Curative external beam radiotherapy in patients over 80 years of age with localized prostate cancer: A retrospective rare cancer network study

1. Introduction 

Rapid population ageing in most areas of the developed world is among the causes of a significant increase in the cancer burden of the elderly. When attention is focused on people over 75 years of age, prostate cancer (PC) becomes the most common tumour diagnosed in males in Europe and North America with more than 44,000 incident cases occurring in the European Union per year in the 2000 [1]. However the appropriate management of elderly patients diagnosed with a localized PC remains controversial. Most urologists consider that benefits of aggressive treatment are greatly reduced since complication rates increased in older men. However, for the case of radiotherapy, several retrospective studies have shown that patient age did not independently influence gastrointestinal or urinary toxicity after radiotherapy either for PC or for other pelvic malignancies [2], [3], [4], [5], [6]. The aim of the present study was to report on patients who were both diagnosed a PC in their 1980s and treated using curative radiotherapy. In this cohort of elderly patients age 80 and greater, the impact of high dose external radiation therapy was analysed in terms of toxicity and survival.

2. Patients and methods 

Information was collected through the database of 12 departments of Radiation Oncology affiliated to the Rare Cancer Network. Between January 1990 and December 2000, 65 patients aged 80 or older were treated with definitive, potentially curative radiation therapy. The median patient age was 81 years (range: 80–89). The indication for curative radiotherapy was considered in accordance with local policies and guidelines, taking into account the decision of the patient. Exclusion criteria included presence of distant metastases, locally advanced disease (bladder and/or rectal involvement) and co-morbidity conditions likely to lead to an extremely high risk of death. The study was conducted in accordance with guidelines of the participating institutions research ethics boards or institutional review boards.

Forty-one (63%), 21 (32%) and 3 (5%) patients had a WHO performance score of 0, 1 and 2 respectively. Fifty percent of patients had co-morbidities requiring permanent medication: 39 patients had only one co-morbidity, 14 patients had 2 and 1 patient had 3 significant co-morbidities. Cardiovascular disease (30%), diabetes (5%) and pulmonary disease (4%) were the most common illnesses observed in the series. Co-morbidities were objectively measured using an instrument described by Greenfield et al. [7] adapted by Fouad et al. [8]. An index of disease severity (IDS) ranging from 0 (absence of co-morbid condition) to 4 (very serious condition carrying to high risk of mortality) was assigned for each individual and for each co-morbid condition. According to data obtained from the medical records, the IDS was scored 0, 1, 2 and 3 for 11 (17%), 18 (27%), 34(52%) and 2 (4%) patients, respectively.

Previous cancer was noted in 12 patients (minimal “other cancer” free survival: 5 years, median follow-up: 7 years). The previous cancer types observed were skin cancers, colonic cancers and laryngeal cancer in 8, 3 and 1 cases, respectively. First clinical signs leading to diagnosis of PC were typically frequency and/or hematuria but DRE and PSA levels revealed PC in 8 and 27 patients, respectively. All patients had histopathological confirmation of adenocarcinoma. Gleason score was obtained in all tissue specimens except two. At the onset of radiotherapy, all 65 patients were free of distant metastases as assessed by CT scan and chest radiography. Bone scintigraphy was not routinely performed. No patients in the current series had a pelvic lymphadenectomy. Eighteen patients (28%) received neoadjuvant hormonal treatment and 6 had adjuvant androgen deprivation therapy (9%) of at least 6 months duration. Hormonal therapy was indicated according to local policy (generally T3 stages and/or Gleason>7). For those who started androgen blockade before radiotherapy, the mean duration of treatment was 6 months. The patients who received systematic hormonal treatment after radiotherapy were treated indefinitely until relapse or death. All patients were irradiated using a >10MV photon beam. Thirteen patients were treated with four opposed fields (box technique) and the 52 others with a three-dimensional (3D) conformal radiotherapy. Twenty-seven patients received pelvic irradiation to a dose of at least 45Gy (41%). The 27 patients whose pelvis was irradiated had at least one of the following criteria: Gleason>7, T2b or T3 stages, pre-treatment PSA>10ng/ml. Not all the patients fulfilling these criteria had their pelvis treated considering their heavy co morbidity. The prostate alone was the clinical target volume for the remaining 38 patients. The mean total dose to the prostate was 69.5Gy (range 60–78) delivered in 35 fractions according to ICRU 62.

