Reduced mouth opening in patients with head and neck cancer treated with radiation therapy: an analysis of the Clinical Registry of Dental Outcomes in Head and Neck Cancer Patients (OraRad)

Abstract

Objectives:

Trismus/reduced mouth opening (RMO) is a common side effect of radiotherapy (RT) for head and neck cancer (HNC). The objective was to measure RMO, identify risk factors for RMO, and determine its impact on quality of life (QOL).

Patients and Methods:

OraRad is an observational, prospective multicenter cohort study of patients receiving curative intent RT for HNC. Interincisal mouth opening measurements (n = 565) and patient reported outcomes were recorded before RT and every 6 months for 2 years. Linear mixed-effects models were used to evaluate change in mouth opening and assess the relationship between trismus history and change in QOL measures.

Results:

Inter-incisal distance decreased from a mean (SE) of 45.1 (0.42) mm at baseline to 42.2 (0.44) at 6-months with slight recovery at 18-months (43.3, 0.46 mm), but no additional improvement by 24-months. The odds of trismus (opening <35mm) were significantly higher at 6 months (OR=2.21, 95% CI: 1.30 to 3.76) and 12 months (OR=1.87, 95% CI: 1.08 to 3.25) compared to baseline. Females were more likely to experience trismus at baseline and during follow up (p<0.01). Patients with oral cavity cancer had highest risk for trismus at baseline and post-RT (p<0.01). RMO was associated with higher RT dose to primary site and receiving concomitant chemotherapy (p<0.01). Trismus was associated with self-reported difficulty opening the mouth and dry mouth (p<0.01).

Conclusions:

Decrease in mouth opening is a common treatment-related toxicity after RT with some recovery by 18-months. Trismus has a significant impact on survivor QOL.

Introduction:

Head and neck cancer (HNC) accounts for approximately 3% of all new cancers in the United States affecting an estimated 66,470 individuals and accounting for over 15,050 deaths in 2022.⁽¹⁾ The overall 5-year survival rate of HNC is approximately 65%. ⁽²⁾ Radiation therapy (RT) is a primary or adjuvant treatment, often delivered with concurrent chemotherapy (CT) and surgery, for most HNC patients.⁽³⁾⁽⁴⁾ Although there have been advances in RT treatment planning and delivery, various toxicities associated with RT remain common and associated with significant morbidity. The head and neck region including the oral cavity is at risk for oral mucositis, dry mouth, dental disease, dysgeusia, osteoradionecrosis, and trismus. These toxicities negatively affect the quality of life (QOL) of a HNC patient both during and after treatment.⁽⁵⁾

Trismus has been defined as a restriction in the range of motion of the jaws and is characterized by limited mouth opening.⁽⁶⁾ Criteria for trismus has been studied in the HNC population with an established functional cut off point of <35mm. ⁽⁷⁾ The self-reported prevalence of trismus among oropharyngeal cancer survivors (n=872) was 31% during a mean time of 7 years of follow up, and was associated with diminished QOL.⁸⁾

The Clinical Registry of Dental Outcomes in Head and Neck Cancer Patients (OraRad) (NIDCR Grant 1U01DE-22939-01; NCT02057510) is an observational study which prospectively followed patients with HNC before RT through two years after RT.⁽⁹⁾ This report describes the change in mouth opening in this population after RT, analyzes risk factors associated with decreased mouth opening, and the relationship between decreased mouth opening and QOL.

Patients and Methods:

Study Cohort:

OraRad is a longitudinal multicenter study of dental and oral outcomes in HNC patients treated with definitive or postoperative RT. ⁽⁹⁾ Institutional Review Board approval was obtained at all sites and informed consent was obtained from every participant. (See Supplemental Materials Table S1 for approval numbers) OraRad enrolled 572 patients from April 2014 to October 2018. Subject eligibility requirements included: age≥ 18; diagnosed with head and neck squamous cell carcinoma (SCC) or salivary gland cancer (SGC) with the intent to receive external beam RT with curative intent (tumor eradication), or patients with a non-SCC, non-SGC malignancy of the head and neck region who had planned to receive at least 4500 cGy RT to the head and neck region; and had no prior RT to the head and neck region. (See Supplemental Materials Table S2 for complete inclusion and exclusion criteria.)

