IMRT for Head and Neck Cancer Aimed at Reducing Dysphagia: Long-term Clinical Outcomes and Dose-Function Relationships

F. Y. Feng,T. H. Lyden,M. J. Haxer,M. Feng,A. G. Sacco,E. Wu,K. A. Vineberg,H. M. Kim,D. B. Chepeha,A. Eisbruch
DOI: https://doi.org/10.1016/j.ijrobp.2008.06.840
2008-01-01
Abstract:We have conducted a prospective trial of IMRT aimed at reducing dysphagia by sparing the important swallowing structures, and previously reported early dose-effect relationships for these structures (IJROBP 2007; 68:1289). We now present long-term tumor control outcomes and dose-function relationships from this trial. 107 patients with stage III-IV oropharyngeal (100) or nasopharyngeal (7) cancer were treated with definitive chemo-IMRT aiming to spare the salivary glands and the portions of the swallowing structures [pharyngeal constrictors (PC), glottic and supraglottic larynx, and esophagus] outside the planning target volumes (PTVs) without underdosing the PTVs. 74 of these patients participated in a longitudinal study of swallowing and underwent videofluoroscopy (VF), patient-reported, and observer-rated (CTCAE) swallowing toxicity scores before and at 3, 12, and 24 months after therapy. At a median f/u of 23 months, actuarial 2-yr locoregional (LR) control is 94%. 6/107 patients (5%) had LR failures as first failure site; all LR failures were within high-risk PTVs and none failed within the spared organs. From pre-therapy to 3 months after therapy, significant (p < 0.05) worsening was observed in most VF parameters (except pharyngeal transit times) and in patient-reported and observer-rated dysphagia. From 3 months to 2 years, the VF parameters remained stable, while almost all patient-reported and observer-rated items improved significantly with many returning to baseline. Mean doses to each swallowing structure was significantly correlated with VF-based aspirations and summary swallow scores at 3 months, and PC mean dose remained significant at both 1 and 2-years post-therapy (p < 0.05); highest correlations were associated with superior PC dose (p < 0.02). Below and above PC V50 = 75%, 7% and 50% of patients, respectively, had VF-based aspiration at 3 months post-therapy, and similar relationships were observed through 2 years. Five patients had VF-based strictures; all had PC V50 >85%. Three patients had aspiration pneumonia after the peri-treatment period; all had PC V50 >90%. Doses to each swallow structure were significantly correlated with grade 2-3 observer-rated dysphagia at 3 months. Similar trends were noted for patient-reported dysphagia. By 1-2 years, grade 2-3 dysphagia was rare, with only 1 patient requiring a feeding tube. IMRT sparing the swallowing structures can be safe without compromising tumor control. The infrequency of aspiration pneumonia and strictures, and lack of increased swallow transit time, suggest a tangible benefit in dysphagia. Among the swallowing structures, doses to the PCs correlated most strongly with dysphagia through two years after therapy.
What problem does this paper attempt to address?