Abstract
Aim: The psychosocial outcomes in long-term survivors of osteosarcoma were investigated. Patients and Methods: A questionnaire concerning the psychosocial status was completed by 30 long-term survivors of osteosarcoma. The family APGAR was employed for the evaluation of family function, social support questionnaire (SSQ) for social support, impact of event scale-revised (IES-R) for posttraumatic stress symptoms (PTSS), and posttraumatic growth inventory (PTGI) for posttraumatic growth (PTG). Results: The family APGAR and SSQ scores were comparable to those of controls. The IES-R score was low, showing a low incidence of PTSS. The PTGI score was high, revealing marked PTG. On multiple regression analysis regarding the IES-R as a dependent variable, the family APGAR was associated with the IES-R. On multiple regression analysis regarding the PTGI as a dependent variable, the age at diagnosis, state of the affected limb and SSQ were associated with the PTGI. Conclusion: The incidence of PTSS was low, and PTG was marked in the long-term survivors of osteosarcoma. A high age at onset and amputation of the affected limb were associated with PTG. Support of the family function reduces PTSS, and the strengthening of social support facilitates PTG.
The prognosis of high-grade osteosarcoma has improved markedly and the number of long-term survivors of treatment for osteosarcoma has increased in recent years, due to multidisciplinary therapy centering around chemotherapy (1). With an increase in the number of long-term survivors of osteosarcoma, the era of simply treating the disease will end, and the quality of life (QOL) of patients will become more important. To improve the QOL of patients with osteosarcoma, it is necessary to clarify the status of and problems regarding the QOL. However, there have been few reports evaluating such patients' QOL (2-4), and so their status and problems remain unclear.
We have performed studies to clarify the status of and problems regarding the QOL in long-term survivors of osteosarcoma who underwent treatment at our hospital, and reported that the state of the affected limb influenced school attendance, there were fewer problems in their employment, and the QOL of long-term survivors was satisfactory excluding physical function on evaluation using SF-36 questionnaire (5-7).
During previous studies, we gained the impression that many long-term survivors of osteosarcoma had a positive attitude, and markedly grew mentally. Psychosocial outcomes, such as posttraumatic stress symptoms (PTSS) and posttraumatic growth (PTG), are important aspects of the QOL, and assessment of these may lead to evaluation of the QOL. We surveyed the psychosocial outcomes in long-term survivors of osteosarcoma who underwent treatment at our hospital to investigate the factors associated with them.
Patients and Methods
Patients. Of 132 patients with high-grade osteosarcoma who were younger than 20 years at the time of diagnosis and underwent treatment at our hospital between 1976 and 2003, the final outcomes were recorded in 122. Of these, 55 were 20 years of age or older at the time of this survey, and had been disease free for 5 or more years after the completion of treatment. A questionnaire concerning the psychosocial status was sent to these 55 survivors by post. The patients were divided based on the state of the affected limb into amputation and limb-sparing groups. Those who had undergone knee rotationplasty were included in the amputation group. Prior to this study, the protocol was approved by the Institutional Research Board. Written informed consent was obtained from each patient.
Therapeutic policy toward high-grade osteosarcoma. The surgical therapy given at our hospital will be described briefly. Several surgical methods such as amputation, rotationplasty, arthroplasty and arthrodesis are available. At our hospital, as many surgical options as possible were given to patients, and the selection of the surgical method was decided by the patient. Where there was no option but amputation, we carried out amputation after having given a full explanation until the patients were ready to accept it.
Clinical characteristics of the patients.
The changes in chemotherapy for high-grade osteosarcoma at our hospital will be described briefly. Adjuvant chemotherapy including doxorubicin and high-dose methotrexate was started from 1976. Cisplatin from 1982, bleomycin, cyclophosphamide and dactinomycin (BCD) from 1987 and ifosfamide from 1991 were each added to adjuvant chemotherapy. Since 1987, intensive-dose chemotherapy using the T10 and T12 protocols (8, 9) as a guide has been given. For patients who unfortunately developed pulmonary metastasis, second-line chemotherapy was administered after aggressive thoracotomy. HELP (vindesine, ifosfamide and cisplatin) regimen (10) and etoposide-cyclophosphamide regimen (11) were administered as second-line chemotherapy.
Evaluation of psychosocial outcomes. Inventories used in preceding studies were employed for the evaluation of the psychosocial outcomes. The family APGAR was employed for evaluation of the family function (12, 13), the 6-item social support questionnaire satisfaction and number (SSQ-S and SSQ-N) for social support and social withdrawal (14, 15), the Japanese-language version of the impact of event scale-revised (IES-R) for posttraumatic stress symptoms (PTSS) (16, 17), and the Japanese-language version of the 18-item posttraumatic growth inventory (PTGI) for posttraumatic growth (PTG) (18-20). Replies to the questionnaire were converted to scores, and the means, standard deviations and minimum and maximum values were calculated.
Distribution of variables.
Statistical analysis. Correlations among the variables of the age at diagnosis, family APGAR, SSQ-S, SSQ-N, IES-R, and PTGI were investigated using Pearson's correlation coefficient. The correlation between the subscales of IES-R and PTGI was also investigated.
