This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Eifel, P. J. Articles by Thames, H. Search for Related Content PubMed PubMed Citation Articles by Eifel, P. J. Articles by Thames, H. Social Bookmarking                            What's this?

Correlation of Smoking History and Other Patient Characteristics With Major Complications of Pelvic Radiation Therapy for Cervical Cancer

From the Departments of Radiation Oncology, Gynecologic Oncology, and Biomathematics, The University of Texas M.D. Anderson Cancer Center, Houston, TX.

Address reprint requests to Patricia Eifel, MD, Department of Radiation Oncology, Box 97, The University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030; email: peifel{at}mdanderson.org

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: The purpose of this study was to identify patient-related factors that influence the risk of serious late complications of pelvic radiation therapy.

PATIENTS AND METHODS: The records of 3,489 patients treated with radiation therapy for International Federation of Gynecology and Obstetrics stage I or II carcinoma of the cervix were reviewed for information about patient characteristics, treatment details, and outcomes. Any complication occurring or persisting more than 3 months after treatment that required hospitalization, transfusion, or an operation or caused severe symptoms or the patient’s death was considered a major late complication. Complication rates were calculated actuarially. The median duration of follow-up was 85 months, and 99% of patients were followed for at least 3 years or until they died.

RESULTS: Heavy smoking was the strongest independent predictor of overall complications (multivariate hazard ratio, 2.30; 95% confidence interval [CI], 1.84 to 2.87). The most striking influence of smoking was on the incidence of small bowel complications (hazard ratio for smokers of one or more packs per day, 3.25; 95% CI, 2.21 to 4.78). Hispanics had a significantly lower rate of small bowel complications than whites, and blacks had higher rates of bladder and rectal complications than whites. Thin women had an increased risk of gastrointestinal complications, and obese women were more likely to have serious bladder complications.

CONCLUSION: Complications of pelvic radiation therapy are strongly correlated with smoking, race, and other patient characteristics. These factors should be considered before the results of clinical studies are generalized to different cultural and racial groups.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
IN 1995, WE REPORTED the incidence and time course of late complications of radiation therapy in 1,784 patients treated for stage IB carcinoma of the uterine cervix.¹ Incidental to that analysis, a comparison of complication rates in women belonging to different racial or ethnic groups revealed that Hispanic women had a significantly lower rate of major gastrointestinal complications than white or black women. This preliminary finding led us to collect data from a larger number of patients treated at The University of Texas M.D. Anderson Cancer Center, Houston, TX, to confirm this finding and to investigate other patient characteristics that might contribute to the differences observed between these groups of women.

Factors that have been investigated for their possible contribution to late effects of pelvic radiation therapy include diabetes mellitus,^2,3 hypertension,³ pelvic infection,^4,5 previous abdominal surgery,^1,4,6,7 and age.⁴ The role of race or ethnicity has not previously been evaluated in large groups of women. The only previous study of the influence of smoking on late complications of pelvic radiation therapy was a 1987 analysis of women treated with radiation for cervical carcinoma in Austria. In that study, Kucera et al² found a significant correlation between smoking and the overall rate of major complications. The only other characteristic evaluated in their analysis was diabetes mellitus, which was not associated with a higher rate of complications.

The purpose of our analysis was to determine the relative influences of race, smoking, and other patient characteristics on the risk of serious late complications of pelvic radiation therapy.

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patients
The medical records of all patients treated with radiation therapy for squamous cell carcinoma of the uterine cervix between January 1960 and December 1994 at M.D. Anderson Cancer Center were retrospectively reviewed. Patients who underwent a hysterectomy as their initial treatment, patients who had already undergone partial hysterectomy when cervical cancer was diagnosed, and those who presented with recurrent disease were excluded. Of the remaining patients, 3,489 were treated with radiation therapy for International Federation of Gynecology and Obstetrics stage I or II tumors. Only studies specifically permitted by the International Federation of Gynecology and Obstetrics⁸ were used to determine tumor stage.

Data regarding patient, tumor, and treatment characteristics were abstracted from the hospital and radiation oncology records of each patient. Follow-up information was obtained from the patients’ records or from communications with patients or their physicians. Patients who were no longer being followed up at M.D. Anderson clinics had been contacted annually by the institution’s Department of Patient Studies to obtain information about tumor status and general medical problems. This information was recorded in each patient’s medical record.

