Radiation-induced stricture in the rectum. The lumen is almost completely obliterated.
A, Radiation enteritis. Diffuse thickening of the small bowel with multiple adhesions to the anterior abdominal wall and between adjacent loops of the small bowel. B, Focal scarring and serosal ecchymosis following pelvic radiotherapy; there is edematous thickening adjacent to the small bowel.
Radiation proctitis. Edematous, friable mucosa with profuse bleeding from telangiectatic blood vessels.
Comparison of radiation complications as a function of the originally treated tumor.
Turina M, Mulhall AM, Mahid SS, Yashar C, Galandiuk S. Frequency and Surgical Management of Chronic Complications Related to Pelvic Radiation. Arch Surg. 2008;143(1):46-52. doi:10.1001/archsurg.2007.7
Copyright 2008 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2008
Refractory complications from pelvic radiotherapy often require surgical treatment. Their management may be dictated by the primary tumor, radiation dose, and type and combination of radiation injuries, and may require transient diversion in most cases to guarantee good outcomes.
Retrospective 10-year cohort analysis compared with statewide epidemiologic data.
During a 10-year period, 14 791 patients in Kentucky were treated with pelvic radiotherapy. Forty-eight were referred to a university colorectal surgical unit for evaluation of refractory radiotherapy complications that had failed conservative medical management.
Main Outcome Measures
Epidemiologic statewide data were compared with hospital data regarding the treatment and outcome of patients with refractory pelvic radiotherapy complications.
Twenty-five patients had received radiotherapy for colorectal carcinoma, 10 for prostate cancer, 7 for carcinoma of the cervix, and 6 for other tumors. Patients presented with 1 or more complications, including radiation enteritis (60%), strictures (53%), fistulae (17%), nonhealing wounds (15%), and de novo cancers in radiated fields (10%). Low anastomotic strictures (10%) were initially treated by dilation under sedation. Six patients (12%) ultimately required permanent diversion. All radiation-induced fistulae required an operation.
Determining the proper treatment requires careful judgment and assessment of the degree and type of injury, patient anatomy, and sphincter function. Patients presenting with colorectal anastomotic and primary bowel strictures as their main complication had the best results, while most patients with severe radiation enteritis and very distal strictures required permanent diversion.
Radiotherapy plays an important role in the treatment of pelvic malignancies and has significantly improved patient prognosis. The potential of radiation to cause injury to normal tissue is well understood. In fact, the first description of enteritis following abdominal exposure to radiation, by Walsh,1 dates back to 1897, only 2 years after the landmark discovery of x-rays. Technical improvements in linear accelerators, such as the development of 3-dimensional conformal radiation therapy,2 intensity-modulated radiotherapy,3,4 novel fractionation schemes,5 and the increasing use of radioprotective agents, have helped reduce the frequency and severity of radiotherapy-related complications in recent years.6 Despite improvements in care, radiation injuries still account for substantial late morbidity in an estimated 20% of patients treated with pelvic radiotherapy.7 Mortality directly related to pelvic or abdominal radiotherapy is difficult to assess in clinical studies, but most authors report mortality rates between 0% and 11%.8- 10
The intestine is an important dose-limiting organ in abdominal and pelvic radiotherapy. Symptoms of acute radiation toxicity occur in most treated patients but are normally transient and resolve after completion of radiotherapy.11 Delayed radiation enteropathy, though less frequent, typically presents 3 months to 6 years after radiotherapy and has been reported to occur in up to 40% of patients receiving abdominal or pelvic radiotherapy.9,11 The incidence and severity of radiation enteropathy, however, depends on the dose of radiotherapy, the volume treated, fractionation schedule, and the length of follow-up. Additional risk factors include age, the presence of comorbidities, tobacco abuse, prior operation, and concurrent chemotherapy.11 In fact, some newer protocols are known to increase the occurrence of local complications, such as impaired anorectal function, following preoperative irradiation in patients treated with low anterior resection for rectal cancer.12
Typical symptoms in patients treated with radiotherapy in the lower abdomen or pelvis include diarrhea, rectal blood and/or mucus, and abdominal pain and can be graded according to the Radiation Therapy Oncology Group from 0 to 5 (Table 1). These symptoms usually originate from radiation enteritis or proctitis and generally respond to conservative management. Chronic or more extensive injuries, such as fistula formation or anastomotic strictures, often require surgical intervention. Historic reports have shown that the frequency of radiation injuries requiring an operation ranges from 2% to 17%.13- 16 In this article, we review our 10-year experience in the management of refractory radiotherapy complications that failed standard therapy to identify the specific distribution of radiotherapy-induced complications and to determine which complication or combination of complications could be treated surgically with satisfactory functional outcomes.