Study Assessments:

Baseline assessments (baseline, V0) occurred after a pre-radiation dental evaluation and any pre-RT extractions but before initiating head and neck RT. Patients were evaluated at 6-month intervals through 24-months post-RT (V06, V12, V18, V24). Participants self-reported sociodemographic information at baseline including race, ethnicity, education, dental insurance, public assistance and sex (See Table 1). Smoking status was also self-reported at baseline. Each visit consisted of self-reported assessments, including the oral pain score and the oral health quality of life using the EORTC QLQ H&N 35 scale, and a comprehensive oral and dental examination, including measurement of mouth opening, performed by a calibrated examiner. Details of the clinical protocol have been previously reported. ⁽⁹⁾

Table 1:

Baseline Clinical and Demographic Characteristics of the Patients. N(%) reported unless otherwise noted.

Characteristic	Study Population (N=565)
Sex
Female	129 (22.8%)
Male	436 (77.2%)
Age (median [range])	59.0 [21.0, 97.0]
Race
Black Only	44 (7.8%)
Other	53 (9.4%)
White Only	468 (82.8%)
Ethnicity
Hispanic	29 (5.1%)
Not Hispanic	536 (94.9%)
Education level
≤ High school	155 (27.4%)
> High school	408 (72.2%)
Declined to answer	2 (0.4%)
Smoking status
Former or current smoker	309 (54.7%)
Never used	254 (45.0%)
Declined to Answer	2 (0.4%)
Dental insurance
Does not have	204 (36.1%)
Has	361 (63.9%)
Receiving public assistance
No	517 (91.5%)
Yes	48 (8.5%)
Type of Cancer
Other	37 (6.5%)
Squamous Cell Carcinoma	462 (81.8%)
Salivary Gland Cancer	66 (11.7%)
Primary Tumor Site
Larynx/Hypopharynx	38 (6.7%)
Oral Cavity	85 (15.0%)
Oropharynx	268 (47.4%)
Other	86 (15.2%)
Salivary Gland	55 (9.7%)
Unknown	33 (5.8%)
T category (AJCC 7th edition)
1/2	340 (60.2%)
3/4	177 (31.3%)
Other	48 (8.5%)
N category (AJCC 7th edition)
0	135 (23.9%)
≥1	423 (74.9%)
Other	7 (1.2%)
Metastases (M):
M0	531 (94.0%)
M1	9 (1.6%)
MX	25 (4.4%)
Total RT dose to primary site (cGy)
Mean ± SD	6579.3 ± 649.1
<6000	33 (5.8%)
≥6000 to <7000	291 (51.5%)
>7000	240 (42.5%)
Unknown	1 (0.2%)
Treatment
RT only	39 ( 6.9%)
RT+CT	218 (38.6%)
Sur+RT	165 (29.2%)
Sur+RT+CT	143 (25.3%)
RT treatment to neck	535 (94.7%)
Bilateral	428 (80%)
Unilateral	107 (20%)
Inter-incisal distance measure
Alveolar ridge-to-alveolar ridge	10 (1.8%)
Tooth-to-alveolar ridge	30 (5.3%)
Tooth-to-tooth	525 (92.9%)

RT: radiation therapy; CT: chemotherapy; Sur: surgery (prior to RT).

Medical records were reviewed to record cancer diagnosis, site, staging (T, N), and management including dose, surgery, and chemotherapy. Primary tumor site was classified into the following 5 anatomical areas: oropharynx (base of tongue, tonsil, oropharynx, or soft palate); oral cavity (oral tongue, oral cavity, gingival/alveolar ridge, mandible, buccal/labial mucosa, floor of mouth, maxilla, retromolar trigone, hard palate, or lip); larynx/hypopharynx (larynx, hypopharynx, or epiglottis); salivary gland (submandibular gland, parotid, or sublingual gland); and other (neck, nasopharynx, pharynx, maxillary sinus, nasal cavity or other sites). Oral cavity cancer that included multiple contiguous subsites was listed as “oral cavity” as an inclusive subsite.

