Characteristics Associated With Swallowing Changes After Concurrent Chemotherapy and Radiotherapy in Patients With Head and Neck Cancer

Objective  To define factors that acutely influenced swallowing function prior to and during concurrent chemotherapy and radiotherapy.

Design  A summary score from 1 to 7 (the swallowing performance status scale [SPS]) of oral and pharyngeal impairment, aspiration, and diet, was assigned to each patient study by a single senior speech and swallow pathologist, with higher scores indicating worse swallowing. Generalized linear regression models were formulated to asses the effects of patient factors (performance status, smoking intensity, amount of alcohol ingestion, and age), tumor factors (primary site, T stage, and N stage), and treatment-related factors (radiation dose, use of intensity-modulated radiation therapy, response to induction chemotherapy, postchemoradiotherapy neck dissection, and preprotocol surgery) on the differences between SPS score before and after treatment.

Setting  University hospital tertiary care referral center.

Patients  The study included 95 patients treated under a multiple institution, phase 2 protocol who underwent a videofluorographic oropharyngeal motility (OPM) study to assess swallowing function prior to and within 1 to 2 months after the completion of concurrent chemotherapy and radiotherapy.

Main Outcome Measures  Factors associated with swallowing changes after chemoradiotherapy.

Results  The mean pretreatment and posttreatment OPM scores were 3.09 and 3.77, respectively. Patients with T3 or T4 tumors (odds ratio [OR], 0.38; 95% confidence interval [CI], 0.15-0.95; P = .04) and a performance status of 1 or 2 (OR, 0.37; 95% CI, 0.15-0.91; P = .03) were less likely to have worsening of swallowing after chemoradiotherapy. There was a trend for worse swallowing with increasing age (OR, 1.04; 95% CI, 0.99-1.09; P = .08). Only T stage (T3 or T4) was associated with improved swallowing after treatment (OR, 8.96; 95% CI, 1.9-41.5; P < .001).

Conclusion  In patients undergoing concurrent chemotherapy and radiotherapy, improved swallowing function over baseline is associated with advanced T stage.

Dysphagia is common in patients with advanced head and neck cancer. Studies have demonstrated that 38% of patients present with mild-moderate to moderate-severe swallowing impairment.¹ In addition, patients with hypopharyngeal and laryngeal primary tumors have been found to have worse swallowing at presentation.¹ While previous investigations have documented severity of dysphagia after definitive or adjuvant radiation therapy or concomitant chemoradiotherapy,^2,3 few studies have documented swallowing function changes from baseline in patients treated with concurrent chemotherapy and radiotherapy.

Of the studies that have reported changes in swallowing function from baseline, most have focused on alterations occurring in the swallowing mechanism. New delayed initiation of the swallow trigger, uncoordinated timing of bolus propulsion, opening of the pharyngoesophageal segment, closure of the larynx, tongue base retraction, and vestibular penetration of barium have all been reported following concurrent chemotherapy and radiotherapy.^4-6 Additionally, increase from baseline in time to velopharyngeal and laryngeal closure and decreased oropharyngeal swallow efficiency have been reported.⁷

While identification of changes in the swallowing mechanism are important to aid in therapeutic intervention for patients undergoing concurrent chemotherapy and radiotherapy, so is the identification of factors predicting improvement or decrement in swallowing function. Therefore, in this study, we sought to identify factors that influenced changes in swallowing during a course of concurrent chemotherapy and radiotherapy.

From November 1998 to August 2002, 222 patients with stage III or IV advanced head and neck cancer were treated under a multiple institution, phase 2 protocol of concurrent chemotherapy and radiotherapy. Organ preservation was the primary goal of the treatment protocol. Resection of small tonsillar or oral cavity primary tumors was allowed, as was pretherapy modified radical neck dissections. Two 21-day cycles of induction paclitaxel (135 mg/m² over 3 hours) and carboplatin (area under the curve, 2; given after paclitaxel) were followed by 4 cycles (for resected cases) or 5 cycles (for unresected cases) of concurrent chemotherapy and radiotherapy. Each chemoradiotherapy cycle consisted of 1 weekly dose of paclitaxel (100 mg/m² after the first dose of radiation), continuous infusion fluorouracil (600 mg/m² for 5 days), oral hydroxyurea (500 mg every 12 hours for 6 days), and twice daily radiation therapy (1.5 Gy per fraction with a minimum 6-hour interfraction interval) followed by a 9-day break. The total radiation dose to gross disease was initially 75 Gy, but later decreased to 72 to 75 Gy based on the response to induction chemotherapy. Initially patients were treated with 3-dimensional conformal radiotherapy, but later some patients were treated with intensity-modulated radiotherapy (IMRT). Following the completion of concurrent chemotherapy and radiotherapy, patients with N2 or higher-stage neck disease underwent a planned neck dissection. Details of the protocol have been outlined previously.⁸

