while the rate of functional neck dissections was lower (11 times/18 times). There were no unusual postoperative complications except a higher rate of tracheotomies in the group with regimen B (24 cases 48%) when compared to regimen A (3 cases 12.5%), which may be due to enhanced swelling of the area perfused with high-dose cisplatin. The side-eects of both i.a. chemotherapy regimens are presented in Table 2. They are divided into gastrointestinal side-eects (nausea/vomiting/diarrhoea), haematological side-eects (anaemia/leukopaenia/thrombopaenia), disturbances of electrolytes (potassium, ferritin), renal and hepatic side-eects (according to WHO: Miller et al., 1981). The regimen B was clearly superior in nearly every case. Only the serum potassium showed a more frequent decrease and the renal side-eects were equal in spite of the much larger dose in regimen B. Creatinine never exceeded a serum value of 1.9 and showed normalisation of values after 5 days on average. Only in regimen A was it necessary to interrupt therapy due to hyperemesis and it had to be supplemented by erythrocyte transfusions or colony stimulating factor (CSF) (twice). Other side-eects were partial temporary alopecia (15% in regimen A and 19% in regimen B) and high tone hearing loss (assessed subjectively in 10% in regimen A and 15% in regimen B). One patient with alcoholic cardiomyopathy developed cardiac decompensation after i.a. chemotherapy. In 212 catheter interventions, only one side-eect was noted (apoplexy possibly due to an arteriosclerotic plaque or thrombus which could have been dislodged by the micro-catheter).
DISCUSSION Chemotherapy cannot be the sole therapy for oral cancer (despite initially high remission rates) if curative treatment is planned. Radical surgery is indispensable (Poulsen et al., 1996) and is not compromised by preceding systemic chemotherapy (Lore et al., 1989). These assertions are valid for i.a. chemotherapy, too, which has enjoyed a renaissance recently. Its advantage is a higher possible local dose of the cytostatic agent compared to the systemic route if a competitive peripheral neutralisation is used (Robbins et al., 1992; Robbins et al., 1994a). Animal experiments and clinical investigation show a better eect from the i.a. route compared to the i.v. route (Bitter, 1976; Harker and Stephens, 1992). The results con®rm the high initial local eectiveness of i.a. cisplatin (Mortimer et al., 1988), especially when in high dose (Robbins et al., 1994a, b). The comparison of two unselected groups regarding staging and number of non-operable patients seemed possible (as TPI and the percentage of patients actually operated on showed) but the main task of the present study was ecacy and practicability of i.a. chemotherapy. Overall, response was better in regimen B. To date, the higher complete remission grades in the regimen A do not seem to be decisive, because they have to be seen in the context of a multimodality therapy and are countered by much greater side-eects when compared with regimen B (Table 2), although less than conventional systemic chemotherapy (Bachaud et al., 1993). It has to be stressed that, in this study, remissions were assessed clinically and not histologically. Nevertheless, the results of this study demonstrate a connection between high-grade
Table 2 ± Side-eects of intraarterial chemotherapy. Group A: regimen with i.a. cisplatin and i.v. 5-¯uorouracil. Group B: regimen with i.a. cisplatin and i.v. sodium thiosulfate. Grades according to WHO (Miller et al., 1981) Grade Nausea (%)
A I II III IV
B
Diarrhaea (%)
Anaemia (%)
Leukopaenia (%)
Thrombocytopaenia (%)
Sideroemia (%)
Hypokalaemia (%)
Serum creatinine (%)
Hepatic enzymes (GOT, GPT, GGT) (%)
A
B
A
B
A
B
A
B
A
B
A
B
A
A
1.5 0 0 0
13.9 8.3 0 0
4.5 13.9 1.5 8.3 1.5 2.8 0 0
4.5 0 0 0
8.3 5.6 0 0
0 1.5 1.5 0
22.2 0 0 0
17.9 0 0 0
16.7 0 0 0
22.4 0 0 0
33.3 34.3 13.9 0 1.5 0 0 0 0 0 0 0
38.9 17.9 2.9 11.1 11.9 0 5.6 0 0 0 0 0
B
B 10.5 0 0 0
306 Journal of Cranio-Maxillofacial Surgery Table 3 ± Relation of grades of remission and survival. Observation times are mentioned in the text Grade of remission
Regimen A Dead
Living
Dead
Living
0 1 2 3 4
3 2 7 1 1
0 2 6 7 7
1 (1.5%) 6 (8.9%) 7 (10.4%) 0 0
1 15 20 13 4
(8.3%) (5.6%) (19.4%) (2.8%) (2.8%)
Regimen B
(5.6%) (16.7%) (19.4%) (19.4%)
(1.5%) (22.4%) (29.9%) (19.4%) (6%)
