Circadian Variation in Radiation-InducedIntestinal Mucositis in Patients With CervicalCarcinomaPragya Shukla, MD1; Deepak Gupta, MD1; Shyam Singh Bisht, MD2; Mohan Chand Pant, MD1;

Madan Lal Bhatt, MD1; Rajeev Gupta, MD1; Kirti Srivastava, MD1; Seema Gupta, MD1;

Ankur Dhawan, MSc1; Deepti Mishra, MD3; and Mahendra Pal Singh Negi, MSc4

BACKGROUND: Mucositis, a radiotherapy-associated toxicity, is an important factor determining morbidity and treat-

ment compliance. Gastrointestinal mucositis in patients undergoing radiotherapy may also depend on time of admin-

istration of radiation in addition to several other factors. The presence of any correlation between the severity of

acute gastrointestinal mucositis in cervical carcinoma patients and the time of irradiation was prospectively eval-

uated. METHODS: A total of 229 patients with cervical carcinoma were randomized to morning (8:00-10:00 AM) and

evening (6:00-8:00 PM) arms. The incidence of mucositis in the 2 arms was assessed and reported in terms of various

grades of diarrhea. RESULTS: Overall (grade I-IV) as well as higher grade (III and IV) diarrhea was found to be signifi-

cantly increased in the morning arm as compared with the evening arm (overall: 87.39 % vs 68.18 %, P < .01; higher

grade: 14.29% vs 5.45%, P < .05). Other radiation-induced toxicity was also higher in the morning arm, but its occur-

rence in the 2 arms did not differ significantly (13.45% vs 12.73%, P > .05). After completion of treatment, patients’

response to radiation in the 2 arms was similar (P > .05). CONCLUSIONS: The significant difference in the incidence

of higher grade diarrhea between the morning and evening arms is indirect evidence of the influence of circadian

rhythm on the intestinal mucosa of the human intestine. This knowledge may facilitate treating patients with

decreased toxicity to the intestinal mucosa. Cancer 2010;116:2031–5. VC 2010 American Cancer Society.

KEYWORDS: rhythm, circadian, radiation toxicity, carcinoma cervix, mucositis.

Exposure to ionizing radiation during radiotherapy of the abdominopelvic region is associated with the development oftreatment-limiting untoward symptoms. Mucosal dysfunction resulting in acute radiation enteritis is caused by crypt celldeath, resulting in breakdown of the mucosal barrier and inflammation.1 This along with ulceration and pseudomem-brane formation is known as mucositis and is characterized by nausea, vomiting, and diarrhea; it may lead to dose reduc-tion, treatment interruptions, increased healthcare costs, and impaired quality of life of patients undergoing irradiation.1

The severity of intestinal mucositis in patients undergoing radiotherapy depends on a number of patient- and treatment-related variables, including age, smoking, total radiation dose, and fractionation schedule.2-4

Whenmice were irradiated at different times of the day, a clear circadian rhythm was observed in the number of apo-ptotic cells in the intestinal crypt, thus demonstrating the time-dependent ability of radiation to cause apoptosis.5,6 Theseobservations have also been confirmed in human rectal mucosa.7

The incidence of gastrointestinal toxicity relates to symptom and functional changes. Making accurate evaluation ofdamage difficult are the problems involved in obtaining sequential biopsy before, during, and after the treatment; thetendency of specimens to be superficial in nature; and the inaccessibility of important segments of the gastrointestinaltract.3

DOI: 10.1002/cncr.24867, Received: January 9, 2009; Revised: June 28, 2009; Accepted: June 29, 2009, Published online February 16, 2010 in Wiley InterScience

(www.interscience.wiley.com)

Corresponding author: Mohan Chand Pant, MD, Department of Radiotherapy, Chatrapati Sahuji Maharaj Medical University, Lucknow, India; Fax: (011)

510-2471553; profmcpant@gmail.com

1Department of Radiotherapy, Chatrapati Sahuji Maharaj Medical University, Lucknow, India; 2Department of Oncology, Delhi State Cancer Institute, New Delhi,

India; 3Department of Surgery, Chatrapati Sahuji Maharaj Medical University, Lucknow, India; 4Biometry and Statistics Division, Central Drug Research Institute,

Lucknow, India

We thank Dr. Rakesh Tuli, Director, CDRI, Lucknow for his encouragement and support. We also acknowledge those patients whose data were evaluated in the

study.