Pretreatment Evaluation
All patients underwent a primary medical evaluation including a detailed history, review of symptoms, and physical examination before a treatment plan was formulated. The format and content of this evaluation were standardized and changed little during the 34-year period covered by the study. The intensity of smoking was usually stated in the medical record (Table 1), but the duration was less frequently recorded. Race was determined from registration information and from descriptions in the medical record. For this study, all patients who were described as Hispanic, Mexican, or Spanish speaking were categorized as Hispanic. Both a gynecologic oncologist and a radiation oncologist examined all patients in a multidisciplinary clinic. The clinical findings, including sites of involvement and tumor size, and the treatment plan were recorded in a clinic note and illustrated in a tumor diagram.

Radiation Treatment
Ninety-four percent of patients (3,296 of 3,489) were treated with a combination of external-beam and intracavitary radiation therapy; 113 patients (3%) were treated with intracavitary radiation therapy alone, and 80 (2%) were treated with external-beam radiation therapy alone. Patients were treated to the whole pelvis 5 days per week using daily fractions of 1.8 to 2 Gy; treatment was delivered using anterior- and posterior-opposed fields (96%) or four fields (4%). In most cases, the upper border of the radiation field was at the L4/5 interspace; however, 132 patients (3.8%) were also treated to the para-aortic nodes. Most patients received a maximum external-beam dose of 40 to 45 Gy to the central pelvis, but 138 patients (4%) received an external-beam dose of more than 50 Gy to the central pelvis. External-beam therapy was delivered using a 25-MV betatron or an 18- to 25-MV linear accelerator in 98% of patients. Fletcher-Suit or Fletcher-Suit-Delclos applicators loaded with radium or cesium were used for all intracavitary therapy.

Analysis of Complications
Complications of radiation therapy that occurred or persisted more than 3 months after treatment and required hospitalization, operation, or caused the patient’s death were considered major late complications. Any complication that fit this description was scored as an event, even if associated symptoms were of short duration. Severe complications that required transfusion or chronic narcotic use or caused severe diarrhea or bladder symptoms uncontrolled with medical management were also scored as major complications.¹ Symptoms that were directly related to recurrent cancer were not scored as complications. However, side effects that occurred in patients whose recurrent disease was at a distant site were scored as complications.

Follow-Up
After the completion of treatment, patients were usually asked to return to the clinic at 3-month intervals for 2 to 3 years, then at 6-month intervals for an additional 2 years, and then yearly. Any patient who did not return to the clinic for more than 1 year was referred to personnel in the M.D. Anderson Department of Patient Studies who attempted to contact the patient by telephone or letter and requested information about their health and disease status. Survival information that was unaccompanied by information about the patients’ health status was not included in this study. Ninety-nine percent of patients were followed up in the clinic, by letter, or by telephone contact for at least 3 years or until they died, and 1,623 patients were followed for more than 10 years. The median follow-up duration was 85 months. The median duration of follow-up in M.D. Anderson clinics was 60 months. One hundred six patients who died less than 5 years after treatment had an unknown or uncertain cause of death; their deaths were scored as treatment-related events in calculating disease-specific survival rates.

Statistical Methods
Complication rates were estimated using the Kaplan-Meier method with all time intervals measured from the date of initial treatment. Complication rates were quoted at 10 years unless otherwise stated. Complications of various types were scored as events on the first date that severe symptoms involving that site or sites were documented. This method was felt to be most appropriate because we were interested in the relative susceptibilities of surviving patients to various types of treatment complications. If there was no major late complication, cases were censored at the date of death or at the date of last contact with the patient, family member, physician, or other person with knowledge of the patient’s health status. Comparisons between actuarial curves were made using the log-rank method.

The nearest integers to the upper and lower quartiles were used as cutoff points for comparisons based on age or body mass index. Patients in the middle two quartiles were used as the baseline comparison groups. The significance of differences between proportions was tested with the ² statistic.

A proportional hazards regression model⁹ was used to evaluate the relative importance of predictive factors for late complications. One hundred twenty-seven patients (3.6%) had unknown values for one or more of the variables and were excluded from the regression analysis. For each end point, factors that did not contribute were eliminated from the model in a stepwise fashion. The primary emphasis of this study was to evaluate the influence of patient characteristics on complications. However, two treatment factors (the dose of radiation delivered to the central pelvis with external-beam irradiation and a history of transperitoneal lymph node dissection) were included in the regression model to rule out any possible confounding bias from selection of patients for these treatments; previous analyses had demonstrated these two factors to be strongly correlated with major complications. Treatments that we have not found to be correlated with the overall incidence of late treatment complications (eg, the use of adjuvant hysterectomy or concurrent chemotherapy)^1,10 were not included in the regression model.

