Complications of Radiotherapy in Laryngopharyngeal Cancer

Abstract

BACKGROUND:

Radiotherapy (XRT) is effective as the primary treatment modality for laryngopharyngeal cancer; however, complications of XRT can result in significant morbidity. Few previous studies have examined the effect of continued smoking on complications of XRT. The authors of this report hypothesized that patients with laryngopharyngeal cancer who successfully quit smoking would have fewer complications of primary XRT.

METHODS:

All patients with head and neck cancer who were smokers at the time of diagnosis were referred prospectively to the Tobacco Treatment Program (TTP). From this group, the patients with laryngopharyngeal cancer who received XRT as the primary treatment modality were retrospectively selected and studied.

RESULTS:

Eighty-six patients were identified and were divided into 3 groups: Seventeen patients attended TTP and quit smoking before the start of XRT (Group 1), 33 patients attended TTP but continued to smoke during XRT (Group 2), and 37 patients refused TTP (Group 3). On the basis of a review of medical records for patients in Group 3, 20 patients quit smoking before starting XRT and were included in Group 1 (abstainers), 11 patients continued to smoke and were included in Group 2 (continued smokers), and 6 patients had incomplete data and were omitted from further analysis. Analyses both with and without Group 3 patients yielded similar results. Abstainers and continued smokers had similar demographic and clinical characteristics. With the exception of skin changes, all complications (mucositis, need for feeding tube, duration of feeding tube, need for hospitalization, pharyngeal stricture, and osteoradionecrosis) were more common in the patients who continued to smoke, although the only complications that were significantly more common were the need for hospitalization (P = .04) and osteoradionecrosis (P = .03). Patients who continued to smoke were more likely to develop osteoradionecrosis (relative risk [RR], 1.32; 95% confidence interval [CI], 1.09–1.6; P = .03) and to require hospitalization during treatment (RR, 1.46; 95% CI, 1.05–2.02; P = .04).

CONCLUSIONS:

Continued smoking during treatment appeared to increase the risk for complications of XRT for patients with laryngopharyngeal cancer and possibly increased hospitalizations. This hypothesis-generating study emphasized the importance of smoking cessation programs in the management of patients with head and neck cancer patients who receive XRT.

Although radiotherapy (XRT) is effective as the primary treatment modality for many head and neck cancers, complications of XRT can result in significant morbidity and a decrease in quality of life. These complications can be classified as acute and late tissue reactions. Acute tissue reactions affect rapidly dividing cells, such as skin and mucosa; occur during or shortly after XRT; and generally are reversible. Late tissue reactions affect more slowly dividing cells, such as bone or cartilage; occur >90 days after the end of XRT; and often are associated with permanent morbidity.¹

Whereas it is known that the development of XRT complications is influenced by multiple treatment-related factors, including total dose of radiation, total number of fractions, time interval between treatment sessions, and total treatment time,^2,3 the role of patient-related factors is less clear. Previous studies have demonstrated that continued smoking after a diagnosis of head and neck cancer has an impact on tumor-related outcomes, including decreased survival time, increased risk of recurrence, and increased risk of second primary malignancies.^4–6 The role of continued smoking on the development of XRT-related complications has not been clearly elucidated.

It has been demonstrated that up to 48% of current smokers who are diagnosed with squamous cell carcinoma of the head and neck will continue to smoke.^7–9 However, several studies have demonstrated promising results in patients with head and neck cancer who were enrolled in smoking cessation programs. Quit rates of up to 70% at 30-month follow-up in patients with head and neck cancer who were enrolled in a smoking cessation program have been demonstrated.¹⁰ Multiple approaches to smoking cessation have been described, with combined interventions, including both medical treatment and intensive counseling, demonstrating the best complete abstinence rates.¹¹

Smokers who present at our institution are offered enrollment in an intensive smoking cessation program (the Tobacco Treatment Program [TTP]), which offers counseling and pharmacologic management for patients who are attempting to quit and support for those who have decided to quit. In addition, objective measures of smoking cessation, such as exhaled carbon monoxide levels, are recorded and used to monitor patient progress. The objective of the current study was to examine the effect of smoking and smoking cessation on patients with laryngopharyngeal cancer who were receiving XRT as their primary treatment modality, and we hypothesized that those patients who successfully quit smoking would have lower rates of XRT-related complications.

