Abstract
This study evaluated the quality of life (QL) of patients who survived oral cancer more than 5 years after treatment, using the Short-Form 36 questionnaire (SF-36), and assessed the impact of factors influencing QL compared with the Spanish population norms. 60 oral cancer patients (65.41 years; 49 males) with cancer-free survival after surgery of >5 years were enrolled. The outcomes of every dimension of the SF-36 questionnaire in every patient were compared with those of a reference Spanish population. Females had statistically significant negative differences in the dimensions of role-emotional, social functioning and vitality. Patients under 65 years had statistically significant negative differences in the dimensions of physical functioning and general health. In the other variables analysed the differences between groups were only clinically relevant. The QL of patients with oral cancer who survive more than 5 years, when assessed with the SF-36 questionnaire, presented similar values to those of the general population, even exceeding these reference values in some dimensions. It is necessary to evaluate QL in the long-term since patients may need a long time to recover from the disease, and to complement QL assessment with other specific questionnaires.
The management of oral cancer usually includes surgical treatment with simultaneous reconstruction and adjuvant radiation therapy, which often results in functional and aesthetic consequences for patients and has an impact on their quality of life (QL). It is important to carry out an assessment of the QL in the immediate postoperative period and also in patients who survive in the long term. Health-related QL is a multidimensional concept that reflects the subjective health status perceived by the patient (general or specifically) and it gathers information through interviews or questionnaires. Long-term QL assessment aids understanding of how patients evaluate and adapt to the after-effects arising from the treatment in the long term, and identifies more effective therapeutic procedures. Most QL studies tend to have a follow-up period not exceeding 12 months and few articles analyse QL over 5 years, focusing on patients who survive oral cancer without recurrence.
QL is a complex and multidimensional concept the assessment of which is performed by questionnaires specifically designed to evaluate both generally and specifically QL in each disease. General questionnaires provide information on various aspects of patient QL which may be useful to complement the specific questionnaires. The Health-Related Quality of Life Health Survey-Short Form (SF-36) is a questionnaire that assesses general QL that has been validated in a general Spanish population. There are other QL questionnaires validated in Spanish and used in patients with head and neck cancer (EORTC QLQ-30, UW-QOQ), but currently there is not a reference group to be compared with the Spanish general population.
Although the SF-36 has been used in patients with head and neck cancer, to the authors’ knowledge, in the English indexed literature, only five papers evaluate the QL in patients with oral cancer. There is no published article reporting the use of SF-36 in oral cancer patients who underwent surgery and survived for more than 5 years. In this study, the authors evaluated QL using the SF-36 questionnaire in a group of 60 patients who survived oral cancer for more than 5 years and assessed the impact of social-demographic (age and gender), clinical (tumour size and stage) and therapeutic (reconstruction and radiotherapy) factors comparing these values with the Spanish population norms.
Material and methods
A cross-sectional study of patients diagnosed with a squamous cell oral cancer primarily treated with surgery with a curative intention and who survived for more than 5 years without any recurrence of the disease, was conducted by the Department of Oral and Maxillofacial Surgery at University Hospital Virgen Macarena (UHVM) in Seville, Spain. Patients who received palliative or neoadjuvant therapy and those who presented a recurrence of the disease were excluded from the study. The study was approved by the Hospital’s Ethics Committee. All the patients were contacted by telephone and arranged to meet the interviewer in UHVM. They signed a specific informed consent form to take part in the study. No patient refused to enrol in the study.
Patients filled out the SF-36 questionnaire in the presence of the same interviewer (JH) who only intervened if the patient did not understand a question, but who did not indicate or explain the answer. The SF-36 questionnaire comprises 36 questions divided into 8 different dimensions: physical functioning, role-physical, role-emotional, vitality, mental health, social function, pain and the social dimension. All questions are scored on a scale ranging from 0 (poor health) to 100 (good health). The authors compared the score for every dimension of every patient with that of a reference for the Spanish population in the same group for age and gender, according to the tables published by Alonso et al. The impact that variables such as age, gender, tumour size, tumour stage, type of reconstructive surgery and adjuvant radiotherapy have on patients’ QL was also analysed.
Statistical analysis
In order to detect differences in relation to the Spanish general population norms, the score of the median of the general population that fit every individual according to age and gender was subtracted for each dimension of every one of these individuals, and the deviation obtained was taken as a result. If this result was higher than 0, positive differences were obtained; that is to say, individuals had better QL than the median of the population with the same age and gender. On the contrary, if the result was lower than 0, so that negative differences were obtained, the QL of the individual was worse than the median of the population with the same age and gender. Wilcoxon’s test was applied to verify if the above mentioned differences were significant for a level of bilateral significance of P < 0.05 in the whole group and for every possible value of the clinical and social variables analysed. A 5-point difference was used as an indicator of clinically and socially relevant change. All patients who fulfilled the inclusion criteria were enrolled in the study, to produce a final sample size that allowed the detection of differences defined for a significance level of 0.05 and a statistical power of 0.80. For the comparison of independent groups, the differences of every variable obtained by the aforementioned method were analysed using the Mann–Whitney U -test, and a bilateral significance level of P < 0.05 was used.