Because this study spanned more than three decades, the regression analysis was run with and without the year of treatment included as a variable. Year of treatment was not significantly correlated with rectal, small bowel, or overall complications and achieved only marginal significance (hazard ratio, 0.98; P = .045) for bladder complications. For bladder complications, inclusion of treatment year only slightly modified the hazard ratios for other treatment variables and did not decrease the significance of any other correlations.

Because this study involved a large number of comparisons, calculations were also performed to correct for multiple testing using the Hochberg method.^11,12 All calculations were carried out using Stata statistical software.¹³

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patient Characteristics
The characteristics of the study patients are listed in Table 1. Significant correlations were found between race and other patient characteristics (Table 2). Hispanics were less likely than women in other ethnic groups to be smokers, and if they smoked, they were less likely to smoke a pack or more of cigarettes per day (Fig 1). Whites were the heaviest smokers. Blacks were more likely to be hypertensive, diabetic, or have a history of pelvic infection and had a greater mean body mass index than women belonging to other ethnic groups. Hispanic women had the youngest median age at diagnosis. Smokers had a lower mean body mass index than nonsmokers (25.7 v 28.2; P < .0005).

View larger version (59K): [in this window] [in a new window]   Fig 1. Correlation between race and smoking habit at the time of initial cervical cancer diagnosis.

Complications of Radiation Therapy
The relationships between patient characteristics and the incidences of major complications involving the bladder, rectum, or small bowel and the overall rate of major complications are listed in Table 3. Hispanics had a significantly lower rate of small bowel complications than whites. Their risk was also significantly less than that of black women in the study (P < .0005). There were no major small bowel complications in the 42 Asian women in the study. However, because of their small number, Asian women were excluded from the comparisons in Tables 3 and 4. Black women had a higher incidence of bladder complications than white women. This may, in part, have reflected the higher incidence of obesity in black women; obesity was also a risk factor for bladder complications. Very thin women (body mass index < 22) had a significantly increased risk of rectal and small bowel complications. Women who had a history of gonorrhea, syphilis, or nonspecific pelvic inflammatory disease were also more likely to experience serious late effects of radiation therapy.

View larger version (79K): [in this window] [in a new window]   Fig 2. Relationship between smoking intensity and the actuarial rate of major small bowel complications at 10 years.

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
These data demonstrate a strong relationship between patient characteristics or behaviors and side effects of pelvic radiation therapy. The correlation between smoking and late complications of radiation therapy is of particular interest. The highly significant difference in the incidence of gastrointestinal complications between smokers and nonsmokers and the correlation between complications and smoking intensity provide compelling evidence of a synergistic effect of smoking and radiation on normal tissue. This effect seems to be independent of other conditions, such as hypertension and thin body habitus, that have been associated with smoking or with complications of radiation therapy. The significance of the correlation also persisted even when the most conservative assumptions were used to correct the P values for multiple testing.

Most studies of the combined effects of smoking and radiation on normal tissue have focused on irradiated tissues in the upper aerodigestive tract. Several studies have suggested that persistent smokers have a higher risk of radiation-related complications, possibly due to a direct irritant effect.^14,15 However, clinical and laboratory studies suggest that persistent smoking may actually reduce the incidence of symptomatic radiation pneumonitis by reducing the inflammatory reaction to radiation.^16,17

The effects of smoking on the late effects of subdiaphragmatic irradiation have not been studied in detail. In 1987, Kucera et al² reported a correlation between smoking and complications in patients treated for cervical carcinoma. The authors suggested that the vasculotoxic effects of smoking and radiation combined to increase the severity of late tissue injury. However, their study did not include a detailed analysis of the types of complications associated with smoking and did not investigate the role of possible confounding factors such as body habitus, hypertension, and race. Although additive vascular injury may play a role in the complications experienced by smokers who undergo pelvic irradiation, a vascular hypothesis probably does not explain the much greater influence of smoking on small bowel than on bladder and rectal complications.

Interestingly, smoking is also a well-recognized risk factor for the development of Crohn’s disease; both the risk and severity of this form of inflammatory small bowel disease are positively correlated with the number of cigarettes smoked.¹⁸ Patients with Crohn’s disease also have a high rate of small bowel complications after radiation therapy.¹⁹ In contrast, smoking has been shown to reduce the incidence and severity of inflammatory bowel disease of the colon (ulcerative colitis). Smoking has a number of effects on intestinal function, including decreased intestinal permeability, reduced rectal blood flow, and increased colonic mucus production.¹⁸ Smoking has also been shown to inhibit production of tumor necrosis factor, interleukin-1, and interleukin-6 and to reduce the ratio of T-helper to T-suppressor cells. Although investigators have scrutinized these associations in efforts to explain the complex relationship between inflammatory bowel disease and smoking, the reasons for this relationship remain uncertain. However, it is interesting to speculate whether the smoking-related effects that contribute to the development of Crohn’s disease might also have a role in the development of severe small bowel injury after radiation therapy.