Of the 222 patients enrolled, 132 eligible patients were treated at our institution. Our study population consisted of 95 of the 132 who had swallowing function assessed both prior to the initiation of concurrent chemotherapy and radiotherapy and within 1 to 2 months after the completion of chemoradiotherapy. Each assessment consisted of an anterior and lateral videofluorographic OPM study, as described previously.^9,10 Briefly, a recording was made by the speech pathologist while patients swallowed small amounts of liquid barium, paste barium, and a barium-coated cookie. Slow motion, frame-by-frame analysis was used to evaluate oral, pharyngeal, laryngeal, and cricopharyngeal function. In addition, attention was given to the presence and cause of aspiration as well as laryngeal sensitivity, response to therapeutic techniques, and percentage aspiration. Multiple components of swallowing were summarized on the OPM record and recorded in a computerized database storage and reporting system, as described previously.¹ Using this standardized format, interrater variability was low, and intrarater variability nonexistent because all studies were conducted at the University of Chicago Speech and Swallowing Center by the senior speech pathologist (E.M.) for reliable and consistent reporting. A sample OPM report is provided as Figure 1.

Individualized therapy was provided based on the results of the OPM study and directed to specific disorders documented during the videofluorographic procedure. In addition, therapeutic techniques were applied during the OPM study to assess response to maneuvers and change in swallow physiology and aspiration. Patients were provided with base-of-the-tongue and laryngeal motility exercises, which they were encouraged to perform during treatment. A safe oral diet with a variety of food textures based on OPM results was encouraged with the goal of maintaining optimal flexibility and range of motion of pharyngeal structures.

Changes in patient swallowing were defined in 2 ways: (1) worsening of swallowing was defined as a posttherapy swallowing performance scale (SPS) score higher than the pretherapy SPS score; (2) improvement in swallowing was defined as a posttherapy SPS score lower than the pretherapy score. The SPS score is a summary score from 1 to 7 of oral impairment, pharyngeal impairment, aspiration, and diet, with higher scores indicating worse swallowing, as outlined in Table 1. This rating scale is designed to summarize the clinical results of the OPM study, and the score provides an overall clinical picture of swallowing function. Because reliable SPS results depend on skilled clinical interpretation, the score is not an additive or formulaic summary. For each patient, the SPS score prior to and following concurrent chemotherapy and radiotherapy was extracted from the OPM report.

Generalized linear modeling was used to model changes in the log SPS score. Backward stepwise logistic regression (P < .10 threshold) was used to model improved or worsened swallowing. Covariates included patient factors (performance status, smoking intensity, amount of alcohol ingestion, and age), tumor factors (primary site, T stage, and N stage), and treatment-related factors (radiation dose, use of IMRT, response to induction chemotherapy, and preprotocol surgery). Statistical analysis was conducted using Stata 7.0 software (StataCorp LP, College Station, Texas).

Clinicopathologic characteristics are summarized in Table 2, while treatment-related characteristics are summarized in Table 3. Most of the patients had a good performance status (Eastern Cooperative Oncology Group [ECOG] score, 0-1), laryngeal (23%; n = 22) or oropharyngeal primary tumors (52%; n = 49), advanced T stage (57% T3 or T4; n = 54), and advanced nodal disease (75% N2 or N3; n = 54). Furthermore, 92% of patients did not have prior surgery (n = 87), and 80% had a response to induction chemotherapy (n = 76). When compared with patients treated on the same protocol but not included in this analysis, our patients were more likely to have been treated with IMRT (57% [n = 54] vs 34%) (P < .01) and to have an ECOG performance status of 0 (57% [n = 51] vs 24%) (P < .01). Otherwise, there were no statistical differences among patients included and other patients.

In general, swallowing function decreased from baseline in patients after concurrent chemotherapy and radiotherapy. Specifically, 59 patients had a worse SPS score after treatment (60%), 17 had no change in their SPS score (17%), and 19 had an improvement in their SPS score (19%). The mean pretreatment SPS score was 3.09, and the mean posttreatment SPS score was 3.77 (P < .01 by the Wilcoxon test). As listed in Table 4 and illustrated in Figure 2, for 82% of patients whose swallowing changed, for worse or better (n = 78), the change was only 1 to 2 points on the SPS scale. The median and mean change values during chemoradiotherapy were 1 and 1.7 points (range, 1-5 points), respectively. Most patients with laryngeal primary tumors had T3 (14%; n = 3) or T4 (82%; n = 18) tumors. When patients with laryngeal cancer were analyzed separately, the mean pretreatment SPS score was 4.3 (range, 1-7), and the mean posttreatment SPS score was 4.2 (range, 2-7). While 4 patients had no change in their swallowing after chemoradiotherapy, 10 had a decrement, and 8 had improvement in their swallowing. The median and mean changes in swallowing scores were 0 and 0.1 (range, 0-5), respectively.