To assess changes in trismus, maximal mouth opening was measured using a Therabite^® Range of Motion Scale and recorded in millimeters. The maxillary incisal tooth-to-mandibular incisal tooth distance was measured, preferably from the upper right central and lower right central incisors. Tooth-tooth distance was also used for patients with a partial denture replacing incisal teeth in one arch. For patients missing anterior teeth in one arch not replaced by a partial denture, a tooth-to-alveolar ridge measurement was recorded. For edentulous patients, an alveolar ridge-to-alveolar ridge measurement was recorded. Follow-up opening measurements were restricted to the same type of distance measurement. Trismus was defined as a mouth opening distance <35mm regardless of measurement type; trismus during follow-up was defined as having a mouth open distance <35mm during at least one follow-up visit.

Patients were administered questions corresponding to RT-specific scales of the EORTC QLQ-H&N35 questionnaire at baseline, and then every 6 months post-RT for 2 years. The chosen questions evaluate the extent (four-point scale ranging from “not at all” to “very much”) to which participants experienced problems with their oral health during the past week. Scores were transformed into 0-to-100 scales, with higher scores representing higher level of symptoms. The specific aspects of oral health included the following scales: swallowing (problems swallowing liquids, pureed food, and solid foods and choking when swallowing), senses problems (problems with sense of smell and taste), teeth, mouth opening, dry mouth, and sticky saliva. A ten unit or more change has been rated as clinically meaningful. ⁽¹⁰⁾ An Academic User Agreement was obtained for use of specific scales from the EORTC QLQ-H&N35 in the OraRad study prior to administration. ⁽¹⁰⁾

Analysis and Statistical Considerations

Risk factors were assessed at baseline and included: sex, smoking status (former/current vs. never smoked), education (at most high school diploma vs. greater than high school), dental insurance, public assistance, tumor stage (T, N), primary tumor site, dose to primary site (treated as continuous and as categorical defined by quartiles), and treatment modalities (defined based on presence or absence of CT and surgery).

Linear mixed-effects models were used to evaluate change in mouth opening distance across study time points. Associations between risk factors and change in distance were evaluated by testing interaction terms between study visits (treated as categorical) and the risk factor. If the interaction term was not significant (omnibus interaction p-value>0.05), it was removed from the model.

Change in the odds of trismus (defined as a mouth opening distance <35 mm) was assessed using a binomial mixed-effects model with 50 gaussian quadrature points. Logistic regression models were used to evaluate the relationship between risk factors and the odds of trismus during any follow-up visit after adjusting for baseline trismus. The relationship between discrete risk factors and baseline trismus was evaluated using Fisher’s exact test with p-values computed by Monte Carlo simulation. Wilcoxon rank sum or Kruskal-Wallis tests were used to compare continuous characteristics between individuals based on baseline trismus.

Linear mixed-effects models were used to assess the relationship between trismus history and change in QOL measures. Trismus history was classified into four categories based on the presence or absence of trismus at baseline and at any point during follow-up.

Subject specific random intercepts were included in all mixed effects models. The supplementary appendix gives results for all models considered. Trends were visualized using locally estimated scatterplot smoothing (loess) curves. ⁽¹¹⁾

Analyses were conducted using R version 4.2.0 and versions 1.1.29, 3.1.3, and 1.7.4.1 of the “lme4”, “lmerTest”, and “emmeans”, and “tableone” packages, respectively. ^{(12) (13) (14)} All p-values are two-sided and have not been adjusted for multiple comparisons.

