230 with oesophago-gastric cancer. The study was designed to detect an improvement in response rates from 35% to 50% with 170 patients per arm providing at least 80% power (a error = 5%). The number of patients with colorectal cancer randomised was increased to 100 per arm to ensure adequate power as a separate study. We observed a significant improvement in response rate from 38% with protracted venous infusion (PVI) 5-fluorouracil (5-FU) alone to 54% with PVI 5-FU plus mitomycin-C (P = 0.024). Due to haemolytic uraemic syndrome associated with a cumulative mitomycin-C dose of 40 mg/m2 (10 mg/m2 per course) the cumulative dose of mitomycin-C was reduced to 28 mg/m2 (7 mg/ m2 per course). The 53 patients treated with the higher dose of mitomycin-C had a response rate of 61% compared to 45% for the 45 patients treated at the lower dose. This difference was not statistically significant. Further subgroup analysis was not performed as the study was not designed to have enough power to be analysed in such a way. After the mitomycin-C dose was reduced, careful surveillance for evidence of haemolysis was instituted and no further mitomycin-C given if haemolysis was observed. Although we reduced the dose of mitomycin-C, the higher dose of mitomycin-C may still be safely used if increased monitoring is instituted, but this needs to be investigated further. Mitomycin-C in combination with PVI 5-FU significantly improves response and failure-free survival but the optimal dose of mitomycin-C with respect to efficacy In their paper published in Annals of Oncology [1], Ross et al. and toxicity requires clarification. describe an advantage of the combination of mitomycin-C and protracted venous infusion (PVI) of 5-fluorouracil as comP. Ross, A. Norman & D. Cunningham pared to PVI-5-fluorouracil alone in the treatment of advanced Department of Medicine, The Royal Marsden Hospital, colorectal cancer. From the data presented, their conclusion is Surrey, SM2 5PT, UK Sutton, inadequate for the following reasons: Because of mitomycin-C associated toxicities, the dose of mitomycin-C has been reduced from 10 mg/m2 to 7 mg/m2 in June 1995. Although I could not find the exact number of patients treated with the lower dose or the higher dose, it appears that about half of the patients in this treatment arm received the lower dose. Response rate in patients receiving toe nails higher dose of mitomycin-C was higher (61%) than the response We report two cases of onycholysis of the finger and 2 rate of patients receiving lower dose (55%). Due to the toxic- following the application of docetaxel (100 mg/m ) given as ities seen, authors recommend treatment with the lower dose single agent chemotherapy for metastatic breast cancer. A 69-year-old female patient underwent lumpectomy and of mitomycin-C. However, their conclusion regarding the effect axillary lymph node dissection in October 1996. Histologic of treatment includes data from patients having received higher dose. Therefore, it can not be concluded that 5-FU and mito- examination revealed breast cancer (ductal, NOS type, pTlc, G2) with two out of 14 lymph nodes positive. The patient mycin-C at lower dose is superior to 5-FU alone. received three cycles of a CMF therapy regimen postoperatively. In March 1997 the CMF therapy was interrupted, beR. Herrmann cause liver ultrasound and X-ray examination revealed multiple Division of Oncology, Kantonsspital, 4031 Basel, Switzer-metastases in the liver, spine, and osseous pelvis. Instead, a high land dose adriamycin (100 mg/m2)-cyclophosphamide (600 mg/m2) combination chemotherapy was given over two cycles. Due to severe leukopenia, the chemotherapy regimen was changed Reference again and docetaxel (100 mg/m2) was given in April 1997. A dose of 171 mg was administered over a one-hour infusion for 1. Ross P, Norman A, Cunningham D et al. A prospective rando- three cycles every three weeks. mised trial of protracted venous infusion 5-fluorouracil with or A 73-year-old female patient received neoadjuvant CMF without mitomycin C in advanced colorectal cancer. Ann Oncol chemotherapy for inflammatory breast cancer. After three 1997; 8 (10): 995-1001. cycles of treatment, radical mastectomy with axillary node dissection was performed in December 1996. Histologic examination revealed multicentric breast cancer (ductal, pT2, This letter was referred to the authors who G2) with seven out of 16 lymph nodes positive. The patient received three cycles of a combination chemotherapy, containrespond as follows: ing 5-fluouracil (1000 mg), cyclophosphamide (800 mg), and mitozantrone (10 mg/m2), at four-week intervals. During this The study which we reported in the October 1997 edition of Annals of Oncology was designed as part of a stratified study treatment the patient developed multiple metastases in the liver 2. Lambert HE, Rustin GJS, Gregory WM, Nelstrop AE. A randomized trial offiveversus eight courses of cisplatin or carboplatin in advanced epithelial ovarian cancer: A North Thames Ovary Group Study. Ann Oncol 1997; 8: 327-33. 3. Colombo N, Maggioni A, Coleoni R et al. Surgery for advanced ovarian cancer. Proceedings of Perugia international cancer conference II on recent advances in treatment of testicular and ovarian cancer 1989. 