Smoking also has been associated with an increased incidence of peptic ulcer disease. Recent studies suggest that smoking may act by augmenting the harmful effects of Helicobacter pylori, both by adversely affecting upper gastrointestinal mucosal protection and physiology and by increasing the risk of H pylori infection.²⁰ Few of our patients were treated with radiation fields that included the stomach (there was only one major gastric complication), and there currently is no evidence that radiation enteritis is associated with H pylori infection. However, it is possible that similar alterations in small intestinal physiology could contribute to the development of small bowel complications in irradiated smokers.

It is not apparent from this study whether the added risk of complications in smokers derives from exposure before, during, or after radiation therapy. Patients who claimed to have quit smoking continued to have an increased risk of small bowel complications; however, the rate of recidivism after treatment was unknown. Although snuff users did not have a high rate of complications in this study, this group of patients was too small and the intensity of their habit too poorly characterized to permit speculation on the risk of exposure to nicotine from sources other than cigarettes, such as nicotine patches. Answers to such questions would have important implications for smoking cessation counseling.

Our original finding of a correlation between race and small bowel complications in patients treated with radiation for stage IB disease was unexpected.¹ However, we have now observed a similar significant correlation in three other groups of patients—in the patients with stage II disease included in the current study, in patients with stage IIIB disease,¹⁰ and in patients treated after hysterectomy for high-risk features (unpublished findings). Because racial categories were determined from the medical record rather than from detailed questionnaires, there probably were some inaccuracies in our identification of patients as Hispanic; we were also unable to identify patients of mixed ethnic backgrounds. However, the consistency of our findings strongly suggest that women who were identified as Hispanic in this study were less likely to suffer serious side effects of radiation. One purpose of the current study was to determine whether race was acting as a surrogate variable for other patient characteristics. The significantly lower rate of small bowel complications in Hispanic women was only partly explained by their tendency to smoke less than black or white women. Other differences—in age, weight, and rates of hypertension, diabetes, or pelvic infection—also failed to explain the strong correlation between race and the risk of major gastrointestinal complications. We had insufficient information to study socioeconomic factors that could be associated with race and radiation complications; however, medical indigence was prominent in all racial groups. Although significant dietary differences are known to exist between ethnic groups, a prospective study would be needed to evaluate the possible influence of dietary factors on treatment-related side effects.

We can only speculate on a possible role of genetic variation on patients’ differing susceptibilities to radiation injury. Genetic studies have suggested the presence of a locus on chromosome 3p that is linked with a susceptibility to inflammatory bowel disease.²¹ A similar genetic susceptibility could be playing a role in the etiology of radiation complications. However, the Mexican immigrants who form most of our Hispanic population are of diverse racial origin, as are the patients comprising our non-Hispanic patient population, reducing the likelihood of a simple genetic explanation for our observations.

An appropriately detailed analysis of the relationship between patient characteristics and tumor control was beyond the scope of this article. Clinical and laboratory studies, though not definitive, have suggested that hypoxia induced by heavy smoking may decrease the sensitivity of tumors to radiation.^2,22 If clinically significant, such an effect would further reduce the therapeutic ratio between tumor control and late complications of treatment.

It is likely that in future studies, smoking and race will be correlated with complications of radiation therapy in other groups of patients. Such studies would make it possible to explore the role of sex. It also would be interesting to evaluate prospectively the influence of other environmental factors, such as alcohol and diet. The results of our study suggest many interesting avenues of research into the mechanisms of radiation injury and the factors that promote its expression. This study also demonstrates that the results of clinical trials, particularly analyses of treatment-related side effects, cannot always be generalized to different cultural and racial groups.

	NOTES

 
Presented at the Forty-Second Annual Meeting of the American Society of Therapeutic Radiology and Oncology, October 22-25, 2000, Boston, MA.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
1. Eifel PJ, Levenback C, Wharton JT, et al: Time course and incidence of late complications in patients treated with radiation therapy for FIGO stage IB carcinoma of the uterine cervix. Int J Radiat Oncol Biol Phys 32: 1289-1300, 1995[CrossRef][Medline]

2. Kucera H, Enzelsberger H, Eppel W, et al: The influence of nicotine abuse and diabetes mellitus on the results of primary irradiation in the treatment of carcinoma of the cervix. Cancer 60: 1-4, 1987[CrossRef][Medline]