Results:

Patient Demographics

A total of 565 patients were enrolled that had mouth opening distance measured, with a median age of 59 (range 21 to 97) years, and the majority being male (77.2%). The cohort was predominately white (82.8%) and non-Hispanic (94.9%). SCC was the most common pathology (81.8%), and oropharynx the most common primary site for RT (47.4%). Although most patients (60.2%) presented with early primary tumor size/extent (T1/T2) disease, nodal involvement was also common (74.9%). The mean (standard deviation) total RT dose delivered to the primary site was 6579.3 (649.2) cGy, with 94.0% of the cohort receiving 6000 cGy or greater dose to primary site. While 6.9% of the cohort had RT alone, 36.8% of patients additionally received CT, 29.2% had RT+surgery, and 25.3% had RT+CT+surgery. Radiotherapy fields often included the nodal regions of the neck (94.7%; bilateral neck RT in 80%; Table 1).

Mouth Opening Distance/Trismus

Most participants (525/565, 92.9%) had a tooth-to-tooth distance measured at baseline. Thirty individuals had a tooth-to-alveolar ridge measurement and ten had an alveolar ridge-to-alveolar ridge measurement at baseline. There were significant changes in inter-incisal distance during the study (omnibus p-value<0.001) with the estimated average distance deceasing from 45.1 (SE=0.42) mm at baseline to 42.2 (SE=0.44) mm at the six-month visit, slight recovery at 12 (42.9, SE=0.45 mm) and 18-months (43.3, SE=0.46 mm), but no additional improvement by 24-months (43.2, SE=0.46 mm). Furthermore, when inter-incisal distance is classified into trismus (<35mm) vs. normal (≥35mm), the odds of trismus was significantly higher at 6 months (OR=2.21, 95% CI: 1.30 to 3.76) and 12 months (OR=1.87, 95% CI: 1.08 to 3.25) compared to baseline (Figure 1, Table 2, Supplementary Table S3).

Figure 1:

Model estimated opening distance at each visit (panel a), and percentage of individuals with trismus (classified as opening distance <35mm) at each visit (panel b).

Table 2.

Change in opening distance in millimeters from baseline to each visit and difference in change based on treatment characteristics. Estimate (95% CI); p-values are presented.

	Change in mouth opening distance	Total RT dose to primary site	Treatment
		per 100 cGy greater dose	RT only vs. RT+CT	RT only vs. Sur+RT	RT only vs. Sur+RT+CT	RT+CT vs. Sur+RT	RT+CT vs. Sur+RT+CT	Sur+RT vs. Sur+RT+CT
p-value	≤0.0001	≤0.0001	0.0073
V06 vs. BL	−2.83 (−3.45, −2.2); ≤0.0001	−0.19 (−0.31, −0.08); 0.0006	1.62 (−0.94, 4.19); 0.2149	−0.84 (−3.45, 1.77); 0.5297	1.85 (−0.8, 4.51); 0.1715	−2.46 (−3.98, −0.94); 0.0015	0.23 (−1.36, 1.83); 0.7751	2.69 (1.02, 4.36); 0.0016
V12 vs. BL	−2.12 (−2.76, −1.47); ≤0.0001	−0.17 (−0.27, −0.06); 0.0018	1.66 (−0.97, 4.29); 0.2149	−0.63 (−3.31, 2.06); 0.6471	1.77 (−0.98, 4.52); 0.2072	−2.29 (−3.85, −0.72); 0.0042	0.11 (−1.57, 1.78); 0.9003	2.40 (0.63, 4.16); 0.0077
V18 vs. BL	−1.74 (−2.41, −1.06); ≤0.0001	−0.23 (−0.33, −0.12); ≤0.0001	3.32 (0.62, 6.02); 0.0159	0.54 (−2.23, 3.30); 0.7030	2.59 (−0.24, 5.42); 0.0728	−2.79 (−4.42, −1.16); 0.0008	−0.74 (−2.47, 1.00); 0.4067	2.05 (0.22, 3.88); 0.0284
V24 vs. BL	−1.83 (−2.49, −1.17); ≤0.0001	−0.22 (−0.32, −0.11); ≤0.0001	1.41 (−1.25, 4.08); 0.2992	−1.71 (−4.42, 1.00); 0.2167	0.73 (−2.07, 3.52); 0.6108	−3.12 (−4.71, −1.53); ≤0.0001	−0.69 (−2.41, 1.04); 0.4361	2.43 (0.64, 4.23); 0.0079

RT: radiation therapy; CT: chemotherapy; Sur: surgery.