Forward stepwise multiple regression analysis was performed regarding the IES-R as a dependent variable. Six independent variables: the gender, age at diagnosis, state of the affected limb at the time of survey, family APGAR, SSQ-S and SSQ-N were included in the original model. Regarding gender, male and female were graded ‘0’ and ‘1’ as dummy variables, respectively. As for the state of the affected limb at the time of the survey, amputation and limb-sparing were graded ‘0’ and ‘1’ as dummy variables, respectively.
Forward stepwise multiple regression analysis regarding the PTGI as a dependent variable was also performed. The original model included the gender, age at diagnosis, state of the affected limb at the time of survey, family APGAR, SSQ-S, and SSQ-N as independent variables.
A p-value <0.05 was considered to be statistically significant for all tests. The data were analyzed with StatView (version 5.0 for Macintosh; Abacus Concepts, Inc. Piscataway, NJ, USA).
Results
Responses to the questionnaire were obtained from 30 patients, a response rate of 54.5% (30/55). The clinical characteristics of the patients are shown in Table I. The amputation and limb-sparing groups consisted of 16 and 14 patients, respectively. Knee rotationplasty was performed in 4, and these were included in the amputation group.
The mean, standard deviation and minimum and maximum values of each variable are shown in Table II. The correlation coefficients among the variables are listed in Table III, and those between the subscales of the IES-R and PTGI in Table IV.
On forward stepwise multiple regression analysis regarding the IES-R as a dependent variable, the coefficient of determination (R2) was 0.250, and the p-value was 0.0049, showing that the fit was significant. The family APGAR was adopted as an independent variable by stepwise analysis (step 1). The standardized regression coefficient of the family APGAR with the IES-R was −0.500 (p=0.0049).
Correlation between variables.
Correlation between IES-R subscales and PTGI subscales.
On forward stepwise multiple regression analysis regarding the PTGI as a dependent variable, the R2 and p-value were 0.509 and 0.0003, respectively, showing that the fit was significant. The age at diagnosis, state of the affected limb at the time of survey, and SSQ-N were adopted by stepwise analysis (step 3). The standardized regression coefficients of these 3 variables with the PTGI were 0.347 (p=0.0197), −0.470 (p=0.0031) and 0.611 (p=0.0002), respectively.
Discussion
The psychosocial characteristics of the participants were investigated by a comparison of the mean scores of the scales with those in preceding reports. The mean family APGAR score in the long-term survivors of osteosarcoma was 7.97 points, whereas that in 527 university students in a survey performed by Smilkstein et al. was 7.61 (13), showing that the family function of the long-term survivors was equivalent to that of the controls. The mean SSQ-N and SSQ-S scores were 4.03 and 5.02 in the long-term survivors, respectively, while those in 178 controls surveyed by Furukawa et al. were 4.43 and 4.89, respectively (15). The SSQ scores decrease in patients experiencing traumatic events. The SSQ scores in the long-term survivors were comparable to those in the controls, showing only slight social withdrawal in the survivors.
The mean IES-R score was 9.70. Posttraumatic stress disorder (PTSD) is suspected when the IES-R score is 25 or above (17). The score was 25 or greater in only 3 out of the 30 patients (10%), showing that the incidence of PTSS was low in the long-term survivors. The mean PTGI score was 51.8, while that in 312 Japanese university students with some traumatic events reported by Taku et al. was 33.9 (19). The PTGI score was high and PTG was marked in the survivors. The long-term survivors of osteosarcoma had few PTSS and their PTG was marked.
Although PTSS and PTG were correlated in preceding reports (21), there was no significant correlation between the IES-R and PTGI in the long-term survivors of osteosarcoma (Table IV). Regarding correlations between the subscales, significant correlations were present only between ‘hyperarousal’ of IES-R and ‘spiritual change and appreciation of life’ of PTGI, and there was no correlation between PTSS and PTG.
As shown in Table III, the correlation coefficients among the variables were moderate or weaker, suggesting no multico-linearity problem in the multiple regression analysis. On forward stepwise multiple regression analysis regarding the IES-R as a dependent variable, the family APGAR was associated with the IES-R. The family function determined based on the family APGAR was the factor most strongly associated with PTSS. Bressound et al. also reported that an unbalanced family structure was associated with PTSS and a poor coping ability (22). The incidence of PTSS was high in patients with a poor family function, and support to improve family function reduces PTSS.
On forward stepwise multiple regression analysis regarding the PTGI as a dependent variable, the age at diagnosis, state of the affected limb, and SSQ-N were associated with the PTGI. Patients with a high age at onset and the amputation group showed marked PTG. Surprisingly, amputation led to positive growth. Growth was marked in patients with strong social support determined based on SSQ-N, showing that the strengthening of social support facilitates PTG.
No definite conclusion could be drawn because the number of cases was small. Moreover, the actual state of all long-term survivors might not necessarily be clarified because only patients who gave consent were surveyed. We will perform a multicenter cooperative study to investigate a great number of cases.
- Received April 17, 2009.
- Revision received July 20, 2009.
- Accepted August 12, 2009.
- Copyright© 2009 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