We conducted a retrospective cohort study analyzing all patients with complications from pelvic radiotherapy admitted to a university colorectal surgery unit during a 10-year period. Patient data were retrieved following approval according to the University of Louisville institutional review board guidelines. Clinical data included age, sex, site and pathology of original tumor, total radiotherapy dose, interval between radiotherapy and referral, symptoms suggestive of radiotherapy injury (graded according to Radiation Therapy Oncology Group guidelines), daily number and quality of bowel movements, treatment for radiotherapy injuries (medical and surgical), degree of continence, presence of stoma, and presence of cancer at last follow-up. To obtain a relative denominator of the number of patients in the state receiving pelvic radiation therapy for malignancy and number hospitalized for radiation-associated complications, demographic data from the state of Kentucky were retrieved from the following 3 independent sources and, whenever appropriate, reflect data from the corresponding 10-year time period: the Kentucky Department of Public Health, Kentucky Cancer Registry, and US Census Bureau (Table 2).
Chronic radiation-induced enterocolitis was suspected clinically by a constellation of symptoms, including diarrhea, abdominal pain, distension, cramping, and/or tenesmus more than 30 days after radiotherapy. The diagnosis of radiation enterocolitis was established by endoscopic, radiologic, and pathologic studies demonstrating characteristic radiation-induced tissue alterations in the absence of other discernible causes of gastrointestinal dysfunction. Diagnostic criteria included endoscopic findings of friable, edematous, bloody, firm, or fibrous mucosa; radiologic or intraoperative findings of strictures, fistulae, or extensive adhesions in formerly radiated areas; and/or pathologic alterations, including mucosal fibrosis and inflammation with microvascular ischemia.
Further analysis included the total dose of radiotherapy and its univariate correlation to the number and severity of subsequent radiotherapy injuries. Data are presented as mean (standard error of the mean) unless otherwise indicated. Groupwise statistical comparisons were performed using the unpaired t test or one-way analysis of variance according to the number of groups. For data that were not normally distributed, we used the Mann-Whitney rank sum test or the Kruskal-Wallis statistic. We performed all statistical computations using SigmaStat 3.01.0 software (Systat Software Inc, Richmond, California). Results were considered significant at P < .05.
Demographic data on the incidence of pelvic malignancies and the numbers of patients treated with pelvic radiotherapy in the state of Kentucky during the 10-year period from 1994 to 2003, including the number of inpatient admissions for pelvic radiotherapy complications, are provided in Table 2. Colorectal cancers were the most frequent pelvic neoplasms in the period studied, closely followed by prostate cancers. Patients with prostate cancer received radiotherapy more frequently (33% vs 14% of patients with colorectal cancers), thus composing the largest population of patients receiving radiotherapy. In 2004, 233 admissions occurred statewide for complications from pelvic irradiation, highlighting the overall frequency of complications requiring hospitalizations. On average, patients admitted for radiotherapy-related injuries were aged 68.0 (1.4) years, stayed 6.7 (0.6) days at the hospital, and caused health care expenditures of $19 205 ($2245) per hospital stay. The precise nature of the radiotherapy complications in these patients could not, however, be retrieved from the data sources available.