MATERIALS AND METHODS

All patients with head and neck cancer who were new to the institution and were current smokers at the time of diagnosis were referred to TTP by their physician or were contacted proactively by the TTP for enrollment. The initial TTP session was conducted face-to-face with the patient, and information on smoking and alcohol history, as well as motivation to quit, was collected prospectively. Those patients who enrolled in the program were followed with a combination of inpatient hospital visits, outpatient clinic visits, and telephone visits. Abstinence rates, as measured by the commonly accepted 7-day point prevalence criterion (ie, abstinence is defined as no smoking for at least 7 consecutive days), were recorded by the TTP staff through patient self-report. Patients were assessed by their TTP clinician in addition to a TTP research assistant throughout the course of treatment. If a discrepancy existed between patient self-reports with various TTP staff, then the patient was asked additional questions to accurately obtain their abstinence rate. The typical course of treatment in the TTP consisted of 6 sessions with a clinician in combination with a tobacco-cessation medication over the course of 3 months.

From the list of patients who were referred to TTP, we selected patients with laryngopharyngeal who were receiving XRT as their primary treatment modality. Patients with cancer of the oral cavity, salivary glands, skin, or thyroid were excluded, because XRT is not the primary treatment for these malignancies. In addition, patients with laryngopharyngeal cancer who did not receive XRT as their primary treatment were excluded. These patients were divided into 3 groups: patients who were enrolled in TTP and quit successfully before XRT (abstainers; Group 1), patients who enrolled in the program but continued to smoke through XRT (continued smokers; Group 2), and patients who were offered the program but did not enroll (Group 3). Two sets of data were analyzed: The first set included only patients who had smoking data collected prospectively (Groups 1 and 2), and the second set also included patients who did not enroll in TTP (Group 3) and had smoking data collected retrospectively.

Medical records were reviewed for medical comorbidities, primary site, clinical stage, treatment, histologic grade, and outcome measures. Medical comorbidities were classified according to the Kaplan-Feinstein Comorbidity Scale, which reflects the severity of medical comorbidities, from none to mild, moderate, or severe disease.¹² Comorbid conditions included the following categories: cardiac, endocrine, hypertension, renal, respiratory, cerebral/psychological, hepatic, locomotor, gastrointestinal, and peripheral vascular disease; alcoholism; other malignancies; and miscellaneous. Histologic grade was defined as “well” if it was recorded as either well or moderately well differentiated, and it was classified as “poor” if it was recorded as either poorly or moderately poorly differentiated. Treatment history, including XRT dose and the number of fractions, duration of therapy, and the use of concurrent or induction chemotherapy, also was collected retrospectively from the medical record. All patients received their treatment at The University of Texas M. D. Anderson Cancer Center.

Data on the onset and severity of XRT complications were extracted from the medical record in addition to data on the need for and duration of a feeding tube and the need for hospitalization during treatment. XRT-associated skin changes and mucositis were graded 0, I, II, III, or IV according the Radiation Therapy Oncology Group Acute Toxicity Scales.

The Student t test was used to compare continuous variables between groups with the Satterthwaite adjustment for unequal variances when indicated by testing for the equality of variances between groups. Chi-square analyses were used to compare categorical variables between groups. For comparisons of categorical variables that had 1 or more cells with <5 patients, Fisher exact analyses were used. SAS software (SAS PROC FREQ, version 9.1; SAS Institute Inc., Cary, NC) was used to calculate relative risks (RRs) with 95% confidence intervals (CIs). All tests were 2-sided, and P = .05 was set as the level of statistical significance. Statistical analyses were performed with STATA software (version 7.0; Stata, Inc., College Station, Tex) or with SAS software (version 9.1; SAS Institute Inc.).

RESULTS

Eighty-seven patients were identified who fit the criteria for this study, including 17 patients who attended TTP and quit before the start of radiation (abstainers; Group 1), 33 patients who attended TTP but continued to smoke during treatment (continued smokers; Group 2), and 37 patients who refused TTP (Group 3). On the basis of a review of the medical records for Group 3 patients, 20 patients quit before they started XRT, 11 patients continued to smoke, and 6 patients had incomplete data and were omitted from further analysis.