Risk Factors Associated with Trismus

Sex

There was not a strong association between sex and change in inter-incisal distance during follow-up (omnibus p-value=0.81; Supplementary Table S4). Males were estimated to have a 5.87 (95% CI: 4.10 to 7.63) mm greater opening distance than females across the study visits. Males were significantly less likely to experience trismus (defined as <35mm) at baseline (11%, vs 24%, p-value=0.0004) and during follow-up (OR: 0.45, 95% CI: 0.27 to 0.76) compared to females (Table 3).

Table 3:

Relationship between risk factors and trismus (defined as <35mm) at baseline and during follow-up.

	BLNORM/ FUNORM	BLNORM/ FUTRIS	BLTRIS/ FUNORM	BLTRIS/ FUTRIS	BL Trismus* p- value	FU Trismus p-value
Sex					0.0004	0.0031
Female	55 (51%)	23 (21%)	7 (7%)	22 (21%)
Male	305 (77%)	47 (12%)	11 (3%)	32 (8%)
Smoking status					0.3284	0.0281
Former or current smoker	202 (75%)	28 (10%)	10 (4%)	31 (11%)
Never used	157 (68%)	42 (18%)	8 (3%)	23 (10%)
T category					0.0001	0.0361
1 or 2	238 (77%)	38 (12%)	11 (4%)	23 (7%)
3 or 4	91 (60%)	24 (16%)	7 (5%)	29 (19%)
N category					0.2548	0.0641
0	75 (64%)	23 (19%)	5 (4%)	15 (13%)
≥1	279 (74%)	47 (12%)	13 (3%)	39 (10%)
Primary Tumor Site					0.0005	0.0027
Larynx/Hypopharynx	27 (87%)	2 (6%)	1 (3%)	1 (3%)
Oral Cavity	30 (45%)	15 (22%)	3 (4%)	19 (28%)
Oropharynx	195 (78%)	30 (12%)	8 (3%)	16 (6%)
Other	58 (77%)	8 (11%)	2 (3%)	7 (9%)
Salivary Gland	28 (55%)	9 (18%)	4 (8%)	10 (20%)
Treatment					0.0255	0.6735
RT only	31 (89%)	3 (9%)	0 (0%)	1 (3%)
RT+CT	138 (73%)	30 (16%)	5 (3%)	16 (8%)
Sur+RT	100 (66%)	21 (14%)	9 (6%)	21 (14%)
Sur+RT+CT	91 (72%)	16 (13%)	4 (3%)	16 (13%)

BLNORM/FUNORM: mouth opening distance ≥35mm during baseline and at all attended follow-up visits; BLNORM/FUNORM: mouth opening distance ≥35mm at baseline, but <35mm during at least one follow-up visit; BLTRIS/FUNORM: mouth opening distance <35mm at baseline, but ≥35mm at all attended follow-up visits; BLTRIS/FUTRIS: mouth opening distance <35mm at baseline and <35mm during at least one follow-up visit.

^*Individuals who had opening distance measured at baseline but not during follow-up were included in this analysis.

Sociodemographic characteristics

There was not a significant association between education level, having dental insurance, or being on public assistance and change in distance or having trismus at baseline or during follow-up (Supplementary Table S4). There was not a strong association between smoking and change in interincisal opening distance during follow-up (p-value=0.59) or baseline opening distance (p-value=0.72). While baseline trismus rates were comparable between former and current smokers compared to individuals who never smoked (15% vs. 12%, p-value=0.33), individuals who never smoked were more likely to experience trismus during follow-up (OR: 1.69, 95% CI: 1.06 to 2.71, Table 3)

Cancer characteristics

Cancer Stage

There was not a strong association between tumor size and change in opening distance during follow-up (omnibus p-value=0.64; Supplementary Table S4). Individuals with a T1/T2 were estimated to have a 3.07 (95% CI: 1.42 to 4.73) mm greater opening distance than individuals with a T3/T4 across the study visits. Individuals with a T3/T4, were significantly more likely to experience trismus at baseline (23% vs. 10% p-value=0.0001) and during follow-up (OR: 1.72, 95% CI: 1.04 to 2.83) compared to individuals with a T1/T2 (Table 3).

