Health-related quality of life after carbon-ion radiotherapy for prostate cancer: A 3-year prospective study

Authors


Abstract

Objectives

To assess 3-year health-related quality of life of patients treated with carbon ion radiotherapy for prostate cancer.

Methods

A total of 213 patients received carbon-ion radiotherapy at a total dose of 66 Gy equivalent in 20 fractions over 5 weeks, and neoadjuvant and adjuvant androgen deprivation therapy were administered for high-risk patients for at least 12 months. A health-related quality of life assessment was carried out at four time-points (immediately before the initiation of carbon-ion radiotherapy, immediately after, 12 and 36 months after completion of carbon-ion radiotherapy) using Functional Assessment of Cancer Therapy General and for Prostate Cancer Patients.

Results

The evaluable response rates among all responses were more than 94%. Overall, a significant decrease in the scores of the health-related quality of life 12 months after carbon-ion radiotherapy returned to their baseline levels at 36 months. Additionally, no significant decrease was observed in the scores at any of the assessment time-points compared with their baseline scores in the group of carbon-ion radiotherapy without androgen deprivation therapy; however, the presence of morbidity and biochemical failure significantly worsened the scores, and the decreases in the scores did not improve even at 36 months after carbon-ion radiotherapy.

Conclusions

An assessment based on a subjective scoring system shows a significant decrease in health-related quality of life at 12 months after carbon-ion radiation therapy, which tends to return to baseline levels at 36 months. The presence of morbidity and biochemical failure significantly worsen health-related quality of life scores. Further controlled studies focusing on health-related quality of life assessment in patients with prostate cancer are warranted.

Abbreviations & Acronyms
ADT

androgen deprivation therapy

BF

biochemical failure

C-ion RT

carbon-ion radiotherapy

EWB

emotional well-being

FACT-G

Functional Assessment of Cancer Therapy-General

FACT-P

Functional Assessment of Cancer Therapy-Prostate Cancer Patients

FWB

functional well-being

GyE

gray equivalents

HRQOL

health-related quality of life

IMRT

intensity-modulated radiation therapy

PCS

prostate cancer subscale

PSA

prostate-specific antigen

PWB

physical well-being

QOL

quality of life

RBE

relative biological effectiveness

RT

radiation therapy

SFWB

social/family well-being

TOI

Trial Outcome Index

Introduction

The opportunity to provide curative treatment for patients with prostate cancer has recently been increasing in Japan mainly because of widespread screening by measuring serum PSA levels.[1] As modern RT methods, such as IMRT, brachytherapy and charged particle therapy under image guidance, can three-dimensionally deliver a large dose to the tumor and allow the sparing of surrounding normal tissues, the rates of tumor control and morbidity after curative RT have been improved.[2-4] Thus, RT has become one of the standard treatments for prostate cancer, as well as radical surgery.

In contrast, it is difficult to determine the best RT method for individual prostate cancer patients, because there are small differences in the tumor control probability and potential adverse events among highly conformal RT methods. It is therefore thought that the HRQOL assessment, a subjective scoring system, is one of the effective tools for comprehensively comparing multiple treatment modalities including their adverse event profiles.[5] We previously reported favorable safety HRQOL results of C-ion RT for prostate cancer based on 1-year follow-up data.[6] However, a prospective long-term study of HRQOL in a sufficient sample size might allow this assessment to be used as an objective index to compare multiple treatments. Thus, the present study aims to report HRQOL results of C-ion RT obtained from more than 200 patients in a 3-year prospective study.

Methods

Patient characteristics

A total of 213 Japanese men with prostate cancer underwent C-ion RT at the National Institute of Radiological Sciences between April 2000 and May 2004. Patients who were eligible for our treatment had primary adenocarcinoma of the prostate with stage T1c-T3bN0M0 according to the 1997 International Union Against Cancer TNM classification,[7] based on diagnostic imaging including bone scanning, chest radiography, computed tomography, transrectal ultrasound and magnetic resonance imaging. Written informed consent was obtained from all participants before registration into the present study, and the participant characteristics are summarized in Table 1.