Cancer April 15, 2010 2031

Original Article

This randomized prospective trial attempted to findthe existence of any relation between the severity of acutegastrointestinal mucositis and the time of radiation inpatients with carcinoma of the cervix. We hypothesizedthat the patients receiving treatment in the evening (6:00-8:00 PM) would have lesser grades of radiation-inducedmucositis than those receiving it in the morning (8:00-10:00 AM), and also that the response of patients whoreceived treatment in the morning (8:00-10:00 AM) orevening (6:00-8:00 PM) would be the same.

MATERIALS AND METHODSThis randomized prospective study was conducted fromJuly 2006 to June 2008 on patients receiving externalbeam radiotherapy for carcinoma of the cervix at theDepartment of Radiotherapy, Chhatrapati Shahuji Maha-raj Medical University, Lucknow, India. Ethical clearancewas obtained from the institutional ethics committee. Intotal, 252 patients registered during this period wereincluded in the study. Informed consent was obtainedaccording to institutional ethics committee requirements.

Subjects

Patients with histologically confirmed nonmetastatic car-cinoma of the cervix with Karnofsky performance statusof 70% or above and stage IIB-IIIB were included in thisstudy. The patients were treated with curative intent andhad not received chemotherapeutic drugs (induction and/or concurrent chemotherapy). Subjects with any intestinalpathology that might interfere with primary endpointassessment and those who smoked during the treatmentwere excluded. Patients in whom treatment was inter-rupted for any cause other than mucositis (>7 days) wereexcluded during the evaluation of the results.

Protocol

After initial workup, which included a hemogram, renalfunction tests, chest x-ray, intravenous pyelography, cys-toscopy, and biopsy, patients began a regimen of externalradiation to the whole pelvis using anterior and posteriorparallel opposing fields to a dose of 50 grays in 25 frac-tions at 5 fractions/wk, using a Cobalt 60 teletherapy unitfollowed by intracavitary brachytherapy.

Before the start of external radiotherapy, the patientswere randomized using computer-generated randomnumbers. Patients allotted even numbers were included inGroup A and received radiation in the morning (8:00-10:00 AM); patients with odd number allocation wereincluded in Group B and received radiotherapy in the eve-ning (6:00-8:00 PM). All the patients were treated with cu-

rative intent, with no intended gaps during externaltherapy. They were given pretreatment dietary counselingthat was reinforced weekly. The overall treatment time forGroup A was 36.33 days and for Group B was 35.64 days.

All subjects followed their usual sleep schedule,sleeping between 10:30 PM and 6:30 AM. No enemas orbowel preparations were used during external beam radio-therapy. Apart from this, patients were given symptomatictreatment such as antimotility drugs and intravenous flu-ids as and when required.

Radiotherapy-induced mucositis was assessedweekly during external radiotherapy by a single observerand graded in terms of diarrhea using the Radiation Ther-apy Oncology Group’s Common Toxicity Criteria. Theobserver was blinded to the patient’s time of radiation.The primary endpoint of the study was to evaluate anydifference in the overall gastrointestinal mucositis andalso to separately compare the incidence of grade III andIV diarrhea in the 2 arms. In patients who developedgrade IV reactions, the treatment was interrupted untilthe reactions decreased by at least 2 grades. For analysis,they were included in the grade IV group in subsequenttreatment weeks. As the patients were treated with cura-tive intent, treatment response as well as radiation-induced toxicities other than gastrointestinal mucositiswere taken as secondary endpoints. The RECIST(Response Evaluation Criteria in Solid Tumors) criteriawere used for the assessment of response to treatment.

Statistical Analysis

The mean difference in continuous variables (age and he-moglobin) between the 2 groups was assessed by 2-samplet test,8 whereas the difference in the 2 proportions wasdone by proportion z test using its continuity for correc-tion.9 A 2-tailed (a ¼ 2) probability value P < .05 wasconsidered to be significant. The retrospective power ofeffect size was evaluated according to Cohen.8 Excel (MSOffice 97-2003; Microsoft, Redmond, Wash) and Prism(version 3; GraphPad Software, San Diego, Calif) wereused for the analysis.

Before conducting the study, we calculated the ap-proximate sample size considering the alpha a ¼ .05,power of the test 1-b ¼ .85, and proportions of totalmucositis (grade I þ II þ III þ IV) in patients of themorning arm (P1 ¼ .80) and the evening arm (P2 ¼0.60), such that the effective size ¼ 0.20, and the samplesize was 91. Therefore, we had initially planned to recruit100 subjects in each group. If the power is 0.85 with other

Original Article

2032 Cancer April 15, 2010

parameters as above, then the sample size should havebeen 103 patients per group.