Results
Between 1992 and 2003, 112 patients who primarily underwent surgery for oral cancer and survived more than 5 years were enrolled in the study, provided that they fulfilled the inclusion criteria. Of these 112 patients, 74 could be located, 14 were excluded (9 patients had died of unrelated causes before they could be contacted; and 5 patients did not attend the consultation) and 38 could not be located. Internal hospital records showed that both groups (14 and 38 patients) had similar characteristics in terms of age distribution, tumour size and stage as the group of patients who completed the questionnaires. Of the 60 patients who took part in the study, 49 were males and 11 females (average age 65.41 years, SD 13.12). Patient survival ranged from 60 to 86 months (average 65.4 months, SD 13.1). The location of the primary tumour was: lateral edge of the tongue (20 cases, 33.3%), floor of the mouth (18 cases, 30%), retromolar trigone (8 cases, 13.3%), lower gum (5 cases, 8.3%), lower lip (4 cases, 6.6%), buccal mucosa (4 cases, 6.6%) and maxilla (1 case, 1.6%). Regarding tumour size, there were two groups: T1–T2 (39 tumours, 65%) and T3–T4 (21 tumours, 35%). There were 17 patients (28%) in stage I, 12 (20%) in stage II, 16 (27%) in stage III and 15 (25%) in stage IV. 43 patients underwent neck dissection (71.6%), complex reconstructive surgery was performed in 18 patients (30%), and postoperative radiotherapy was administered in 22 cases (36.6%). The following major reconstruction procedures were used: forearm free flap in 11 cases; fibula free flap in 3; and pectoralis major muscle flap in 2. Mandibular reconstruction plates were employed in two cases.
Table 1 shows the measures of centre and spread of the dimensions for the SF-36 in every group of patients analysed as well as the levels of significance of those comparisons that resulted in significant differences. Clinically relevant differences are also reflected in the table.
| Physical functioning | Role limitation physical | Role limitation mental | Social functioning | Mental health | Vitality | Pain | General health | |
|---|---|---|---|---|---|---|---|---|
| Male (49) vs PN | ↑ 82.95 (23.11) 90 (80–97.5) |
↓↓ P < 0.024 90.30 (28.32) 100 (100–100) |
95.91 (17.52) 100 (100–100) |
↓↓ P < 0.000 90.62 (14.38) 100 (87.5–100) |
78.69 (16.31) 80 (66–94) |
↑↑ P < 0.002 75.51 (18.17) 80 (65–90) |
↓↓ P < 0.000 86.63 (13.65) 90 (77.5–100) |
↑↑ P < 0.001 72.26 (19.57) 75 (52.5–90 ) |
| Female (11) vs PN | ↓ 62.72 (22.28) 65 (40–90) |
↓ 81.81 (40.45) 100 (100–100 ) |
↓ 72.72 (46.70) 100 (0–100) |
↓↓ P < 0.011 69.31 (23.95) 62.5 (50–100) |
↓ 55.63 (24.88) 52 (36–80) |
↓ 44.54 (24.13) 40 (20–55) |
↓ 63.18 (23.92) 65 (45–80) |
↓ 50 (30.41) 40 (20–80) |
| Female vs male | ↓↓ P = 0.028 | ↓↓ P = 0.018 | ↓ | ↓↓ P = 0.07 | ↑ | |||
| <65 years (23) vs PN | 88.26 (13.45) 95 (80–10) |
↓ 94.56 (21.26) 100 (100–100) |
↓ 91.30 (25.06) 100 (100–100) |
↓↓ P < 0.011 88.17 (18.75) 100 (75–100) |
75.47 (20.89) 80 (56–92) |