No significant differences were observed in the distribution of patients according to age, sex, ethnicity, or medical comorbidities between abstainers and continued smokers (Table 1). The mean age (±standard deviation) was 56.2 ± 11.9 years for abstainers (median, 55 years; range, 22–75 years) and 55 ± 9.1 years for continued smokers (median, 55 years; range, 44–70 years; P = .6).

Table 1.

Demographic and Comorbid Characteristics of Abstainers and Continued Smokers

	Abstainers, n=17		Continued Smokers, n=33
Variable	No.	%	No.	%	P *
Age, y
≤55	7	41.2	16	48.5	.62
>55	10	58.8	17	51.5
Sex
Men	13	76.5	25	75.8	.95
Women	4	23.6	8	24.2
Ethnicity
Non-Hispanic white	12	70.6	29	87.9	.13
Other	5	29.4	4	12.2
Kaplan-Feinstein Comorbidity Scale
None-mild	11	33.3	25	75.8	.41
Moderate to severe	6	56.7	8	24.2

^*Chi-square or Fisher exact test, as appropriate.

There were no differences between abstainers and continued smokers in the distribution of primary site, stage, or grade (Table 2). In addition, a similar proportion of each group received XRT alone versus XRT with chemotherapy (P = .6) (Table 2). The mean XRT dose, number of XRT fractions, and XRT treatment time all were similar between abstainers and continued smokers (Table 2). Although the mean follow-up was shorter for continued smokers than for abstainers, the difference was not significant (Table 2).

Table 2.

Tumor and Treatment Characteristics of Abstainers and Continued Smokers

	Abstainers, n=517		Continued Smokers, n=533
Variable	No.	%	No.	%	P *
Site of primary tumor
Nasopharynx	0	0	1	3	.87
Oropharynx	11†	64.7	20	60.6
Hypopharynx	1	5.8	3	9.1
Larynx	5	29.4	9	27.3
Tumor grade/differentiation ‡
Well/moderately well	2	11.8	3	9.1	.53
Moderate	7	41.1	17	51.5
Poor/moderately poor	7	41.1	8	24.2
Tumor classification
T0-T2	7	41.1	18	54.5	.37
T3-T4	10	58.9	15	45.5
Lymph node status
N0	5	29.4	11	33.3	.78
N1-N3	12	70.6	22	66.7
Overall stage
I	2	11.8	4	12.1	.44
II	0	0	4	12.1
III	5	29.4	6	18.1
IV	10	58.8	19	57.7
Treatment
XRT alone	7	41.1	10	30.3	.57
Ind. CTX→XRT	0	0	3	9.1
Concurrent CTX-XRT	7	41.1	15	45.5
Ind. CTX→CTX-XRT	3	17.8	5	15.1
No CTX	7	41.1	10	30.3	.44
Received CTX	10	58.8	23	69.7
Variable	Mean ± SD	Median (Range)	Mean ± SD	Median (Range)	P
Radiation dose, Gy	68.3 ± 3.3	70 (63–72)	68.9 ± 4.3	70 (48–72)	.65
No. of radiation fractions	32.9 ± 3.17	33 (30–42)	33.2 ± 5.0	33 (12–42)	.84
Radiation schedule, d	44.2 ± 4.3	44 (39–51)	43.9 ± 2.9	44 (38–51)	.40
Follow-up, d	533.4 ± 227.4	519 (103–1026)	515.8 ± 395.0	395.7 (111–1210)	.60

XRT indicates radiation therapy; CTX, chemotherapy; Ind., induction; SD, standard deviation; Gy, grays.

^*The chi-square or the Fisher exact test was used for categorical variables, as appropriate, and t tests were used to compare means (with adjustment for unequal variances, as appropriate) for continuous variables.

^†Includes 1 patient who had cervical metastasis of unknown primary origin. Radiation fields for this patient were consistent with oropharyngeal fields.