RESULTSOf a total of 252 patients registered, 23 patients wereexcluded. Of 23 excluded patients, 8 did not completetreatment because of reasons other than treatment-relatedproblems, and 15 had radiation interrupted during thetreatment because of complications developed other thanmucositis, such as hemoglobin <10 mg/dL, leukopenia,and infection. Thus, 219 patients were found eligible,with 119 in the morning arm (Group A) and 110 in theevening arm (Group B). The overall treatment time of the2 arms (36.33 days vs 35.64 days) was similar (P>.05).

Patients’ baseline (pretreatment) characteristics inthe 2 groups are summarized in Table 1. On comparison,the baseline characteristics of the 2 groups were found notto differ significantly (P>.05).

The overall radiotherapy-induced mucositis (gradesI-IV) in patients of the 2 groups is summarized in Table 2.

This was found to be significantly higher in patients in themorning arm (z ¼ 3.36, P< .01) than in those irradiatedin the evening. Similarly, higher grade of mucositis (IIIand IV) in patients in the morning arm was also found tobe significantly (z¼ 2.00, P< .05) more frequent than inthe evening arm.

Other radiation-induced toxicity in patients withhigher grade (III and IV) disease is summarized for the 2groups in Table 3. On comparison, the proportions oftoxicity in patients of the 2 groups did not differ signifi-cantly (P > .05). Similarly, overall (total) proportions oftoxicity in patients of the 2 groups also did not differ sig-nificantly (z¼ 0.04, P> .05).

With a total of 229 patients in the 2 arms, the retro-spective power is �80% to detect a 19.21% difference(effect size) in overall proportions of mucositis in patientsof the morning arm (87.39%) and the evening arm(68.18%), at a 5% type I error rate and 2-sided alternative(a¼ 2) with 95% confidence.

Table 1. Pretreatment Characteristics of the 2 Groups

Characteristics Morning Arm,n5119, No. (%)

Evening Arm,n5110, No. (%)

Statistic, t and z,[df5227]

Age, y, mean�SD 49.81�4.76 48.93�5.14 1.35a

Well differentiated 34 (28.57) 31 (28.18) 0.08a

Moderately differentiated 37 (31.09) 38 (34.55) 0.41a

Poorly differentiated 21 (17.65) 18 (16.36) 0.08a

Undifferentiated 12 (10.08) 11 (10.00) 0.20a

Unknown 15 (12.61) 12 (10.91) 0.19a

Hemoglobin, g/dL, mean�SD 10.83�1.21 10.52�1.27 1.89a

KPS >70 >70 —

StageIIB 73 (61.34) 66 (60.00) 0.07a

IIIA 14 (11.76) 10 (9.09) 0.44a

IIIB 32 (26.89) 34 (30.91) 0.52a

SD indicates standard deviation; KPS, Karnofsky performance status.

Not significant, P > .05.

Table 2. Mucositis by Grade in Patients of the 2 Groups

Grade MorningArm, n5119,No. (%)

EveningArm, n5110,No. (%)

z Statistic[df5227]

I 55 (46.22) 52 (47.27) 0.03a

II 32 (26.89) 17 (15.45) 1.95a

III 13 (10.92) 5 (4.55) 1.55a

IV 4 (3.36) 1 (0.91) 0.82a

Total 104 (87.39) 75 (68.18) 3.36b

aNot significant, P>.05.bP < .01.

Table 3. Other Radiation-Induced Toxicity in Grade III and IVPatients of the 2 Groups

Toxicity MorningArm, n5119,No. (%)

EveningArm, n5110,No. (%)

z Statistic[df5227]

Skin reaction 5 (4.20) 7 (6.36) 0.44a

Nausea and

vomiting

5 (4.20) 2 (1.82) 0.66a

Bladder toxicity 1 (0.84) 1 (0.91) 0.65a

Hematological

toxicity

5 (4.20) 4 (3.64) 0.12a

Total 16 (13.45) 14 (12.73) 0.04a

aNot significatnt, P>.05.

Circadian Rhythm in Intestinal Mucositis/Shukla et al

Cancer April 15, 2010 2033

The response to radiation in patients of the 2 groupsafter completion of treatment is summarized in Table 4.The proportion of patients showing complete response,partial response, static disease, and progression of diseasein the 2 groups did not differ significantly (P> .05).