There was not a significant association between nodal involvement and change in distance during follow-up (omnibus p-value=0.92; Supplementary Table S4) or opening distance across the study visits (p-value=0.12). While the proportion of individuals with trismus at baseline was comparable between those with N0 vs. N≥1 (17% vs. 13%, p-value=0.25), there was a borderline statistically significant relationship between nodal involvement and trismus during follow-up (OR: 0.61, 95% CI: 0.36 to 1.03 for N≥1 vs N0; Table 3).

Primary Tumor Site

There was not a significant relationship between tumor site and change in distance during follow-up (omnibus p-value=0.75; Supplementary Table S4). There were significant differences in baseline opening distances between the groups defined by primary cancer site (p-value≤0.001; Supplementary Table S6). Patients with a primary tumor site of larynx/hypopharynx had the largest opening distance at baseline (48.8 mm, SE=1.41) which was significantly greater than the distance for oral cavity (38.8mm, SE=0.96) and salivary gland (40.8mm, SE=1.16).

Type of interincisal distance significantly varied between groups (p=0.018) with individuals who had a primary tumor site of larynx/hypopharynx being the least likely to have a tooth-to-tooth measurement (82%) and oropharynx and salivary gland the most likely (96% in both groups; Supplementary Table S5).

There was a significant relationship between primary tumor site and both baseline (p-value=0.0005) and follow-up trismus (p-value=0.003), with patients who had a primary site of oral cavity the most likely to have trismus at baseline (28%) or during follow-up (51%) and those who had a primary site of larynx/hypopharynx the least likely to have trismus at baseline (8%) or during follow-up (10%; Table 3, Supplementary Table S6).

Treatment characteristics

Dose to Primary Site

Dose to the primary site was significantly associated with change in mouth opening distance (p-value<0.001; Table 2). For each additional 100 cGy dose to the primary site, mouth opening distance is estimated to decrease by 0.19 (SE=0.06) mm from baseline to the 6-month visit, 0.17 (SE=0.05) mm from baseline to the 12-month visit, 0.23 (SE=0.05) mm from baseline to the 18-month visit, and 0.22 (SE=0.05) mm from baseline to the 24-month visit. Dose was also classified into quartiles (quartile 1 ≤6000 cGy, quartile 2 ≤6600 but >6000 cGy, quartile 3 ≤ 7000 but >6600 cGy and quartile 4 >7000 cGy) to evaluate non-linear trends. Quartile of dose was significantly associated with change in distance (p-value<0.001, Figure 2; Supplementary Table S7). Individuals with doses classified into quartile 2, 3, and 4 experienced greater decreases in mouth opening distance compared to individuals classified into quartile 1. Individuals classified into quartile 2 experienced borderline significantly less of a decrease in mouth opening distance from baseline to the 6-, 18- and 24-month visit compared to quartile 3, but there was no significant difference in change in distance between individuals classified into quartile 2 and 4 across study visits. There also was no significant difference in change in distance across study visits between individuals classified into quartile 3 and 4.

Figure 2:

Loess curve for opening distance (mm) by quartile of radiation therapy dose. Visit month has been jittered for visualization purposes. Quartile 1= RT dose≤ 6000 cGy; Quartile 2= RT dose 6000 to ≤6600; Quartile 3= RT dose 6600 to ≤7000; Quartile 4=RT dose >7000.

There was not a significant relationship between dose and baseline trismus (p-value=0.35), but there was a borderline statistically significant relationship between dose and follow-up trismus (OR=1.04, 95% CI: 1.00 to 1.09 for each 100cGy higher dose). Evaluating dose by quartiles, quartile 1 was the least likely to experience trismus during follow-up (18%), compared to around 30% of individuals in the remaining 3 quartiles (Table 3).