During radiotherapy, acute toxicity was evaluated weekly by a physician and retrospectively scored according to the Common Toxicity Criteria version 2.0. Follow-up visits were scheduled according to each centre policy, but at least two visits per year were typically carried out. For late side effect evaluation, the EORTC/RTOG toxicity score was used. Biochemical failure was validated in most centres after three rising PSA values at 1/2-month intervals (ASTRO 1997 consensus guidelines) since Phoenix consensus criteria could not be assessed retrospectively in a large number of cases. Clinical local failure was assessed either by DRE and or/transrectal ultrasonography. Biopsy was not systematically performed. No systematic attempt was made to evaluate the quality of life before and after treatment. The median follow-up was 65 months (22–177). Estimations of overall survival and event free survival were performed using the Kaplan–Meier method. Statistical analysis was performed using SAS v.9.1 software.

3. Results 

3.1. Toxicity 

Considering the WHO performance status, 16 patients (25%) noted a deterioration of their WHO score during the course of radiotherapy, 14 patients (21%) moved from a score 0 to a score 1 and one patient started with a score 0 and ended his irradiation with a score 2. The last patient died soon after the end of radiotherapy and the cause of death was attributed to a general physical deterioration. Interruptions in radiotherapy were observed in 4 cases (6%): holidays, machine maintenance and acute side effect were the reasons of interruption in 2, 1 and 1 cases, respectively. The mean duration for treatment was 51.8 days (range 39–80). The clinical pathologic features and the treatment modalities of the patients are summarized in Table 1.

Table 1. Clinical features and therapeutic patterns.

		n	%
	Tumour stage
	T1	10	15.4
	T2	40	61.5
	T3	15	23.1
	Gleason score
	2–4	2	3.2
	5–7	45	71.4
	8–10	16	25.4
	Not available	2
	PSA (ng/ml)
	<10	18	28.6
	10–20	23	36.5
	>20	22	34.9
	Not available	2
	Performance status
	0	42	64.5
	1	20	32
	2	3	3.5
	3	–
	Treatment 0
	RT only	41	63
	RT+hormone	24	37

The incidence of genitourinary and rectal acute and late toxicities reported according to the CTC and RTOG/EORTC criteria are shown in Table 2. No patients had grade 4 gastrointestinal or genitourinary toxicity. In the group of patients who received a pelvic radiotherapy, 10 (37%) had grades 2 and 3 acute digestive side effects (mainly diarrhea) compared to 4 patients (10.5%) in the group of patients who did not. No difference of toxicity was observed between patients irradiated using a four-fields box technique and those treated with 3D conformal radiotherapy. Among the 11 late urologic effects observed, 3 patients had an urethral stricture necessitating an internal urethrotomia. The 8 patients with late genitourinary side effects mainly suffered from frequency and occasionally from hematuria. None had incontinence or needed surgery. Digestive late effects (proctitis and/or rectal bleeding) concerned 6% of the patients. Most patients with diarrhoea recovered soon after the end of the treatment without sequellae. Finally 17 patients (26%) experienced one or more late urinary or digestive complications related to radiotherapy, the majority of these events occurring during the first 3 years of follow-up. Previous digestive disease or abdominal surgery was not recorded for the patients in this series and could not be correlated to the digestive side effects.

Table 2. Acute and late toxicity.

	Acute toxicity	Bladder, urethral	Digestive
	Grade 0	11 (18%)	21 (34%)
	Grade 1	29 (44%)	30 (45%)
	Grade 2	17 (26%)	8 (12%)
	Grade 3	8 (12%)	6 (9%)
	Grade 4a

	Late toxicityb	Hematuria, urinary frequency/retention		Proctitis, rectal bleeding
	Grade 1	0	4	3
	Grade 2	2	3	3
	Grade 3	1	2	1

a One death occurred immediately after the end of radiotherapy due to general physical deterioration. b One patient with a hip fracture occurring 6 years after radiotherapy.

3.2. Disease free survival and overall survival 

Median follow-up was 65 months. Eighteen patients developed a biochemical failure during follow-up (25%). Among them, 4 patient's PSA values never returned to a normal range after therapy. The lack of PSA normalization following radiotherapy correlated with a clinical local failure in 3 cases (75%). Out of the 14 other patients, 3 had a clinically proven local failure and 2 had bone metastases. The total number of patients with a clinical local failure was 4 (6%). The number of biochemical, local and distant failures are shown in Table 3 according to the modality of irradiation and the use of either neo-adjuvant or adjuvant hormonal. 54 of 65 patients were found NED at the time of last evaluation, 38 of which remained alive, 11 which had died, and 5 which were lost to follow-up. Eleven patients were found to have relapsed, 9 of which had died and 2 of which were lost to follow-up. After prostate treatment, 3 patients had a second malignancy (1 colonic, 1 bladder and 1 pancreatic tumour), all of them died shortly after their diagnosis. The death causes appear in Table 4.