4. Bruzzone M, Repetto L, Chiara S et al. Chemotherapy versus radiotherapy in the management of ovarian cancer patients with pathological complete response or minimal residual disease at second look. Gynecol Oncol 1990; 38: 392-5. 5. Krafft W, Glaser FH,Wenzel U et al. Moglichkeiten zur Erhaltung der kompletten Remission beim zytoststich behandelten Ovarialkarzinom. In Meerpohl HG, Pfleiderer A, Profous CZ (eds): Das Ovarialkarzinom. Berlin/Heidelberg/New York: Springer-Verlag 1993: 137^0. 6. Advanced Ovarian Trialists Group. Chemotherapy in advanced ovarian cancer: An overview of randomised clinical trials. BMJ 1991; 303: 884-93. Onycholysis in patients treated with docetaxel Downloaded from http://annonc.oxfordjournals.org/ by guest on September 9, 2014 5-Fluorouracil and mitomycin-C in colorectal cancer: Unacceptable conclusion 231 A. Obermair, M. Binder, M. Barrada, D. BancherTodesca, E. Asseryanis & E. Kubista University Hospital Medical School Vienna, A-1090 Vienna, Austria References 1. Curran CF. Onycholysis in doxorubicin-treated patients. Arch Dermatol 1990; 126: 1244 (Letter). 2. Creamer JD, Mortimer PS, Powles TJ. Mitozantrone-induced onycholysis. A series offivecases. Clin Exp Dermatol 1995; 20:459-61. 3. Obermair A, Vavra N, Kurz C et al. Onycholysis of the finger and toenails following the application of high-dose oral etoposide (1250 mg/m2) given as 200- and 150-mg single doses from days 1-10 every three weeks. Gynecol Oncol 1995; 57: 436 (Letter). 4. Trudeau ME. Docetaxol (Taxotere): An overview of first-line monotherapy. Semin Oncol 1995; 22: 17-21. 5. Slee PHTJ. Nail changes after chemotherapy. N Engl J Med 1997; 337:168 (Case report). Quality of life assessment in patients receiving adjuvant therapy for breast cancer: The IBCSG approach Figure 1 of the recently published paper [1] is labeled incorrectly. The labeling of the lines is reversed. The information in the text, however, is correct: quality of life scores at time points when patients were assigned to receive CMF therapy were systematically lower compared to timepoints without assigned chemotherapy, although the difference diminished over time. We apologize for this error. The figure will be corrected on all reprints which are available by the undersigned. The corrected figure is printed herewith. J. Bernhard IBCSG Coordinating Center, 3008 Bern, Switzerland 85 O 80 CIS! i < 0. i P 75 70 0 A,B C,D ^ / A,C / 9r / 65 i 60 0'B • CMF O No CMF 55 50 , —1 1-—1 1 1 1 —1 1 1 H K 1 1 1 1 12 15 0.0005 731 0.0003 726 Month p N 0.16 797 0.0001 683 0.002 708 Figure 1. Median PACIS scores in trial VII (postmenopausal nodepositive patients) at time points where patients received chemotherapy. Groups are combined according to whether or not the patients were receiving chemotherapy at a given time point. All patients also received Tamoxifen. Higher scores indicate better adjustment, /"-values are from ANOVA's at each time point using 100 minus square root of PACIS, controlling for culture and excluding patients who recurred within the first 18.5 months. A, B, C, D refer to treatment groups (see text for description). Reference 1. Bernhard J, Hiirny C, Coates AS et al. for the International Breast Cancer Study Group (IBCSG). Quality of life assessment in patients receiving adjuvant therapy for breast cancer: The IBCSG approach. Ann Oncol 1997; 8: 825-35. Downloaded from http://annonc.oxfordjournals.org/ by guest on September 9, 2014 and in the skin of the left arm. Therefore, in May 1997 she was given docetaxel as first line chemotherapy for metastatic breast cancer. A dose of 169 mg was administered over a one-hour infusion for two cycles every three weeks. Both patients developed burning pain, hyperaemia, and appreciable epidermolysis between the fingers and the toes after two cycles of treatment. While the first patient continued docetaxel treatment, the second discontinued the therapy. After three cycles of docetaxel the first patient presented with multiple and bilateral onycholysis of the finger and toe nails with brown discoloration as a sign of bleeding beneath the nails. The second patient presented with so-called Beau-Reil lines, indicating that nail growth had stopped due to chemotherapy. Additionally, beginning onycholysis was observed at the basal layer of multiple finger and toe nails. In both patients a dermatological examination showed no evidence of infectious origin of the alterations. To date, only a few case reports and letters have been published on the developent of onycholysis associated with anthracyclines [1], mitozantrone [2] and etoposide [3]. After using docetaxel as first-line chemotherapy in the treatment of metastatic breast cancer, nail disorders, including ridging, pain, and onycholysis were reported to occur in 35% of the patients but rarely to be severe [4]. Recently, Slee reported a case of nail changes following seven courses of docetaxel at three-week intervals [5]. Although both of our patients had undergone previous treatment with anthracyclines and mitozantrone, the clinical course clearly showed that the nail changes were related to docetaxel. We therefore suggest that investigators should pay special attention to this uncommon side effect, in particular to the frequency of docetaxel-related onycholysis.