Treatment modality

There was a significant relationship between treatment modality and change in opening distance (omnibus p-value=0.0073; Figure 3; Table 2). Patients who received RT+CT experienced a greater decrease in mouth opening distance from baseline to all subsequent visits compared to those who received RT+surgery and from baseline to 18-months compared to individuals who had RT alone. Patients who received RT+CT+surgery experienced a greater decrease in distance from baseline to all subsequent visit compared to individuals who had RT+surgery.

Figure 3:

Loess curve for opening distance (mm) by treatment. Visit month has been jittered for visualization purposes. RT=Radiation therapy, CT=chemotherapy, and Sur=surgery

There was a significant relationship between treatment and baseline trismus (p-value=0.026), but not follow-up trismus (p-value=0.67). Individuals with RT alone were least likely to have trismus at baseline (3%) and individuals with RT+surgery were most likely (19%); Table 3) However individuals who received RT were least likely to have a tooth-to-tooth measurement (87%) compared to the other treatment modalities (92-94%; p=0.0435; Supplementary Table S5).

Quality of Life Associated with Trismus

Trismus history was associated with changes in difficulty opening mouth wide (p-value<0.0001) and dry mouth (p-value=0.0002), Supplementary Table S8-S9; Figure 4).

Figure 4:

Loess curve for selected EORTC scales by trismus history. Visit month has been jittered for visualization purposes. Statistical comparisons are provided in Supplementary Tables S6 and S7. BLNORM-FUNORM: mouth opening distance ≥35mm during baseline and at all attended follow-up visits; BLNORM-FUNORM: mouth opening distance ≥35mm at baseline, but <35mm during at least one follow-up visit; BLTRIS-FUNORM: mouth opening distance <35mm at baseline, but ≥35mm at all attended follow-up visits; BLTRIS-FUTRIS: mouth opening distance <35mm at baseline and <35mm during at least one follow-up visit.

Patients who had trismus at baseline but not during follow-up experienced a clinically meaningful (≥ 10 point) decrease in difficulty opening mouth wide and less of an increase in dry mouth from baseline to subsequent follow-up visits compared to individuals who had normal mouth opening distance at baseline. They additionally experienced a decrease in difficulty opening mouth wide (baseline to 6- and 24-months) and less of an increase in dry mouth (baseline to 12-, 18-, and 24-months) compared to individuals with trismus at baseline and during follow-up (Supplementary Table S9, Figure 4).

Patients who had normal mouth opening distance at baseline, but experienced trismus during follow-up experienced a clinically meaningful (≥ 10 point) increase in problems opening mouth from baseline to all subsequent visits compared to all other groups of individuals categorized by trismus history. They additionally experienced an increase in dry mouth (baseline to 18-months) compared to individuals who had trismus at baseline and during follow-up (Supplementary Table S9, Figure 4).

Individuals who had trismus at baseline and during follow-up experienced decreases in self-reported difficulty opening mouth (baseline to all subsequent visits) and less of an increase in dry mouth (baseline to 6-months) compared to individuals with normal mouth opening distance at all visits. (Supplementary Table S9, Figure 4).

Discussion:

The OraRad study provided insights into the impact of RT HNC treatment and adjunctive modalities on mouth opening or trismus and associated oral health related QOL to 24 months post RT. Both a significant decrease in interincisal distance and increase in trismus was noted at 6 months with a slight recovery up to 18 months but no additional recovery at 24 months. This finding is similar to other studies although variability in recovery from restricted mouth opening, is likely related to specific patient, cancer, or treatment characteristics. ^{(15) (16) (17)}

Males had greater opening at all time points, but there was not a significantly different change in interincisal distance based on sex. Females were more likely to experience trismus at baseline and during follow up likely due to starting with a smaller mouth opening distance at baseline. There have been inconsistent findings regarding risk of trismus based on sex in patients who received RT as well as those who never received RT. ^{(15) (18) (19) (20) (21) (22)}