3. Potish RA: Importance of predisposing factors in the development of enteric damage. Am J Clin Oncol 5: 189-194, 1982[Medline]

4. Perez CA, Breaux S, Bedwinek JM, et al: Radiation therapy alone in the treatment of carcinoma of the uterine cervix: II. Analysis of complications. Cancer 54: 235-246, 1984[CrossRef][Medline]

5. Van Nagell JR, Parker JC, Maruyama Y, et al: The effect of pelvic inflammatory disease on enteric complications following radiation therapy for cervical cancer. Am J Obstet Gynecol 128: 767-771, 1977[Medline]

6. Fine BA, Hempling RE, Piver MS, et al: Severe radiation morbidity in carcinoma of the cervix: Impact of pretherapy surgical staging and previous surgery. Int J Radiat Oncol Biol Phys 31: 717-723, 1995[Medline]

7. Potish RA, Dusenbery KE: Enteric morbidity of postoperative pelvic external beam and brachytherapy for uterine cancer. Int J Radiat Oncol Biol Phys 18: 1005-1010, 1990[Medline]

8. Benedet J, Odicino F, Maisonneuve P, et al: Carcinoma of the cervix uteri. J Epidemiol Biostat 3: 5-34, 1998[Medline]

9. Cox DR: Regression models and life tables. J R Stat Soc Ser B 34: 187-220, 1972

10. Logsdon MD, Eifel PJ: FIGO stage IIIB squamous cell carcinoma of the uterine cervix: An analysis of prognostic factors emphasizing the balance between external beam and intracavitary radiation therapy. Int J Radiat Oncol Biol Phys 43: 763-775, 1999[CrossRef][Medline]

11. Brown B, Russell K: Methods correcting for multiple testing: Operating characteristics. Stat Med 16: 2511-2528, 1997[Medline]

12. Hochberg Y, Benjamini Y: More powerful procedures for multiple significance testing. Stat Med 9: 811-818, 1990[Medline]

13. StataCorp: Stata Statistical Software: Release 7.0. College Station, TX, Stata Corporation, 2001

14. Benninger MS, Gillen J, Thieme P, et al: Factors associated with recurrence and voice quality following radiation therapy for T1 and T2 glottic carcinomas. Laryngoscope 104: 294-298, 1994[Medline]

15. Kluth EV, Jain PR, Stuchell RN, et al: A study of factors contributing to the development of osteoradionecrosis of the jaws. J Prosthet Dent 59: 194-201, 1988[Medline]

16. Johansson S, Bjermer L, Franzen L, et al: Effects of ongoing smoking on the development of radiation-induced pneumonitis in breast cancer and oesophagus cancer patients. Radiother Oncol 49: 41-47, 1998[CrossRef][Medline]

17. Bjermer L, Cai Y, Nilsson K, et al: Tobacco smoke exposure suppresses radiation-induced inflammation in the lung: A study of bronchoalveolar lavage and ultrastructural morphology in the rat. Eur Respir J 6: 1173-1180, 1993[Abstract]

18. Andus T, Gross V: Etiology and pathophysiology of inflammatory bowel disease: Environmental factors. Hepatogastroenterology 47: 29-43, 2000[Medline]

19. Willett CG, Ooi CJ, Zietman AL, et al: Acute and late toxicity of patients with inflammatory bowel disease undergoing irradiation for abdominal and pelvic neoplasms. Int J Radiat Oncol Biol Phys 46: 995-998, 2000[CrossRef][Medline]

20. Parasher G, Eastwood G: Smoking and peptic ulcer in the Helicobacter pylori era. Eur J Gastroenterol Hepatol 12: 843-853, 2000[Medline]

21. Hampe J, Lynch NJ, Daniels S, et al: Fine mapping of the chromosome 3p susceptibility locus in inflammatory bowel disease. Gut 48: 191-197, 2001[Abstract/Free Full Text]

22. Siemann DW, Hill RP, Bush RS: Smoking: The influence of carboxyhemoglobin (HbCO) on tumor oxygenation and response to radiation. Int J Radiat Oncol Biol Phys 4: 657-662, 1978[Medline]

Submitted November 1, 2002; accepted May 15, 2002.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				E. R. Gritz, C. Dresler, and L. Sarna Smoking, The Missing Drug Interaction in Clinical Trials: Ignoring the Obvious Cancer Epidemiol. Biomarkers Prev., October 1, 2005; 14(10): 2287 - 2293. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Eifel, P. J. Articles by Thames, H. Search for Related Content PubMed PubMed Citation Articles by Eifel, P. J. Articles by Thames, H. Social Bookmarking                            What's this?