Table 1. Participant characteristics
Mean age (range)69.4 years (53–92)
Median PSA (range)18.0 ng/mL (3.4–500)
 No. patients (%)
Stage (International Union Against Cancer 1997)
T1c81 (38.0)
T2a34 (16.0)
T2b32 (15.0)
T3a-b66 (31.0)
Gleason Score
≤651 (24.0)
798 (46.0)
8≤64 (30.0)

C-ion RT

The technique of C-ion RT for prostate cancer has been previously described elsewhere.[8, 9] Briefly, C-ion RT was carried out only on the prostate and seminal vesicles without pelvic irradiation, and it was given once a day, 4 days a week (from Tuesday to Friday), and one port was used in each session.[4, 8, 9]

Irradiation doses were expressed in the photon gray equivalent (GyE) doses (physical carbon-ion dose [Gy] × RBE).[4, 8, 9] Irrespective of the size of the spread out Bragg peak, the RBE value for carbon ions was estimated to be 3.0 at the distal part of the SOBP. In principle, a total dose of 66 GyE was delivered in 20 fractions with a fractional dose of 3.3 GyE, and it was equivalent to 86.4 Gy with a fractional dose of 2 Gy when the α/β ratio of prostate cancer was 1.5 Gy using the linear-quadratic model.[10] For three (1.4%) patients with uncontrolled severe diabetes mellitus, the total dose was reduced to 63 GyE in 20 fractions with a fractional dose of 3.15 GyE (equivalent to 83.7 Gy with a fractional dose of 2 Gy) to reduce the severity of the side-effects.

ADT

ADT consisting of medical or surgical castration with or without anti-androgen was delivered according to the risk group of prostate cancer. Participants meeting all of the criteria (stage ≤T2a, Gleason score ≤6, and PSA ≤ 20 ng/mL) were defined as the low-risk group (n = 40), and C-ion RT alone was carried out for them. The remaining 173 participants were defined as the high-risk group, and C-ion RT in combination with ADT was delivered. ADT was initiated 2–6 months before the initiation of C-ion RT, and the therapy was continued for at least 1 year after the completion of C-ion RT.[4]

In 73 (42%) patients, ADT was completed within 36 months after the completion of C-ion RT, and the median duration of ADT after C-ion RT was 13 months, but the median duration of ADT after C-ion RT was 43 months in the remaining 100 participants.

QOL questionnaire

The assessment of HRQOL in the present study was carried out using the FACT-G and FACT-P questionnaire version 4 (Japanese version).[11] To explain briefly, the FACT-G instrument is a self-report questionnaire consisting of 27 question items, and it comprises four subdomains: PWB, FWB, SFWB and EWB. The PCS contained in the FACT consists of 12 questions particularly designed to measure the HRQOL in patients with prostate cancer. The FACT-P is the combination of the four subdomains of the FACT-G and the PCS. The TOI is created by summing the PWB, FWB and PCS, and is considered to be a sensitive index focusing on the physical aspect of HRQOL.[12, 13]

HRQOL assessment was carried out at the following four time-points: (i) immediately before the initiation of C-ion RT; (ii) immediately after; (iii) 12 months after the completion of C-ion RT; and (iv) 36 months after the completion of C-ion RT. The questionnaire sheet immediately before the initiation and immediately after completion of C-ion RT was given to each participant in person, and was collected on completion, whereas it was received from each participant through the mail at 12 and 36 months after completion of the therapy.

Statistical analysis

Calculation of statistics was carried out using spss version 11.5 (SPSS, Chicago, IL, USA). Linear mixed models were used to compare data between groups, and between the baseline and each assessment time point, and the Bonferroni method was used for adjustment for multiple comparisons. Significant differences were assessed using two-sided tests with P < 0.05.

Survival times were calculated using the Kaplan–Meier method.[14] Cumulative survival and biological relapse-free rates were compared by the log–rank test among the different subgroups.

Results

Survival and questionnaire submission rate

The curves of overall survival and biochemical failure-free survival in all participants are shown in Figure 1. The 3-year rates of overall survival and biochemical failure-free survival were 93.4% and 91.4%, respectively. The submission rate of the QOL questionnaire and the response compliance rate are shown in Table 2. At the time of 12 months after the initiation of C-ion RT, one participant died due to causes unrelated to prostate cancer. At 36 months, four participants died of progressive disease of prostate cancer and 10 participants died of intercurrent diseases unrelated to prostate cancer.