↑ 70.65 (24.13) 80 (50–90) |
↓↓ P < 0.001 85.76 (15.30) 90 (70–100) |
↑ 68.31 (23.67) 75 (50–90) |
| >65 years (37) vs PN | ↑ 73.64 (27.52) 85 (52.5–95) |
↓↓ P < 0.049 85.13 (35.08) 100 (100–100) |
↓ 91.89 (27.67) 100 (100–100) |
↓↓ P < 0.000 85.81 (18.20) 87.5 (75–100) |
73.83 (19.78) 76 (60–92) |
↑↑ P < 0.009 69.32 (22.02) 75 (55–82.5) |
↑↑ P < 0.009 80.2019.73) 80 (67.5–100) |
↑↑ P < 0.000 68.10 (23.43) 75 (50–85) |
| >65 years vs <65 years | ↑↑ P = 0.003 | ↓ | ↑ | ↓ | ↑ | ↑ | ↑↑ P = 0.007 | |
| No RT (38) vs PN | 78.81 (26.99) 92.5 (65–96.25) |
94.07 (22.83) 100 (100–100) |
↓ 91.22 (27.60) 100 (100–100) |
↓↓ P < 0.000 87.90 (16.12) 100 (75–100) |
75.89 (18.77) 78 (56–93) |
↑↑ P < 0.000 70.39 (20.04) 75 (55–81.25) |
↓↓ P < 0.001 82.89 (16.17) 83.75 (69.37–100) |
↑↑ P < 0.002 72.13 (22.40) 80 (50–90) |
| RT (22) vs PN | 80 (18.64) 87.5 (67–95) |
↓↓ P < 0.044 79.54 (39.81) 100 (87.5–100) |
92.42 (25.05) 100 (100–100) |
↓↓ P < 0.005 84.65 (21.79) 100 (71.87–100) |
↓ 72 (22.35) 74 (65–89) |
↑ 68.86 (27.07) 72.5 (47.5–91.25) |
↓↓ P < 0.019 81.36 (21.70) 90 (67.5–100) |
61.36 (23.81) 62.5 (47.5–77.5) |
| RT vs no RT | ↓ | ↓ | ||||||
| T1T2 (39) vs PN | ↑ 78.20 (24.50) 90 (65–95) |
↓↓ P < 0.017 85.25 (34.27) 100 (100–100) |
↓↓ P < 0.034 88.88 (30.90) 100 (100–100) |
↓↓ P < 0.000 85.65 (19.84) 100 (75–100) |
73.02 (21.29) 76 (56–92) |
↑ 68.20 (21.71) 75 (50–80) |
↓↓ P < 0.000 79.23 (18.91) 80 (67.5–100) |
↑↑ P < 0.043 67.33 (25.02) 75 (50–90) |
| T3T4 (21) vs PN | ↑ 81.19 (23.81) 90 (75–97.5) |
95.23 (21.82) 100 (100–100) |
96.82 (14.54) 100 (100–100) |
↓↓ P < 0.004 88.69 (15.26) 100 (81.25–100) |
77.14 (17.71) 80 (68–92) |
↑ 72.85 (24.57) 80 (57.592.5) |
↓↓ P < 0.040 88.09 (15.73) 90 (78.75–100) |
↑ 69.76 (20.27) 75 (55–87.5) |
| T1T2 vs T3T4 | ↓ | ↓ | ↓ | ↓ | ||||
| No reconstruction (42) vs PN | ↑ 78.81 (23.81) 90 (65–95) |
↓↓ P < 0.017 94.07 (33.22) 100 (100–100) |
↓↓ P < 0.034 91.22 (29.89) 100 (100–100) |
↓↓ P < 0.000 87.90 (18.82) 100 (75–100) |
75.89 (21.27) 80 (56–93) |
↑ 70.39 (22.30) 75 (53.75–90) |
↓↓ P < 0.000 82.89 (19.05) 80 (67.5–100) |
↑↑ P < 0.005 72.1324.71) 77.5 (50–90) |
| Reconstruction (18) vs PN | ↑ 80 (25.45) 90 (65–100) |
↓ 79.54 (23.57) 100 (100–100) |
92.42 (15.71) 100 (100–100) |
↓↓ P < 0.003 84.65 (17.24) 87.5 (71.87–100) |
72 (17.39) 74 (62–86) |
↑ 68.86 (24.12) 77.5 (47.5–90) |
↓↓ P < 0.007 81.36 (15.76) 90 (7937–100) |
61.36 (19.85) 70 (44.65–76.25) |
| Reconstruction vs no reconstuction | ↑ | ↑ | ↓ | ↑ | ↑ | ↓ | ||
| Radial (11) vs PN | ↑81.07 (25.20) 90 (72.5–100) |
100 (0) 100 (100–100) |
100 (0) 100 (100–100) |
↓↓ P < 0.011 86.82 (15.33) 87.5 (75–100) |
76 (15.76) 74 (67–92) |
↑ 74.64 (20.33) 80 (61.25–90) |
↓↓ P < 0.010 89.64 (14.30) 90 (87.5–100) |
66.87 (18.66) 70 (49.05–78.75) |
Figs. 1–3 show box plots representing the medians of differences for every dimension together with the medians of Spanish population norms by age and gender for all the assessed variables (gender, age, tumour size, stage and treatment) as well as their interquartile range.