remissions and survival similar to those reported by Ervin et al. (1987) who stated a correlation of response to systemic induction chemotherapy with failure-free survival. Dierent locations of primary tumours, dierent mixtures of clinical stages of the treated populations and dierent drugs and administration routes are the reasons for diculties in comparison. The fact that remissions mainly occurred after the ®rst cycle could allow the early distinction of responders from non-responders. The assessment of remission, prognostically, could be another future task for i.a. chemotherapy, as well as a starting point for dierential therapy. The actual aim of i.a. chemotherapy is not necessarily complete remission, but to restrain the tumour in its local and possibly metastatic aggressiveness before radical surgery without the very high side-eects of systemic chemotherapy. Therefore, stabilization of disease was also de®ned as a response. The eradication of known micro-metastases is still a problem. Although other side-eects show that free cisplatin reaches the peripheral organs and it is known that i.a. chemotherapy without a neutralising agent can reach the same peripheral cytostatic levels (Sileni et al., 1992), doubts exist about i.a. chemotherapy in this regard (Forastiere et al., 1987). Therefore, subsequent treatment with this multimodality therapy consists of a simultaneous radiochemotherapy with weekly systemic application of docetaxel. Postoperative radio-chemotherapy was reported to show improvement of survival (Bachaud et al., 1991). Nevertheless, locoregional recurrence during the ®rst 2 years after treatment remains the main problem in oral cancer patients (Brady and Davis, 1988). The present study focuses on the feasibility and acceptance of i.a. chemotherapy. The data requirements regarding side-eects compared well with other studies (Eckardt and Kelber, 1994; Robbins et al., 1994a; Kerber et al., 1998) and serve to demonstrate that i.a. chemotherapy, especially the high dose variant, was particularly well tolerated. It is also worthy of note that sodium thiosulfate was administered only during the chemotherapy perfusion in contrast to other investigations (Robbins et al., 1994a, b; Kerber et al., 1998). The catheter interventions also carried a very low risk. Observation times were too short to state de®nitive advantages regarding patient survival, and for subgroup analysis of the patient population according to tumour stages and the actual treatment. Small
numbers only allow preliminary conclusions. Further studies must include control groups. Nevertheless, it shows that all T1 and T2 tumours can be treated because of the signi®cantly reduced side-eects. Earlier studies all reported on treatment of more advanced cancer stages and hence the aim could only be palliative in most cases. Logically, general survival can only be assessed for improvement when smaller tumour stages are also included in which a curative approach is adopted. There are hints in prospective studies that this may be the correct conclusion (Volling and SchroÈder, 1995). A multi-modality therapy which is supposed to be superior to a single therapy can be carried out only when the individual modalities have reduced side-eects which are not cumulative. The reduction of morbidity may also have prognostic importance. At the present time there is no clear evidence from controlled studies that chemotherapy has any bene®t in long-term control of oral squamous cell carcinoma, and the high level of adverse side-eects makes its use hard to justify. Using this technique as described, it is possible to reduce the side-eects to allow further studies into the role of dierent chemotherapy regimens. References Bachaud JM, David JM, Boussin G, Daly N: Combined postoperative radiotherapy and weekly cisplatin infusion for locally advanced squamous cell carcinoma of the head and neck: preliminary report of a randomized trial. Int J Radiat Oncol Biol Phys 20: 243±246, 1991 Bachaud JM, David JM, Shubinski RE et al.: Predictive factors of a complete response to and adverse eects of a CDDP-5FU combination as primary therapy for head and neck squamous carcinomas. J Laryngol Otol 107: 924±930, 1993 Baker SR, Wheeler RH: Intraarterial chemotherapy for head and neck cancer. Part I. Theoretical considerations and drug delivery systems. Head Neck Surg 6: 664±682, 1983 Bitter K: Pharmacokinetic behaviour of bleomycin-cobalt-57 with special regard to intraarterial perfusion of the maxillo-facial region. J Maxillofac Surg 4: 226±231, 1976 Brady LW, Davis LW: Treatment of the head and neck by radiation therapy. Semin Oncol 15: 29±38, 1988 Eckardt