When factors associated with worsening of swallowing were analyzed, only a trend for worse swallowing with increasing age was found (odds ratio [OR], 1.04; 95% confidence interval [CI], 0.99-1.09) (P = .08). Surprisingly, patients with T3 or T4 tumors (OR, 0.38; 95% CI, 0.15-0.95) (P = .04) and an ECOG performance status of 1 to 2 (OR, 0.37; 95% CI, 0.15-0.91) (P = .03) were less likely to have worse swallowing after a course of concurrent chemotherapy and radiotherapy. Only advanced T stage (T3 or T4 tumors) was associated with improved swallowing after chemoradiotherapy (OR, 8.96; 95% CI, 1.6-41.6) (P < .001).

A total of 51 patients had gastrostomy tubes placed before, during, or within 6 months of the completion of concurrent chemotherapy and radiotherapy (54%). This included 11 patients with gastrostomy tube placement prior to the initiation of protocol treatment (12%), 36 patients with gastrostomy tubes placed during the course of concurrent chemotherapy and radiotherapy (38%), and 4 had gastrostomy tube placement within 6 months of completing concurrent chemotherapy and radiotherapy (4.2%). At final follow-up, 12 patients had gastrostomy tubes in place (13%), and 77 patients were free of gastrostomy tubes (81%). In the remaining 6 patients, information on gastrostomy tube status was not available.

Multiple studies have demonstrated improved locoregional control and in some cases survival with definitive or adjuvant chemoradiotherapy over radiotherapy.^11-14 As control and survival continue to improve in patients with advanced head and neck cancer, functional, cosmetic, and quality-of-life issues become more important. While control is improved with concurrent chemotherapy and radiotherapy, rates of dysphagia are also higher, as seen in the report of Radiation Therapy Oncology Group (RTOG) 91-11¹¹ in which 23% of patients treated with concomitant cisplatin and radiotherapy were able to swallow only soft foods compared with 9% of patients receiving sequential cisplatin/fluorouracil irradiation and 15% of patients receiving radiotherapy alone.

Quality-of-life studies demonstrate that for patients with advanced head and neck cancer, difficulties and embarrassment caused by eating in public, distorted speech, hoarseness, and mouth pain are important predictors of overall quality of life.¹⁵ Furthermore, after “being cured of my cancer” and “living as long as possible,” patients prioritize “being able to swallow all foods and liquids” only behind “having no pain,” “returning to regular activities,” and “having a normal amount of energy.” Given the high level of importance that patients place on swallowing, we sought to determine in this study the factors that could predict detrimental or improved swallowing outcomes following a course of chemoradiotherapy.

One of the interesting findings from our study population was the preservation of function in patients with advanced laryngeal cancer. Recent reports demonstrate that patients with larynx cancer treated with concurrent chemotherapy and radiotherapy had improved quality of life over surgically treated patients and that patients with an intact larynx were more likely to obtain nutrition orally without supplements.¹⁶ Furthermore, patients with advanced larynx cancer treated with concurrent chemotherapy and radiotherapy were found to have intelligible communication and efficient swallowing.⁶ Our analysis demonstrated that patients with laryngeal cancer had a worse mean baseline swallowing score than the entire study population. However, after treatment, these patients had little change in their swallowing score compared with a mean decrement of 0.78 in the population as a whole. These results indicate that patients with advanced laryngeal cancer treated with concurrent chemotherapy and radiotherapy may be able to preserve swallowing function.

Perhaps the most interesting finding from this analysis is that patients with more advanced T stages were less likely to have worsened swallowing and more likely to have improved swallowing after a course of concurrent chemotherapy and radiotherapy. To our knowledge, this has not been reported in the literature. Our group's prior investigations with pretherapy swallowing studies revealed no association between T stage and aspiration status, cervical esophageal impairment, pharyngeal impairment, oral impairment, and SPS score.¹ Other investigators have not found an association between T stage and dysphagia, but the population studied was heterogeneous and included patients treated with definitive and adjuvant radiation and chemoradiation.^2,17 Our finding that patients with advanced T stages were less likely to have worsening of their swallowing and more likely to have improved swallowing will aid practitioners in counseling patients as to expectations during treatment. However, it must be noted that patients were strongly encouraged to maintain oral intake as long as possible during treatment and did not undergo routine percutaneous endoscopic gastrostomy tube placement. This strategy was used to encourage patients to exercise the swallowing mechanism throughout treatment.