^‡Tumor grade was not recorded for 1 abstainer and for 5 continued smokers; moderately to well differentiated was listed as well differentiated, and moderately to poorly differentiated was listed as poorly differentiated.

Although moderate-to-severe skin changes were more common in the abstainers, these results were not statistically significant, and all other complications were more common in continued smokers (Table 3). Continued smoking was associated with an elevated risk of developing moderate-to-severe mucositis (Table 3) (P = .1); however, this elevation was not statistically significant. Three patients who continued to smoke required pharyngeal/esophageal dilatation for stricture compared with none of the abstainers. The need for hospitalization was significantly greater for continued smokers compared with abstainers (Table 3) (P = .04). The chief reason for hospitalization was dehydration in both groups (78% of continued smokers and 75% of abstainers), followed by airway management, pneumonia, and pain control. Finally, osteoradionecrosis was more common in continued smokers than in abstainers (Table 3) (P = .03). Of the 8 continued smokers who developed osteoradionecrosis, 4 required major resection for treatment. None of the patients in the abstainer group developed osteoradionecrosis.

Table 3.

Complications of Radiation Therapy in Abstainers and Continued Smokers and the Relative Risks for Each Complication Associated With Continued Smoking

	Abstainers, n=517		Continued Smokers, n=533
Complication	No.	%	No.	%	P *	RR of Complication Associated With Continued Smoking (95% CI)
Skin changes
Grade 0-I	7	41.2	20	60.6	.19	0.7 (0.36–1.28)
Grade II-IV	10	58.8	13	39.4
Mucositis
Grade 0-I	7	41.2	7	21.2	.13	1.9 (0.81–4.63)
Grade II-III	10	58.8	26	78.8
Feeding tube needed
No	7	41.2	11	33.3	.58	0.9 (0.42–1.75)
Yes	10	58.8	22	66.7
Hospitalization during treatment
No	15	88.2	20	60.6	.04	1.5 (1.05–2.02)
Yes	2	11.8	13	39.4
Pharyngeal stricture requiring dilatation
No	17	(100)	30	(90.9)	.20	1.1 (.99–1.23)
Yes	0	(0)	3	(9.1)
Osteoradionecrosis
No	17	(100)	25	(75.8)	.03	1.3 (1.09–1.6)
Yes	0	(0)	8	(24.2)
Osteoradionecrosis needing major resection
No	17	(100)	29	(87.8)	.13	1.1 (1.0–1.29)
Yes	0	(0)	4	(12.2)
Variable	Mean ± SD	Median [Range]	Mean ± SD	Median [Range]	P *	RR (95% CI)
Feeding tube duration, d†	250.0 ± 119.6	225 [48–362]	208.2 ± 225.4	152 [42–1043]	.31	—
Hospitalization duration, d	3.5 ± 2.4	3.5 [2–5]	8.7 ± 12.8	3.5 [2–43]	.26	—

RR indicates relative risk; CI, confidence interval; SD, standard deviation.

^*The chi-square or Fisher exact test, as appropriate, was used for categorical variables, and t tests were used to compare means (with adjustment for unequal variances. as appropriate) for continuous variables.

^†Includes 12 patients (3 abstainers and 9 continued smokers) who had a feeding tube present at last contact.

When Group 3 patients were included in the analysis, there were no statistically significant differences between continued smokers and abstainers with respect to age, sex, ethnicity, comorbidities, tumor characteristics, or treatment characteristics. We again noted that all complications, with the exception of skin changes, were more common in continued smokers (Table 4). The increased risk of developing esophageal/pharyngeal stricture in patients who continued to smoke approached statistical significance (P = 06). Like in the previous group, excluding Group 3, a statistically significant association was noted between continued smoking and the development of osteoradionecrosis (P = .02) and the need for hospitalization (P = .03). Also, continued smokers had a statistically significant increase in length of stay compared with abstainers (P = .01).

Table 4.