Characteristics of the cancer were associated with changes in jaw opening. Patients with a smaller local tumor volume (T1/T2 tumor) were less likely to experience trismus compared to patients with a larger tumor volume/extent (T3/T4 tumor). Interestingly, patients without regional lymph node involvement were moderately more likely to have trismus at follow up compared to individuals with regional lymph node involvement. However, patients with no regional lymph node involvement were more likely to have an oral cavity or salivary gland cancer. ⁽²³⁾ Individuals with a primary tumor site of oral cavity had the highest risk of trismus at baseline and during follow-up whereas individuals with a primary tumor site of larynx/hypopharynx, had the lowest risk. While we identified a significant relationship between primary cancer site and risk of trismus, we did not identify a relationship between primary tumor site and change in interincisal opening.

Treatment of the cancer had a significant impact on mouth opening. There was a relationship between dose to the primary site and change in opening distance. Those who received <6000cGy had significantly less reduced mouth opening in follow-up and a greater risk of trismus compared to those who received >6000cGy. Since the incidence of human papilloma virus (HPV)-associated oropharyngeal SCC is increasing and associated with a favorable prognosis and long-term survival, strategies to mitigate side effects of treatment are of significant importance. ^{(24) (25) (26)} Study protocols now include de-escalation of RT dose and volume, ^{(27) (28)}providing an opportunity to consider even further strategies which can mitigate trismus.

Patients who received RT+CT experienced a greater decrease in opening from baseline to subsequent visits compared to RT alone, which has been reported in other studies. ^{(16) (8)}The combination of RT+CT+surgery resulted in even a greater decrease in opening from baseline to subsequent visits compared to RT+surgery (without CT). We did not identify a significant difference in change in mouth opening distance between individuals who received RT+CT+surgery compared to individuals who received RT+CT, suggesting that CT plays a substantial role in mouth opening. However, there was not a significant relationship between treatment modality and baseline trismus or follow-up trismus. This could be related to the small sample size for those who had RT alone. Although there was no relationship between treatment modality and change in trismus in our cohort over the 2 years of observation, others have observed this over longer periods of time. ⁽⁸⁾

Quality of life was significantly associated with trismus as reported in other studies. ^{(29) (8) (30)} Specifically, individuals who did not have trismus at baseline but experienced trismus during follow-up experienced a clinically meaningful increase in problems opening mouth and dry mouth. Those who had trismus at baseline but did not during follow up experienced less of an increase in problems in these QOL assessments.

Although this study followed a large cohort of HNC patients that received contemporary RT doses ≥ 45cGy for two years, there were limitations. The population was predominately Caucasian, non-Hispanic and male. All patients were treated in major cancer centers within the Eastern United States; thus, the results may not be generalizable. Furthermore, our cohort was followed for 2 years while some studies report a more gradual decline in mouth opening after two years. ^{(8) (17)}Although patients who received RT+CT had a greater decrease in opening from baseline to subsequent visits compared to patients who received RT alone, information on specific CT agents, doses, or cycles was lacking. The trismus threshold of <35mm was used for both men and women as sex-based trismus thresholds have not been firmly established in the literature. It is unlikely that these limitations negate the main findings from our study, namely that RT for HNC is related to decreases in mouth opening, negatively impacting QOL among survivors. As deescalated RT algorithms are being investigated in patients with HPV-associated oropharyngeal SCC ^{(31) (32) (33) (34) (27) (28)} proactive management for those at risk for trismus, may have significant impact the QOL of HNC survivors.

Conclusion:

A significant decrease in mouth opening is a common treatment related toxicity after RT, which has a significant impact on survivor QOL. Variability in restricted mouth opening after RT is related to specific patient, cancer, or treatment characteristics. This study clarifies the risk factors associated with decreased mouth opening after RT therapy for HNC and its impact on a survivor’s QOL.

Statement of Clinical Relevance:

Decrease in mouth opening is a common treatment related toxicity after RT, which has a significant impact on survivor QOL. Variability in restricted mouth opening after RT is related to specific patient, cancer, or treatment characteristics.