DISCUSSIONThe hypothesis that the response of patients who receivedtreatment in the morning or the evening would be thesame was found to be true. Overall (grade I-IV) as well ashigher grade (III and IV) diarrhea was found to be signifi-cantly more frequent in the morning arm as comparedwith the evening arm (overall, 87.39% vs 68.18%, P <

.01; higher grade: 14.29% vs 5.45%, P< .05). The mech-anism behind this significant difference is not yet clear.We hypothesized that the significant difference in theincidence of higher grade diarrhea between the morningand evening arms is indirect evidence of the influence ofcircadian rhythm on the intestinal mucosa of the humanintestine.

A total of 229 patients were finally evaluated in thisstudy. Given this sample size, the effective size of totalmucositis is 19.21 and at a¼ .05, the retrospective poweris 87%. Thus, for total mucositis, the sample size wasfound to be adequate. As overall radiation induced muco-sitis was the primary endpoint, and this study did not findany significant difference in mucositis between patients inthe 2 groups at each grade or in radiation-induced toxicityat higher grades, the sample size was evidently not suffi-cient, although the accrual goal was met within the antici-pated period of time.

Mucositis is an important factor determining mor-bidity and treatment compliance. Prolonging overalltreatment time results in a decrease in tumor control of0.55%-1.4% per day of treatment elongation.10

According to the law of Bergonie and Tribondeau,18

tissue appears to be more radiosensitive if its cells are less

well differentiated, have greater proliferative capacity, anddivide more rapidly. Radiotherapy-associated mucositis iscaused by the effect of ionizing radiation on proliferatingepithelial stem cells and partly results from damage to en-dothelial cells.3 Paris et al11 proposed that crypt cell deathis an indirect consequence of endothelial cell apoptosis.The lower 2=3 of the crypt is considered the normal prolif-erative zone, and the upper 1=3 is occupied by nondividingdifferentiated cells.12 Terpstra et al13 reported that epithe-lial proliferation is fairly uniform along the entire lengthof the large and the small bowel.

This randomized prospective trial was conducted todiscover any correlation between radiation-related muco-sitis and the time of radiation. We have evaluated the inci-dence of grade III/IV mucositis in patients withcarcinoma of the cervix treated between 8:00 and 10:00AM and between 6:00 and 8:00 PM. In our study, we haveseen that the patients undergoing radiation in the morn-ing showed a significantly higher incidence of grade IIIand IV mucositis. The progression trend also suggests thatthe patients given radiation in the morning exhibitedhigher grades of mucositis. This lower incidence of gradeIII/IV mucositis in patients who were treated in the eve-ning may be because of the effect of circadian rhythm inthe cell cycle of normal mucosa.

Studies on the effect of radiation on mice haveclearly demonstrated that the peak time of day for induc-ing apoptosis is between 6:00 and 9:00 AM, with thetrough occurring between 6:00 and 9:00 PM.5,6 Studies ofcellular proliferation in the human rectal mucosa haveshown the existence of a circadian rhythm of cell prolifera-tion in rectal mucosa, with the highest proliferative activityoccurring in the morning between 3:00 and 11:30 AM andthe least activity occurring 12 hours later.7 The differencein the grade of mucositis in the 2 study groups could beexplained on the basis of these studies.

Klevecz et al14,15 and Smaaland et al16 have reporteda phase opposition between DNA synthesis rhythms inhealthy target organs and in tumor. This might explainthe similar response rate achieved in the 2 groups of thisstudy (82.35% in morning irradiated patients and80.91% in evening irradiated patients).

Conclusions

The difference in the incidence of grade III and IV diar-rhea in the morning and the evening groups is clinical evi-dence of the influence of circadian rhythm on the humanintestinal mucosa. To the best of our knowledge, this isthe first study to evaluate the existence of any correlation

Table 4. Response Rate in Patients of the 2 Groups

Response MorningArm, n5119,No. (%)

EveningArm, n5110,No. (%)

z Statistic[df5227]

Complete

response

98 (82.35) 89 (80.91) 0.11a

Partial response 20 (16.81) 20 (18.18) 0.10a

Static 0 (0.00) 0 (0.00) —

Progression

of disease

1 (0.84) 1 (0.91) 0.65a

aNot significant, P>.05.