Complications of Radiation Therapy in Abstainers and Continued Smokers and the Relative Risks for Each Complication Associated With Continued Smoking (Group 3 Patients Included)

	Abstainers, n=37		Continued Smokers, n=44
Complication	No.	%	No.	%	P*+	RR for Complication Associated With Continued Smoking (95% CI)
Skin changes
Grade 0-I	21	56.8	30	68.1	.29	0.8 (0.59–1.18)
Grade II-IV	16	43.2	14	31.9
Mucositis
Grade 0-I	16	43.2	12	27.2	.13	1.5 (0.86–2.91)
Grade II-III	21	56.8	32	72.8
Feeding tube needed
No	16	43.2	16	36.4	.53	1.1(0.69–2.04)
Yes	21	56.8	28	63.6
Hospitalization during treatment
No	32	86.5	29	65.9	.03	1.3 (1.02–1.68)
Yes	5	13.5	15	34.1
Pharyngeal stricture requiring dilatation
No	37	100	40	90.9	.06	1.1 (1.0–1.21)
Yes	0	0	4	9.1
Osteoradionecrosis
No	36	97.3	35	79.5	.02	1.2 (1.04–1.43)
Yes	1	2.7	9	20.5
Osteoradionecrosis needing major resection
No	37	100	40	90.9	.06	1.1 (1.0–1.21)
Yes	0	0	4	9.1
Variable	Mean ± SD	Median [Range]	Mean ± SD	Median [Range]	P *	RR (95% CI)
Feeding tube duration, d†	206.6 ± 138.3	186 [48–449]	193.3 ± 202.7	152 [5–1043]	.54	—
Hospitalization duration, d	3.8 ± 2.2	3 [2–7]	8.2 ± 11.8	4 [2–43]	.01	—

RR indicates relative risk; CI, confidence interval; SD, standard deviation.

^*The chi-square or Fisher exact test, as appropriate, was sued for categorical variables, and t tests were used to compare means (with adjustment for unequal variances, as appropriate) for continuous variables.

^†Includes 16 patients (6 abstainers and 10 continued smokers) who had a feeding tube present at the last contact.

DISCUSSION

The current study demonstrated the success of the TTP, which resulted in 35% of patients with laryngopharyngeal cancer who were enrolled in the program quitting by the end of treatment (3 months). In addition, we demonstrated that, with the exception of XRT-related skin changes, early and late complications of XRT were more common in patients who continued to smoke during treatment for laryngopharyngeal cancer.

Early complications of XRT can have a profound impact on a patient’s quality of life and their ability to undergo complete treatment. Mucositis, dysphagia, odynophagia, and the need for narcotic analgesics all are well correlated in patients with head and neck cancer who receive XRT.³ Severe mucositis and the associated decrease in oral intake commonly require hospitalization for dehydration and may necessitate a feeding gastrostomy tube. Data from previous series regarding the effect of smoking on the development of mucositis are variable. In a study of 41 patients with locally advanced head and neck cancer who received XRT, Rugg et al observed a statistically significant correlation between smoking during or after treatment and the development of severe mucositis.¹³ However, in a similar patient population, Browman et al observed no difference in the rate of mucositis between continued smokers and abstainers. An important consequence of mucositis and other acute complications of XRT is interruption of treatment, which can result in a higher risk of persistent and recurrent disease.

In the current study, we observed a significant association between the development of moderate-to-severe mucositis and continued smoking, and patients who continued to smoke more commonly required a feeding gastrostomy tube, likely related to the degree of mucositis these patients suffered. We also noted a statistically significant increase in need for and longer duration of hospitalization in patients who continued to smoke, both including and excluding data from Group 3. It is possible that the increased need for hospitalization was linked to the greater incidence of and less tolerance for the early complications of treatment. Patients who are receiving XRT for laryngopharyngeal cancer also are at risk of developing chronic dysphagia, often as a result of pharyngeal and esophageal stenosis.¹⁴ Furthermore, it is well established that concurrent chemotherapy increases XRT-associated dysphagia compared with XRT alone.^2,3,15 Although a very similar proportion of abstainers and continued smokers received concurrent chemotherapy (59% and 61%, respectively) in the current series, we observed that the rates of feeding tube placement and pharyngeal-esophageal stricture requiring intervention were higher in continued smokers.