Throughout the 10-year period, 48 patients with refractory radiotherapy complications were evaluated at the Section of Colon and Rectal Surgery of the University of Louisville Department of Surgery. Patients ranged in age from 37 to 85 years (mean, 62  years) and had received pelvic radiotherapy for a variety of malignancies (Table 3). The interval from radiotherapy to complication varied greatly and ranged from 0 to 49 years. The delivery of radiation also varied as the population studied received treatment from 1950 to 2004. Radiation doses ranged from 30 Gy to 65 Gy with a mean and median of 47.33 Gy and 50 Gy, respectively. Nine patients (16%) received brachytherapy. Most patients (65%) had more than 1 radiation-associated complication: 29% had 2 complications, 21% had 3, and 15% of our patients had 4 or more. Following univariate analysis, a correlation existed between the radiation dose applied and the number of ensuing complications. Patients with 1 diagnosed complication of radiotherapy were treated with a mean dose of 41.9 (1.8) Gy, whereas patients with 3 or more radiotherapy complications received 47.6 Gy (2.4; P = .03).
The complications encountered in this series can be grouped into 3 broad categories: radiation enterocolitis, wound-healing disturbances leading to ulceration and fistula formation, and development of primary malignancies within irradiated tissue (Table 4). Although most patients had more than 1 radiotherapy complication, treatment was usually dictated by the most prominent form of radiotherapy injury in each patient. Table 5 presents the principal complications and subsequent treatment. Most patients had radiotherapy-induced bowel strictures, most frequently occurring at surgical anastomoses (Figure 1). Of the 16 patients who had anastomotic strictures, 11 occurred at a colorectal anastomosis, whereas 5 strictures affected coloanal anastomoses. In 8 of 11 patients with colorectal anastomosis strictures, the anastomotic segment was resected and converted to either ultra-low colorectal anastomosis or coloanal anastomosis with acceptable postoperative functional results with respect to the number of daily bowel movements and fecal continence (Table 5). Three patients underwent abdominoperineal resections owing to extremely poor sphincter function following radiotherapy. Dilatation was the initial treatment measure in all 5 patients with coloanal anastomotic stenosis. Two patients continued self-dilatations using a Hegar dilator with acceptable functional outcome, whereas 3 patients subsequently required permanent diversion.
Of the 4 patients with primary stenosis of an irradiated bowel, 1 patient with anal stenosis was treated successfully by dilatation, whereas the remaining 3 patients (2 rectal and 1 sigmoid stenosis) all required resection. Of these patients, 1 underwent coloanal anastomosis, which had to be converted to colostomy owing to poor functional results secondary to incontinence and severe radiotherapy enteritis. The other 2 patients had satisfactory functional results. One patient, however, subsequently died from cardiomyopathy suspected to result from the combination of whole-abdomen radiotherapy and adriamycin toxicity.
Of the 8 patients who had severe radiation enteritis (Figure 2), 5 developed short-bowel obstruction following radiotherapy, of whom one eventually developed multiple organ failure and died. The 3 other patients required multiple blood transfusions owing to associated telangiectasia, which was ultimately fatal in 1 patient.
Eight patients sustained radiotherapy-induced fistulae, of which, 3 were enterocutaneous, 3 were enterovaginal, 1 was enteroenteric, and 1 was rectourethral. All patients required resections with successful fistula repair. Finally, there were 2 cases of isolated radiation proctitis (Figure 3), one of which could be managed conservatively; the other, however, required rectal resection with coloanal anastomosis, which later had to be converted into abdominoperineal resection owing to unsatisfactory sphincter function.
Five patients had nonhealing ulcers and wounds diagnosed following radiotherapy. Two patients had rectal ulcers following radiotherapy for prostate cancer, 1 from a nonhealing abdominoperineal resection wound 6 years postoperatively and abdominoperineal resection despite numerous plastic surgery procedures. One patient had an anal canal ulcer eroding through both internal and external anal sphincters following a combination of external beam radiotherapy and brachytherapy for anal squamous cell carcinoma. This did not heal during an 18-month period despite diversion and hyperbaric oxygen therapy.