Figure 1.

The Kaplan–Meier curves of overall survival and biochemical failure-free survival in all participants. image, Overall Survival; image, Biochemical Failure Free.

Table 2. Submission rate of the HRQOL questionnaire and response compliance rate
 Before C-ion RTAfter C-ion RT
1 Month12 Months36 Months
No. living participants213213212199
No. responses205192209186
No. adaptive responses196181197177
% of responses of all participants96.290.198.187.3
% of responses of living participants96.290.198.693.5
% of adaptive responses of all responses95.694.394.395.2

Relationship of late morbidity with QOL scores

QOL

The mean and standard errors of scores of the individual subdomains, FACT-P, FACT-G and TOI obtained from all of the 213 participants are shown in Table 3. In comparison with the corresponding baseline scores, a significant decrease was observed in all FACT-G, FACT-P and TOI scores at 12 months. However, these scores returned to the baseline levels, and there was no significant difference in the scores between before and 36 months after C-ion RT.

Table 3. Mean (standard error) of scores of the individual subdomains, FACT-P, FACT-G and TOI obtained from all of the 213 participants
SubdomainsFormulaFull scoreBefore C-ion RTAfter C-ion RT
Immediately12 Months36 Months
  1. *P < 0.05 in comparison with the corresponding baseline scores.
Physical well-being(1)2824.7 (0.3)24.3 (0.3)24.1* (0.3)24.5 (0.3)
Social/familial well-being(2)2818.2 (0.4)18.2 (0.4)16.1* (0.4)15.8* (0.4)
Emotional well-being(3)2418.5 (0.3)19.7* (0.3)19.4* (0.3)19.1* (0.3)
Functional well-being(4)2821.2 (0.4)21.0 (0.4)20.8 (0.4)21.5 (0.4)
Prostate cancer subscale(5)4834.9 (0.5)32.5* (0.5)33.6* (0.5)34.3 (0.5)
FACT-G score(1 + 2 + 3 + 4)10882.7 (1.0)83.2 (1.0)80.3*(1.0)80.9 (1.0)
FACT-P score(1 + 2 + 3 + 4 + 5)156117.5 (1.3)115.7 (1.3)113.9* (1.3)115.3 (1.3)
TOI score(1 + 4 + 5)10480.8 (1.0)77.9* (1.0)78.4*(1.0)80.5 (1.0)

Participants were divided into two groups by concomitant ADT to carry out a subset analysis (Fig. 2). In the group of C-ion RT alone, no significant decrease was observed in FACT-G, FACT-P or TOI scores at any of the assessment time-points compared with their baseline scores. In the group of C-ion RT in combination with ADT, significant decreases were observed in TOI scores immediately after the therapy, and in FACT-G, FACT-P and TOI scores at 12 months compared with their baseline scores. However, at 36 months, no significant decrease was observed in any scores compared with their baseline scores.

Figure 2.

Comparison of changes in the HRQOL scores between C-ion RT alone and combination with ADT. *P < 0.05 in comparison with the corresponding baseline scores with Bonferroni correction for multiple comparisons. image, C-ion RT alone (n = 40); image, with ADT (n = 173).

Late morbidity

In the present study, late adverse events were assessed according to the Radiation Therapy Group/European Organization for Research and Treatment of Cancer late morbidity criteria,[15] and are summarized in Table 4. Neither grade 3 nor greater late adverse events were observed at any assessment time-point. In 161 participants, grade 1–2 late adverse events in the rectum and/or bladder/urethra were observed, but no adverse events were observed in the remaining 52 participants.

Table 4. Incidence of adverse events in all of the 213 participants
No. patients (%)Bladder/urethraRectum
Grade 0Grade 1Grade 2Grade 3Grade 0Grade 1Grade 2Grade 3
Maximum64 (30.0)136 (63.8)13 (6.1)0 (0.0)171 (80.3)37 (17.4)5 (2.3)0 (0.0)
Last follow up179 (84.0)30 (14.1)4 (1.9)0 (0.0)197 (92.5)14 (6.6)2 (0.9)0 (0.0)

In the group with the late adverse events (group A), a significant decrease was observed in FACT-G and FACT-P scores, even at 36 months, compared with their baseline scores, and FACT-P and TOI scores in the group without the events (group B) were significantly higher than those in group A (Fig. 3).

