portion of the circumference of the adventitia furthest from the catheter in their trial. Fourth, we state that our findings indicate that it may be necessary to deliver therapeutic doses of radiation 3– 4 mm from the luminal border because proliferating cells are found there. Parikh and Nori state that another study that has examined more than 21 vascular brachytherapy trials found that the dose delivered at 0.75 mm from the lumen most closely correlates with success. Although the details of this failure analysis are not known, they may rely on a shaky assumption. There must have been some assumption about the position of the brachytherapy source relative to the vessel wall. Intravascular ultrasound (IVUS) may have been used for this purpose. As inaccurate as this is, it is as good as one can get with current technology. As a result, they point out that these failure analyses based on dosimetric parameters may best be thought to rely on surrogates for dose to a “true target layer” rather than the real dose. We are simply unable to tell what the true dose is at any point in the target vessel using existing devices. Regardless of how dose is described, it is clear from the various failure analyses that success is related to the dose received to some portion of the vessel as indicated by proxy from the “dose surrogates.” The SCRIPPS failure analysis we cite in our publication seemed more descrip- tive of the dosimetric methods they used, prompting us to use it in our discussion (5). Without more details, it is not possible to reconcile these two conflicting failure analyses. Fifth, Parikh and Nori highlight the problems with IVUS interpretation and caution against its use with vascular brachytherapy. Currently, in interventional cardiology, IVUS is not widely used in the average practice. This is typical of relatively new technologies. Prior to widespread use, physicians must become familiar with its operation and applicability. Arguably, this is a great opportunity for radiation oncologists to bring in their skills with image interpretation and dosimetry at a relatively early stage on the IVUS learning curve. The non-IVUS-dependent guidelines formulated by the AAPM are a good start, but this field is far from mature so amendments are sure to be on the horizon. The inclusion of image guidance in these amendments will depend on its value to the field. Clinical trials will be necessary to clarify the picture. JAY P. CIEZKI, M.D. URS O. HÄFELI,PH.D. Department of Radiation Oncology The Cleveland Clinic Foundation 9500 Euclid Avenue Cleveland, OH 44195 PII S0360-3016(00)00413-2 1. Glagov S, Weisenberg E, Zarins C, Stankunavicius R, Kolettis G. Compensatory enlargement of human atherosclerotic coronary arteries. N Engl J Med 1987;316:1371–1375. 2. Guyton J, Hartley C. Flow restriction of one carotid artery in juvenile rats inhibits growth of arterial diameter. Am J Physiol 1985;248:H540 – H546. 3. Mintz G, Kent K, Pichard A, Satler L, Popma J, Leon M. Contribution of inadequate arterial remodeling to the development of focal coronary artery stenoses: An intravascular ultrasound study. Circulation 1997; 95:1791–1798. 4. Wiedemann J, Leavy J, Amols H, et al. Intracoronary irradiation acutely impairs endothelial and smooth muscle cell function as assessed by intravascular ultrasound. Circulation 1992;86:I–188. 5. Teirstein P, Massullo V, Jani S, et al. A subgroup analysis of the Scripps coronary radiation to inhibit proliferation poststenting trial. Int J Radiat Oncol Biol Phys 1998;42:1097–1104. FIVE FRACTIONS OF PREOPERATIVE RADIOTHERAPY FOR RECTAL CARCINOMA To the Editor: We studied with special interest and attention the paper by Myerson RJ et al. published this year in the Journal (1). We address this letter to you because we recently started (quite a few years later, of course) a similar scheme of preoperative radiotherapy (RT), because the impres- sion is diffuse that patients with rectal carcinoma need additional adjuvant or neoadjuvant treatment apart from basic radical surgery with total me- sorectal excision. Administration of a short course of preoperative RT— five fractions of 5 Gy each—was recently evaluated adequately (2–5). We follow a similar approach—that is, administration of a 5-day short course of preoperative RT (total dosage 25 Gy) followed by surgery within 7 days from completion of treatment. RT is delivered through four portals, with the patient lying in a prone position. A CT scan is used to define treatment volume. Superior border of the fields is at the level of the sacral promon- tory. The localization and extension of the primary growth determine the inferior border (always 3–5 cm below the lower edge of the tumor). The posterior border lies 1 cm behind the sacrum; the anterior border lies 3 cm in front of the sacral promontory and encompasses the posterior two-thirds of the femoral heads. The lateral margins lie 1 cm lateral to the pelvic sidewalls. It is obvious that by using the LQ isoeffect model, 25 Gy in five daily fractions corresponds to a 21  2-Gy scheme for both tumor and late-reacting tissues. Nevertheless, major complications are not expected if meticulous treatment planning is followed. Because we started our study in January 1999, we are unable to evaluate results regarding the first branch of the published study, which is “long- term tumor control.” On the other hand, because it is nearly 10 months since the beginning of the method, we feel we can judge the results of the second branch of the study, which is “tolerance to treatment.” Until now, we have treated 26 patients with operable rectal carcinoma. Sixteen of them were men, and 10 women. Their age ranged between 54 and 72 years (mean 63 years). Twenty-four of them have had low anterior resection of the rectum; the remaining two, abdominoperineal excision of the rectum. Preoperative RT was well tolerated regarding acute toxicity. There were no technical problems as a result of the administered RT, whereas macro- scopic radiation colitis was detected in 18 patients. Patients were evaluated postoperatively by the following. CLINICAL EXAMINATION Radiation dermatitis occurred in 18 of 26 patients, atypical abdominal pain in eight of 26, watery stools in six of 26, and bloody stool in four of 26. There were no major postoperative complications such as wound infection or dehiscence, anastomotic breakdown, bowel obstruction, or persistent colonic bleeding. There were no postoperative deaths. Compli- cations therefore ranged between Grade 1 and 2, and in any case less than Grade 3. COLONOSCOPY All patients had colonoscopy performed 3– 6 weeks postoperatively. Prominent findings were intense edema, hyperemia, and coverage of the anastomosis area with newly formed tissues. PATHOLOGIC EXAMINATION OF RESECTED SPECIMEN We found mild to moderate atrophy of the mucosa. Swollen epithelial cells with eosinophilic cytoplasm; large, dense, colored nuclei; and an increase in nucleokinesias were noted, as well as fibrosis with dilated thick-walled vessels in the submucosal layer. Moderate lymphoplasmato- cytic infiltration of the chorion with the presence of abundant eosinophils occurred. In conclusion, we can claim that use of a short course of preoperative RT in patients with operable tumors of the rectum did not significantly influ- ence the postoperative morbidity or mortality of these patients. The study is ongoing. K. PISTEVOU-GOBAKI G. A. PLATANIOTIS L. S. PAPADOPOULOS Department of Radiotherapy AHEPA Hospital Aristoteles University of Thessaloniki 1, S. Kyriakidi, 546 36 Thessaloniki Greece V. PENOPOULOS P. DELIKARIS Second Department of Surgery G. Papanikolaou National General Hospital Thessaloniki Greece 263 Correspondence