Original Article

2034 Cancer April 15, 2010

between the time of irradiation and the severity of mucosi-tis. This study does not substantiate the correlation withdata, but provides a theoretical possibility of treating thepatients more effectively when the tumor cells are in a ra-diosensitive phase, with treatment at the same time beingless toxic to the surrounding tissues, which are in radiore-sistant phase. Marking the time of the day when normalcells and tumor tissues are in different phases is an experi-mental possibility that could be investigated in studies ofcyclin-dependent kinases and other cell cycle markerproteins.

CONFLICT OF INTEREST DISCLOSURESThe authors report no conflicts of interest.

REFERENCES1. Walsh D. Deep tissue traumatism from roentgen ray expo-

sure. BMJ. 1897;2:272-273.2. Dodd MJ, Miaskowski C, Shiba GH, et al. Risk factors for

chemotherapy-induced oral mucositis: dental appliances,oral hygiene, previous oral lesions, and history of smoking.Cancer Invest. 1999;17:278-284.

3. Sonis ST, Fey EG. Oral complications of cancer therapy.Oncology (Williston Park). 2002;16:680-686.

4. Barasch A, Peterson DE. Risk factors for ulcerative oralmucositis in cancer patients: unanswered questions. OralOncol. 2003;39:91-100.

5. Ijiri K, Pottents CS. Circadian rhythms in the incidence ofapoptotic cells and number of clonogenic cells in intestinalcrypts after radiation using normal and reversed light condi-tions. Int J Radiat Biol Relat Stud Phys Chem Med.1988;53:717-727.

6. Ijiri K, Potten CS. The circadian rhythm for the numberand sensitivity of radiation-induced apoptosis in the cryptsof mouse small intestine. Int J Radiat Biol. 1990;58:165-175.

7. Buchi KN, Moore JG, Hrushesky WJ, Sothern RB, RubinNH. Circadian rhythm of cellular proliferation in thehuman rectal mucosa. Gastroenterology. 1991;101:410-415.

8. Cohen J. Statistical Power Analysis for the Behavioral Scien-ces. 2nd ed. Mahwah, NJ: Lawrence Erlbaum; 1988.

9. Zar JH. Biostatistical Analysis. Englewood Cliffs, NJ: Pren-tice-Hall; 1974.

10. Fyles A, Keane TJ, Barton M, Simm J. The effect of treat-ment duration in the local control of cervix cancer. Radio-ther Oncol. 1993;28:179-180.

11. Paris F, Fuks Z, Kang A, et al. Endothelial apoptosis as theprimary lesion initiating intestinal radiation damage inmice. Science. 2001;293:293-297.

12. Marra G, Anti M, Percesepe A, et al. Circadian variation ofepithelial cell proliferation in human rectal crypts. Gastroen-terology. 1994;106:982-987.

13. Terpstra OT, van Blankenstein M, Dees J, Eilers GA.Abnormal pattern of cell proliferation in the entire colonicmucosa of patients with colon adenoma or cancer. Gastroen-terology. 1987;92:704-708.

14. Klevecz R, Shymko R, Braly P. Circadian gating of Sphase in human ovarian cancer. Cancer Res. 1987;47:6267-6271.

15. Klevecz R, Braly P. Circadian and ultradian cytokineticrhythms of spontaneous human cancer. Ann NY Acad Sci.1991;618:257-276.

16. Smaaland R, Lote K, Sothern RB, et al. DNA synthesis andploidy in non-Hodgkin’s lymphomas demonstrate variationdepending on circadian stage of cell sampling. Cancer Res.1993;53:3129-3138.

17. Farker K, Merkel U, Wedding U, Hippius M, Hoffken K,Hoffmann A. Chronomodulated chemotherapy with oxali-platin, 5-FU and sodium folinate in metastatic gastrointesti-nal cancer patients: original analysis of non-hematologicaltoxicity and patient characteristics in a pilot investigation.Int J Clin Pharmacol Ther. 2006;44:31-37.

18. Hall EJ. Radiosensitivity and cell age in the mitotic cycle.In: Hall EJ, ed. Radiobiology for the Radiobiologist. 5th ed.Philadelphia, PA: JB Lippincott; 2000:339-340.

19. Mormont MC, Levi F. Cancer chronotherapy: principles,applications, and perspectives. Cancer. 2003;97:155-169.

20. Mormont MC, Levi F. Circadian-system alterations duringcancer processes: a review. Int J Cancer. 1997;70:24-47.

Circadian Rhythm in Intestinal Mucositis/Shukla et al

Cancer April 15, 2010 2035