Figure 3.

Comparison of changes in the HRQOL between the groups with and without grade 1–2 late adverse events. *P < 0.05 in comparison with the corresponding baseline scores with Bonferroni correction for multiple comparisons. **P < 0.05 in comparison with two groups during overall period before to 36 months after C-ion RT. image, Grade 0 (n = 52); image, Grade 1–2 (n = 161).

To confirm the effect of adverse events on the HRQOL scores, further investigation was carried out in each point separately. At 12 and 36 months after C-ion RT, grade 1–2 late adverse events were observed in 81 (38%) and 92 (43%) patients, respectively. At 12 months, only TOI score in group B was significantly higher than that in group A, but all of the FACT-G, FACT-P and TOI scores in group B were significantly higher than those in group A 36 months after C-ion RT (Fig. 4).

Figure 4.

Comparison of changes in the HRQOL between the groups with and without grade 1–2 late adverse events at (a) 12 months and (b) 36 months after C-ion RT. *P < 0.05 with Student's t-test. (a) image, Grade 0 (n = 116); image, Grade 1–2 (n = 81); (b) image, Grade 0 (n = 82); image, Grade 1–2 (n = 96).

BF

Assessment of BF was carried out according to the Phoenix criteria,[16] and BF was observed in 28 of 213 participants. Participants were also divided into two groups by the presence of BF to carry out a subset analysis (Fig. 5). A significant interaction between the presence of BF and the assessment time-points was noted in FACT-P and TOI scores. This interaction reflects a difference in the change in scores over time between the groups with and without BF. In the group without BF, significant decreases observed in FACT-P and TOI scores immediately after the therapy or at 12 months also disappeared at 36 months because of a return of the scores to the baseline levels. In contrast, in the group with BF, the HRQOL scores all decreased at any time-points significantly compared with their baseline scores, and no obvious improvement of the scores was observed.

Figure 5.

Comparison of changes in the HRQOL between the groups with and without BF. *P < 0.05 in comparison with the corresponding baseline scores with Bonferroni correction for multiple comparisons. A significant interaction between the presence of BF and the assessment time points was noted in FACT-P and TOI scores. image, without BCF (n = 185); image, with BCF (n = 28).

Discussion

Studies on the analysis of HRQOL assessment in prostate cancer patients carried out for up to 2 years have been reported,[6, 17-20] but many of the studies were carried out as cross-sectional studies, and results from a prospective longitudinal study have rarely been reported. Taking the long investigation period of 3 years, and a high response rate of the assessment in the present study into account, the results of the study might be qualitatively credible.

It is suggested that the longer the investigation period lasts, the more difficult it is to obtain necessary responses for a proper analysis. Although participants were asked to complete and send the QOL questionnaire sheet by mail at the 12-month and 36-month assessment time-points, the evaluable response rate was approximately 95% at all assessment time-points; therefore, this collection method of the questionnaire sheet was considered to be acceptable.

For HRQOL assessment in malignancies with a short natural course, such as lung cancer and esophageal cancer, the analysis of scores with missing data discarded would result in a significant overestimation.[21, 22] Compared with these types of malignancies, prostate cancer has a relatively longer natural course. However, in a study carried out for a longer period of time, the number of deaths increased even in prostate cancer treatment. In fact, as seen in the 14 cases of death (6.6%) in the present study, the validity of such an analysis in disregarding these QOL data might be uncertain. Therefore, the linear mixed model analyses, which can provide a valid analysis when data are missing at random, were used in the present study to avoid overestimation of the data.

The SFWB score at 36 months in the present study remained significantly lower than its baseline score. Among the subdomains of FACT-P, the change in the SFWB score was very different from those in the other scores. Fujimoto et al. evaluated the validity in Japan of the FACT-G questionnaire for lung cancer, and have raised potential cross-cultural issues regarding the SFWB subdomain.[23] Thus, analysis of the SFWB among the subdomains of FACT-P might require sufficient considerations in Japan. In contrast, the TOI score, which does not include SFWB domain, is considered to be a sensitive index focusing on the physical aspect of HRQOL (Table 3).