Our investigation found no correlation between the use of IMRT and swallowing dysfunction. However, this is not surprising, since our main goal of IMRT was to decrease the dose to the parotid glands, skin, larynx, and oral cavity, while no attempt was made to spare the pharyngeal constrictors owing to their close proximity to the retropharyngeal nodal region. Other investigators have found that radiation dose to the upper, middle, and lower pharyngeal constrictors and the supraglottic and glottic larynx have been associated with swallowing dysfunction.¹⁸ Preliminary results from patients treated prospectively with IMRT to spare these structures demonstrated statistically significant correlations between aspiration risk and partial volumes of the pharyngeal constrictors and glottic and supraglottic larynx receiving radiation doses between 50 and 65 Gy. Furthermore, dose to the pharyngeal constrictors was associated with stricture risk and worsening of liquid and solid swallowing.¹⁹ We are currently investigating the interactions of chemotherapy and IMRT in regard to swallowing function, neck fibrosis, and saliva flow. Patients analyzed in the present study were more likely to have been treated with IMRT, which may enhance the applicability of these results because 76% to 82% of practitioners are currently using IMRT to treat patients with head and neck cancer.²⁰

These data indicate that 54% of patients required gastrostomy tube nutritional assistance during the course of or shortly after concurrent chemotherapy and radiotherapy (n = 51). These numbers are consistent with the entire protocol cohort from which these patients were drawn in which 62% of patients had gastrostomy tubes placed prior to (26%) or during (36%) the course of chemoradiotherapy. However, at final follow-up, only 13% had gastrostomy tubes in place (n = 12). This is probably owing to many causes, the first of which is that patients with gastrostomy tubes in place are more likely to have multiple swallowing assessments and therefore would have data making them eligible for this analysis. Furthermore, mucosal healing and swallowing therapy enable patients to regain swallowing function.

Our study is limited by the retrospective nature of the analysis as well as the short follow-up period. The goal of this study was only to assess what factors were associated with swallowing changes during a course of chemoradiotherapy. Swallowing function for these patients will continue to evolve with longer follow-up. Furthermore, the tools used to assess swallowing limit the analysis. The SPS itself has not been validated by statistical analysis; rather, it is a quick, clinically relevant scoring system for patients with head and neck cancer and a simple, single-step technique for classifying swallowing functional status. While other more specific measures are available to score videofluorographic swallowing test results (eg, oropharyngeal swallow efficiency), these methods are extremely labor intensive and are best suited for the laboratory setting. By design, our study was only intended to assess acute swallowing changes of concurrent chemotherapy and radiotherapy. Further investigation will be needed to determine long-term swallowing changes. However, all the patient data was collected prospectively by a single senior speech pathologist (E.M.). Furthermore, all patients described in this study were managed uniformly with organ-preserving intent by an experienced multidisciplinary team of head and neck surgeons, medical oncologists, and a single radiation oncologist (D.J.H.). The value of the SPS might have been more limited if it had been administered by multiple raters, with subjective scoring changes leading to interobserver variability. However, because all of the OPM findings were interpreted by a single senior speech pathologist, our data were obtained without interobserver variability. To validate these conclusions, we plan on expanding our data set with patients treated at our institution with similar chemoradiotherapy regimens. In addition, swallowing function continues to change over time in patients treated with chemoradiotherapy for head and neck cancer. Therefore, we plan to determine which factors are associated with chronic swallowing function changes and esophageal stricture.

In conclusion, we found that patients with advanced T stages (T3 and T4) and worse performance status (ECOG 1 and 2) were less likely to have worsening of swallowing during a course of concurrent chemotherapy and radiotherapy. Additionally, patients with T3 or T4 tumors were more likely to have an improvement in swallowing function after the completion of their treatment. Patients with advanced laryngeal tumors, while initially presenting with worse swallowing function, had less of a decrement in swallowing function after chemoradiotherapy.

Correspondence: Joseph K. Salama, MD, Department of Radiation and Cellular Oncology, 5758 S Maryland Ave, MC 9006, Chicago, IL 60637 (jsalama@radonc.uchicago.edu).

Submitted for Publication: October 15, 2006; final revision received October 24, 2007; accepted October 30, 2007.

Author Contributions: Dr Salama had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Salama, List, Mell, Blair, Vokes, and Haraf. Acquisition of data: Salama, Stenson, Mell, MacCracken, Cohen, Blair, and Vokes. Analysis and interpretation of data: Salama, Stenson, List, Mell, Cohen, and Haraf. Drafting of the manuscript: Salama, Mell, Vokes, and Haraf. Critical revision of the manuscript for important intellectual content: Salama, Stenson, List, Mell, MacCracken, Cohen, Blair, Vokes, and Haraf. Statistical analysis: Mell. Administrative, technical, and material support: Stenson, List, MacCracken, Cohen, Blair, Vokes, and Haraf. Study supervision: Salama, Stenson, List, Cohen, and Haraf.

Financial Disclosure: None reported.

Previous Presentation: This article was presented at the 2006 American Head and Neck Society Annual Meeting & Research Workshop; August 18, 2006; Chicago, Illinois.