Table 3. Type of failures according to pelvic radiotherapy and hormonal treatment.

		Pelvic RT (27)a	No Pelvic RT (38)	Hormonal T (18)	No Hormonal T (47)
	Biochemical
	Failures	7(26%)	11(28%)	4(22%)	14(30%)
	Local	2(7%)	2(5%)	1(5%)	3(6%)
	Failures
	Distant	0	2(5%)	0	2(4%)
	Failures

a Number of patients in each group.

Table 4. Death causes (20 patients died).

	Local failure	1
	Distant failure	4
	General deterioration	12
	Other cancer	3
	Total	20

The Kaplan–Meier disease free survival and overall survival rates at 5 years were 73% and 77%, respectively (Fig. 1, Fig. 2). The crude 5-year survival rate was 43%. According to T stage, Gleason score and PSA level, only 7 cases (10%) could be considered as favourable (T1 or T2 and Gleason score <7 and PSA<10ng/ml). Therefore and due to this very small number, no statistically significant difference could be observed in terms of toxicity and failure rates between these 7 patients with favourable disease status and the 58 others. By both univariate and multivariate analysis, co-morbidities, PSA value, Gleason score, T stage, pelvic radiotherapy and hormonal treatment were not significantly associated with either local or distant failure. The only statistically significant factor observed in the multivariate analysis related to toxicity was the association of pelvic radiotherapy and increases of gastrointestinal toxicity.

View Large Image Download to PowerPoint

Fig. 1. Overall survival curve.

View Large Image Download to PowerPoint

Fig. 2. Event-free survival curve.

4. Discussion 

Treatment of PC in the elderly at ages greater than 80 remains rarely advised due to the perception that this population would not benefit from any curative form of therapy due to their short expected life spans and the perceived indolent nature of their malignancies. Currently, the curative treatment of PC is justified by a combination of different parameters which include the proposed treatment modality (i.e. radical prostatectomy vs. radiotherapy), their consequences in terms of implied morbidity, the evaluated length of life expectancy, the symptoms related to the disease (frequency, hematuria) and the patient's personal demand. Therefore, an in-depth analysis of these criteria should be made individually rather than arbitrarily rejecting any form of aggressive therapy based on age alone. One of such approach is to restrict curative treatment to men with a life expectancy of 10 years or more [8], i.e. any patient aged under 70 years in the majority of western countries. In fact the median non-prostate cancer survival is more than 10 years for men up to age 80 years who are selected and receive curative radiation treatment for prostate cancer [9] However, individual life expectancy is directly derived from both age and co-morbidity. In most physicians’ experience, age alone is often advocated as an argument to favour watchful waiting, especially when the disease is small and strictly limited, the PSA level is under 10ng/ml and the Gleason score is less than 7. From 1993 to 1997, a retrospective population-based analysis of treatments for PC observed that the use of radical prostatectomy decreased by 34% among men aged 70–74 years and 50% for men aged 75 years and more [10]. The fact that a given patient is not eligible for radical prostatectomy would not preclude his ability to tolerate high dose radiation therapy. Remarkably, another recent study form The Netherlands [11] observed that the proportion of patients who received curative radiotherapy for PC varied between 30% and 39% in patients under 79 years (33% in the range 70–79 years) but fell down to 5% in patients older than 80 years. General under-treatment of older men was reported in several papers [3], [12], [13] suggesting that age should not be used as an unique and independent factor in treatment decision making. Obviously, the number and the severity of co-morbidities are often correlated to age. Co-morbidity was shown to be a significant independent factor causing death [14], [15]. A study using a population-base sample of decedents showed that co-morbidity independently predicted death in patients older than 65 years with PC [16]. The likelihood of dying of intercurrent causes was strongly associated with cardiovascular disease [11], [16], [17]. The presence of two or more co-morbid conditions significantly decreased survival [11]. Finally, the ratio of men dying from co-morbidities rather than from PC ranges grossly between one third and one half of the patients [16], [17], [18], [19]. It is difficult to ascertain what proportion of elderly patients will suffer from a lack of definitive treatment for PC. Nevertheless, recent literature suggests that radiotherapy used in a curative fashion may be useful in elderly patients with PC [2], [6], [20] both from the standpoint of improving their survival as well as preserving quality of life by avoiding high levels of treatment toxicity. Geinitz et al. [2] found that high dose conformal radiation therapy to the prostate was well tolerated in patients aged 75 years and older. The survival in older patients was at least as long as it was in younger patients treated during the same period of observation. Improvement in survival of men over 65 with PC treated with radiotherapy was also suggested by Alibhai et al. [13] based on a decision-analytic model. Authors concluded that potentially curative therapy may be advised in reasonably healthy men up to age 80 years who have high grade disease. Good tolerance to curative doses of radiation was formerly demonstrated in 1619 patients enrolled in 9 different EORTC trials initiated from 1975 to 1991 who received pelvic irradiation [4]. Among them, 454 patients with PC were analysed according to age under or over 70, toxicity and survival. There were no trends toward more aged patients to experience gastrointestinal or urinary side effects. Moreover, survival of patients included in PC trials was not different in each age group. External beam radiotherapy efficacy in elderly patients with clinically localised adenocarcinoma was retrospectively reported in a single institution [6] and a small subset of 23 patients over 80 years were compared to 160 men aged 70–79. The 3-year overall, disease specific and bNED survival rates were 62.5%, 66.6%, and 75.2%, respectively while the acute toxicity and late sequellae were comparable in 70–79 and >80 patient groups. Another report from Japan [20] showed that age > or < than 75 years did not influence patterns of external radiotherapy delivering 65Gy, survival and late toxicity in 140 elderly patients compared to 304 younger patients. Among our patients, we observed 6% of grades 2 and 3 proctitis and rectal bleeding while Villa et al. [6] and Geinitz et al. [2] reported 4.3% and 23%, respectively for similar late effects in elderly patients. Nevertheless, increased bowel symptoms after external beam therapy were noted by Fransson and Widmark [21] in 181 patients (mean age 78 years) compared with 141 age-matched controls.