Five patients developed de novo cancers within the previously radiated area with tissue types different than the lesion requiring radiotherapy. Three of these tumors were rectal and 1 was uterine, all 4 being adenocarcinoma, and all led to the patients' death within 5 years of diagnosis. The fifth patient developed a squamous cell carcinoma within a presacral radiation ulcer, which was resected. The overall mortality in this series of 48 patients was 15% (7 patients), with a mean follow-up of 3 (range, 1-7 years) years. Three deaths were related to specific complications of radiotherapy, and 4 to secondary de novo cancers developing in the radiation field.
Figure 4 shows the distribution of radiotherapy-induced complications grouped according to the initial cancer for which the patients received radiotherapy. Radiation-induced proctitis was diagnosed more often in patients treated for prostate cancers (P < .01). Primary bowel strictures, anastomotic strictures, and small-bowel enteritis was observed more often in patients receiving radiotherapy for colorectal cancer (P = .08). In patients treated for cervical cancer, proctitis, strictures, and small-bowel enteritis occurred with equal frequency.
Radiotherapy is an established part of the oncologic care for a variety of abdominal and pelvic neoplasms and has markedly improved patient prognosis. The purpose of our study was to identify the spectrum of refractory radiotherapy complications and to determine which complication or combination of complications could be treated surgically with satisfactory functional outcomes. Using the Kentucky Cancer Registry database, we found that 61 000 patients or roughly 1.5% of the state population had a diagnosed pelvic tumor during the past 10 years and that approximately 24% of these patients received radiotherapy in the same period. Although we have no specific data regarding the type of radiotherapy complications requiring hospitalization statewide, those requiring surgical treatment are much less frequent.
Treatment-related complications occurring late after radiotherapy need to be distinguished from those occurring during or immediately after therapy, which are usually self-limiting and/or respond well to conservative treatment. Late complications of pelvic radiotherapy comprise a number of specific entities, including radiation enteritis, proctitis, and stricture or fistula formation. These changes typically arise after 6 months, are sometimes progressive, especially in severe forms, and may persist indefinitely.18 Radiation proctitis with light to moderate bleeding not requiring transfusion often resolves without specific treatment,18,19 whereas strictures, fistulae, and abscesses almost invariably require surgical intervention. The occurrence of these severe complications depends on a variety of patient- and treatment-related factors, such as radiotherapy dose and fractionation, patient age, tobacco abuse, concurrent chemotherapy, prior abdominal operation, or preexisting pelvic inflammatory disease.20 The reported incidence of radiotherapy injuries following irradiation of pelvic neoplasms ranges between 5% and 40%. Miller et al,9 in their series of 386 patients undergoing radiotherapy for rectal carcinoma, reported that 17% of patients developed clinically apparent radiotherapy-induced enteritis or proctitis, while Perez et al21 reported an 8% incidence of moderate proctitis in patients undergoing radiotherapy for prostate cancer within 10 years. In his review on late radiation proctitis, Babb22 reported incidence rates between 5% and 20% in patients treated for pelvic neoplasms. Most of these studies focus on the occurrence of radiotherapy injuries in a select number of patients treated for a specific tumor and do not specifically focus on treatment or severity of complications. It was the purpose of our study to evaluate the distribution of refractory complications and their response to surgical treatment. This implies, by nature of the study, a selection bias, because our patients have been referred with more severe complications, which is apparent from the mean late radiation toxicity grade of 3.4. We do not intend to imply that all such patients were referred to our unit; no doubt many were referred elsewhere. However, because of the referral pattern of our state with only 2 major medical centers, we believe that our series is representative of the spectrum of refractory radiation injuries statewide.