The presence of treatment-related adverse events, particularly late adverse events, is an essential factor for evaluating feasibility of treatment modalities for malignancies. Although there was no grade 3 or greater late adverse events in the present study, grade 1 or 2 late adverse events were observed in 76% of the participants. However, it is often difficult to accurately assess these minor events, especially for the grade 1 events. It was interesting that the TOI scores changed over time in a similar manner in the groups with and without adverse events, although there was a significant difference between these groups overall. This result is considered to possibly reflect the presence of minor adverse events that do not allow clinically objective assessments. Therefore, HRQOL assessment could be an alternative measure for assessing minor adverse events.

The present study showed that changes in HRQOL scores over time are also significantly different between patients with and without BF. There are reports stating that HRQOL can be a factor for survival in patients with malignancies,[24-26] but the present results cannot argue whether HRQOL affected survival of prostate cancer patients in the present study. It is likely that a minor increase of PSA during several follow-up examinations or the knowledge of the presence of BF led to impairment of the psychological aspect of HRQOL. Furthermore, the start of salvage therapy might have an influence on the physical aspect of HRQOL.

In the present study, the baseline of HRQOL assessment was set before the initiation of C-ion RT rather than before the initiation of the entire treatment, and this is possibly a limitation of the study. Because ADT is associated with various side-effects, including hot flashes, obesity and erectile dysfunction, it could possibly have a major impact on HRQOL, and decrease the baseline score. However, the scores of FACT-G, FACT-P or TOI further decreased in the group receiving a combination of ADT with C-ion RT, whereas no obvious change was observed in the group receiving C-ion RT alone. Additionally, the scores in the combination group at 36 months returned to baseline levels, although adjuvant ADT was still continued for more than half of the participants. These results could indicate that the effects of C-ion RT reflect change of the HRQOL scores in the present study.

Lee et al. carried out a prospective longitudinal study to analyze the HRQOL assessment using FACT questionnaires for up to 12 months after external beam radiotherapy, radical prostatectomy or interstitial brachytherapy for prostate cancer.[17] As HRQOL assessment in their study was different from the present study in view of its version, in addition to the missing data at 3 months in the study, it is impossible to directly compare their results with the present results. Nevertheless, baseline HRQOL scores in the case of C-ion RT tended to be maintained at a higher level 12 months after the treatment when compared with other treatment modalities, and they returned to the baseline levels at 36 months.

In addition to lack of baseline HRQOL information, a potential limitation of the present study was that the assessment of HRQOL was carried out using the FACT questionnaire only; although other questionnaires, such as University of California-Los Angeles Prostate Cancer Index, Expanded Prostate Cancer Index Composite and Short Form-36 have been also introduced.[18, 27, 28] However, the results of the first prospective study carried out continually over 3 years in prostate cancer treatment with charged particle beam RT will give us the opportunity to compare the patients' QOL after various treatment modalities, because the FACT-P has been widely spread and well validated by many previous studies undergoing RT, surgery and chemotherapy.[17, 29, 30] Additionally, the HRQOL scores evaluated in the present study have a significant relationship with occurrence of late adverse events, and we believe that the results using the FACT questionnaire must be a useful tool for evaluation of patients' QOL after C-ion RT.

A 3-year prospective HRQOL study was carried out in C-ion RT for prostate cancer using the FACT questionnaires, and favorable compliance was obtained. The present study suggests that the assessment, a subjective scoring system, is one of the most effective methods for comprehensively comparing individual treatments including their adverse event profiles. Overall, a significant decrease in the FACT-P and TOI scores at 12 months after C-ion RT tended to return to their baseline levels at 36 months, but the presence of morbidity and biochemical failure significantly worsened the scores. Further controlled studies focusing on HRQOL in prostate cancer patients are warranted to compare the HRQOL after C-ion RT with those treated with other recent RT modalities.

Conflict of interest

None declared.

Ancillary