A, Kelber A: Palliative, intraarterial chemotherapy for advanced head and neck cancer using an implantable port system. J Oral Maxillofac Surg 52: 1243±1246, 1994 Ervin TJ, Clark JR, Weichselbaum RR et al.: An analysis of induction and adjuvant chemotherapy in the multidisciplinary treatment of squamous-cell carcinoma of the head and neck. J Clin Oncol 5: 10±20, 1987 Forastiere AA, Baker SR, Wheeler R, Medvec BR: Intra-arterial cisplatin and FUDR in advanced malignancies con®ned to the head and neck. J Clin Oncol 5: 1601±1606, 1987 Goel R, Cleary SM, Horton C et al.: Eect of sodium thiosulfate on the pharmacokinetics and toxicity of cisplatin. J Natl Cancer Inst 81: 1552±1560, 1989 Harker GJ, Stephens FO: Comparison of intra-arterial versus intravenous 5-¯uorouracil administration on epidermal squamous cell carcinoma in sheep. Eur J Cancer 28: 1437±1441, 1992 Jaulerry C, Rodriguez J, Brunin F et al.: Induction chemotherapy in advanced head and neck tumors: results of two randomized trials. Int J Radiat Oncol Biol Phys 23: 483±489, 1992 Kerber CW, Wong WH, Howell SB, Hanchett K, Robbins KT: An organ-preserving selective arterial chemotherapy strategy for head and neck cancer. AJNR Am J Neuroradiol 19: 935±941, 1998
Intraarterial chemotherapy as neoadjuvant treatment of oral cancer 307 Lore JM, Bonilla JA, Spaulding M, Lee KY, Kaufman S, Sundquist N, Smeeding D: Preoperative adjuvant chemotherapy for advanced head and neck cancer: a surgical evaluation. J Surg Oncol Suppl. 1: 2±6, 1989 Martin M, Hazan A, Vergnes L et al.: Randomized study of 5¯uorouracil and cisplatin as neoadjuvant therapy in head and neck cancer: a preliminary report. Int J Radiat Oncol Biol Phys 19: 973±975, 1990 Miller AB, Hoogstraaten B, Staquet M, Winkler A: Reporting results of cancer treatment. Cancer 47: 207±214, 1981 Mortimer JE, Taylor ME, Schulman S, Cummings C, Weymuller E, Laramore G: Feasibility and ecacy of weekly intraarterial cisplatin in locally advanced (stage III and IV) head and neck cancers. J Clin Oncol 6: 969±975, 1988 Munro AJ: An overview of randomized controlled trails of adjuvant chemotherapy in head and neck cancer. Br J Cancer 71: 83±91, 1995 Platz H, Fries R, Hudec H: Retrospektive DOÈSAK Studie uÈber Karzinome der MundhoÈhle. TherapieabhaÈngiger Prognoseindex TPI. Dtsch Z Mund-Kiefer-GesichtsChir 7: 5±12, 1983 Poulsen M, Aldren C, Tripcony L, Walker Q: Is surgery necessary in stage III and stage IV cancer of the head and neck that responds to induction chemotherapy? Arch Otolaryngol Head Neck Surg 122: 467±471, 1996 Robbins KT, Storniolo AM, Kerber C, Seagren S, Berson A, Howell SB: Rapid superselective high-dose cisplatin infusion for advanced head and neck malignancies. Head Neck 14: 364±371, 1992 Robbins KT, Storniolo AM, Kerber C et al.: Phase I study of highly selective supradose cisplatin infusions for advanced head and neck cancer. J Clin Oncol 12: 2113±2120, 1994a Robbins KT, Vicario D, Seagren S et al.: A targeted supradose cisplatin chemoradiation protocol for advanced head and neck cancer. Am J Surg 168: 419±422, 1994b
Rooney M, Kish J, Jacobs J et al.: Improved complete response rate and survival in advanced head and neck cancer after threecourse induction therapy with 120-hour 5-FU infusion and cisplatin. Cancer 55: 1123±1128, 1985 Sileni VC, Fosser V, Maggian P et al.: Pharmacokinetics and tumour concentration of intraarterial and intravenous cisplatin in patients with head and neck squamous cancer. Cancer Chemother Pharmacol 30: 221±225, 1992 Sobin LH, Wittekind C: UICC: TNM classi®cation of malignant tumours. New York: John Wiley & Sons Inc., 1997 Stell P, Rawson NSB: Adjuvant chemotherapy in head and neck cancer. Br J Cancer 61: 779±787, 1990 Sullivan RD, Jones R, Schnabel TG, Shorey J: The treatment of human cancer with intraarterial nitrogen mustard utilising a simpli®ed catheter technique. Cancer 6: 121±134, 1953 Teicher BA, Holden SA, Kelley MJ et al.: Characterization of a human squamous carcinoma cell line resistant to cisdiamminedichloroplatinum(II). Cancer Res 47: 388±393, 1987 Volling P, SchroÈder M: VorlaÈu®ge Ergebnisse einer prospektiv randomisierten Studie zur primaÈren Chemotherapie bei MundhoÈhlen -und Pharynxkarzinomen. HNO 43: 58±64, 1995 Dr. med. Dr. med. dent. AdorjaÂn KovaÂcs Klinik und Poliklinik fuÈr Kiefer-und Plastische Gesichtschirurgie UniversitaÈtsklinikum Haus 21 Theodor-Stern-Kai 7 D-60590 Frankfurt am Main Germany Tel.: 069-63016610 Fax: 069-63015644 Paper received 23 June 1999 Accepted 16 September 1999