Most patients with refractory complications had more than 1 radiotherapy-associated complication and this may have contributed to the refractory nature. Our finding that two-thirds of our patients developed 2 or more complications from radiotherapy is consistent with previous literature. Kimose et al23 found a combination of radiotherapy injuries affecting the colon, rectum, small bowel, or the urinary tract in 62% of their 182 patients. Most patients presented herein required major operations, many with fecal diversion and subsequent restorative operations, because of the severity of radiotherapy injuries.
Good progress has been made in recent years with nonoperative treatment modalities for radiation injuries of a lesser degree; these modalities include medical treatment, hyperbaric oxygen, endoscopic formalin therapy, argon plasma coagulation, or laser therapy.24 Also, recent studies have suggested that preoperative radiotherapy may be more beneficial than postoperative radiotherapy with respect to long-term bowel function. In their series of 109 rectal cancer patients, Nathanson et al25 found that postoperative radiotherapy led to significantly more postoperative bowel movements and more episodes of clustered bowel movements compared with preoperative radiotherapy, presumably because of milder fibrosis in the reconstructed rectum.
Five patients (10%) in our series developed secondary cancers in previously radiated areas, 3 of which were rectal neoplasms. This is consistent with a recently published large retrospective cohort study26 that found an independent association between prostate radiation and the future development of rectal cancer in more than 30 000 patients receiving radiotherapy. The authors reported an adjusted hazards ratio for the development of rectal cancer following prostatic radiotherapy of 1.7 compared with patients treated with an operation alone. Other studies, however, have produced conflicting results of the potential risk of secondary rectal cancer development following radiotherapy of prostate cancer.27,28 Brenner et al27 have reported a 34% increase in solid tumors of the pelvis in patients surviving longer than 10 years after radiotherapy of prostate cancer, whereas Neugut et al28 have found only a slightly increased risk of bladder cancer following prostatic radiotherapy but no increased risk of rectal cancer.
The delivery of radiotherapy has changed dramatically during the last several years with the advent of 3-dimensional conformal and intensity-modulated therapy and the more frequent use of radioprotectors. The success of these measures to reduce complications has been reported in several sites.29- 32 Only with continued advances and review of failures and successes can we improve the control of these cancers and minimize complications.
In summary, most patients with refractory complications from pelvic radiotherapy develop radiation enteritis and colonic strictures; these complications occur most frequently at sites of previous anastomoses. Other, less frequent complications included radiation proctitis, fistula formation, and small-bowel obstruction, which occurred combined in two-thirds of all patients. Indications for operation typically included anatomic problems, such as symptomatic strictures, and the control of symptoms, such as diarrhea, incontinence, or bleeding that was unresponsive to nonoperative therapy. Aggressive surgical therapy, often involving temporary fecal diversion for fistulae and colorectal anastomotic strictures and resection of primary bowel strictures, produced the best results and often led to a dramatic improvement in symptoms. Patients with severe radiation enteritis and those with very distal strictures who presumably had associated radiation injury to the anal sphincter fared worse, and most of these patients required permanent fecal diversion.
Correspondence: Susan Galandiuk, MD, Department of Surgery, Ambulatory Care Bldg, 550 S Jackson St, University of Louisville, Louisville, KY 40292 (firstname.lastname@example.org).
Accepted for Publication: July 25, 2006.
Author Contributions:Study concept and design: Turina, Mahid, and Galandiuk. Acquisition of data: Turina and Mulhall. Analysis and interpretation of data: Turina, Mahid, and Yashar. Drafting of the manuscript: Turina and Mulhall. Critical revision of the manuscript for important intellectual content: Mahid, Yashar, and Galandiuk. Statistical analysis: Turina, Mulhall, and Mahid. Obtained funding: Galandiuk. Administrative, technical, and material support: Yashar and Galandiuk. Study supervision: Turina, Mahid, Yashar, and Galandiuk.
Financial Disclosure: None reported.
Additional Information: Data were obtained from the Kentucky Cancer Registry and the Commonwealth of Kentucky Cabinet for Health and Family Services with permission.