Although we noted that continued smokers had a shorter duration of gastrostomy-tube dependence and lower rates of acute XRT-associated skin changes compared with abstainers, these findings were not statistically significant. However, a dramatic early effect of smoking cessation may be improved skin oxygenation, which may allow for greater XRT-associated acute skin toxicity, whereas continued smoking and less skin perfusion may provide some level of protection from acute skin toxicity.

Osteoradionecrosis is a serious late complication of XRT that can lead to intolerable pain, trismus, fistulas, oral dysphagia, and the need for multiple surgical procedures. The pathophysiology of osteoradionecrosis involves reduced vascularity of the oral hard and soft tissues as a result of radiation-induced endarteritis. In addition, it has been demonstrated that XRT reduces the proliferation of periosteal, bone marrow, and endothelial cells.¹⁶ The development of osteoradionecrosis is multifactorial and can be influenced by both treatment-related and patient-related variables. Treatment-related variables that reportedly increase the risk of developing osteoradionecrosis include doses >66 grays, increased total duration of treatment, and increased XRT field size. Patient-related variables include poor dental status, male sex, and increased age,¹⁷ but these variables may be confounded by or associated with smoking in these groups.

Tobacco use has been implicated as a risk factor for the development of osteoradionecrosis; however, this assertion comes primarily from small, descriptive series and anecdotal experiences. A frequently cited study by Kluth et al retrospectively compared 14 patients who developed osteoradionecrosis with 28 controls. Those authors observed that 86% of patients who developed osteoradionecrosis had continued smoking and heavy alcohol use during and after treatment compared with 21% of controls.¹⁸ Similarly, Katsura et al observed an association between smoking after XRT and the development of osteoradionecrosis (P = .011) and suggested that smoking may enhance the occurrence of mandibular hypovascularization after XRT.¹⁹ However, Goldwaser et al observed no increased risk of osteoradionecrosis in patients who continued to smoke.²⁰

Unlike most previous series that examined the complications of XRT in laryngopharyngeal cancer, we excluded former smokers and nonsmokers from the analysis. All patients in the current study were current smokers at the time of enrollment, which likely accounts for the high overall rate of osteoradionecrosis noted in this series (16%). It is noteworthy that, in this study, we demonstrated the benefit of smoking cessation before beginning XRT in the prevention of this complication. There was no difference in comorbidities, smoking pack-year history, alcohol use, or tumor and treatment characteristics between patients who quit smoking before treatment and those who continued to smoke. However, patients who successfully quit before starting XRT and remained abstinent were significantly less likely to develop osteoradionecrosis.

There are several shortcomings associated with this study, including a small sample size, mixed primary tumor sites, and retrospective collection of data on XRT complications. However, all patients in this series were newly diagnosed and identified prospectively, and both groups (abstainers and continued smokers) were similar in the distribution of demographic, exposure, and clinical characteristics. Another limitation of this study was the heterogeneity of treatment, in particular, the possibility of confounding of XRT complication risk by chemotherapy. Although we observed no significant difference in treatment modalities between the groups (Table 2) or in XRT complications between patients who did and did not receive chemotherapy (data not shown), this remains a limitation of this pilot study. Finally, although detailed data on smoking history and current status were collected prospectively as part of an intensive cessation program, some patients (Group 3) had their status defined in a retrospective fashion. These patients were included only in the confirmatory dataset (Table 4) in which the results were very similar.

The objective of the current study was to better define the role of smoking cessation on rates of early and late complications of XRT. Specifically, we sought to examine whether enrollment in an intensive smoking cessation program and abstaining from smoking during XRT would prevent complications of treatment. Although heterogeneity in treatment, mixed tumor sites, and a small sample size made this a hypothesis-generating study only, there were several interesting findings. We observed a significant association between patients who continued to smoke with increased need for hospitalization and increased osteoradionecrosis, as well as a trend toward more severe mucositis. In addition to decreased quality of life during and after treatment, smokers were at increased risk of treatment interruption and decreased response to XRT, given the higher rate of complications. This study emphasizes the importance of smoking cessation in the prevention of XRT complications in patients with laryngopharyngeal cancer. It also highlights the role of intensive tobacco cessation programs in the multidisciplinary management of patients with head and neck cancer.