In our study, crude survival rate at 5 years was only 43% which was much less than the Kaplan–Meier disease specific survival estimate of 90%. We think this difference is due to the relatively high number of patients either who died form intercurrent disease or who were lost to follow-up (10 and 8 patients, respectively). Our multicentric retrospective study had several limitations such as lack of predetermined specific inclusion and exclusion criteria leading to highly selected population but these potential drawbacks are consistent with those presented in other previous studies. The participating centres were asked to give their ratio of population ages (<80 and >80 years) for patients newly diagnosed with prostate cancer. The mean ratio was 9.5% (ranging from 6% to 11%). Getting an accurate count of the patients not treated was much harder as about 35–50% of the patients we recommended treatment on were given radiotherapy outside our institutions and the majority of participating centres did not keep internal records of patients observed prospectively. Therefore, event though investigators included all consecutively treated, eligible patients during the timeframe of the study, only 3–7% of men diagnosed with prostate cancer after age 80 during this time period received RT as their primary treatment. Including patients up to 2003 or 2005 would probably gives us a more robust series but would likely have a detrimental effect on the homogeneity of the treatment modalities since at that time, Intensity Modulated Radiation Therapy became more and more popular and preferred to 3D conformal radiation therapy as the standard curative technique in prostate cancer. In our series, the majority of the patients were treated using conformal 3D radiation therapy delivering a mean total dose to the prostate of 69.5Gy instead of IMRT which actually delivers total doses >76Gy. In this selected group of patients referred by urologists for curative radiotherapy, this treatment appeared to be rather well tolerated. In our study, relatively high number of digestive grades 2–3 acute complications were reported in those patients who received a pelvic irradiation compared to those who did not. The authors therefore recommend to avoid prophylactic pelvic irradiation in patients over 80 years of age. Medical co-morbidities, frailty and performance status become significant factors in deciding to pursue a more aggressive treatment in men over the age of 80 with prostate cancer. Different classifications have been proposed to help provide an objective definition of frailty [22], [23], [24]. In considering the management of individual patients over the age of 80 with prostate cancer, it is important to try to answer successive questions in order to choose an appropriate course of therapy. Is the prostate cancer indolent or aggressive? Does it deserve an immediate curative treatment? Is the patient likely to tolerate the complete course of therapy? What are the wishes of the patient and his family with respect to the proposal? The question raised by Fentiman et al. in the Lancet in 1990 “Cancer in the elderly: Why so badly treated” [25] and re-addressed by Wymenga et al. [26] pleads for the development of new fields of research in geronto-oncology.

5. Conclusions 

Radiation therapy given with curative intent was rather well tolerated in our group of patients over 80 years with localised prostate cancer and may positively affect survival. Therefore the authors suggest that radiation therapy with appropriate techniques (i.e. 3D conformal or IMRT) should be considered for elderly patients with localised prostate cancer that are in good health and have a life expectancy of >5 years.

Reviewers 

Bruno Cutuli, MD, Polyclinique de Courlancy, Radiation Oncology, 38 rue de Courlancy, F-51100 Reims, France.

René-Olivier Mirimanoff, MD, CHUV – Centre Hospitalier Universitaire Vaudois, Department of Radiation Oncology, Rue du Bugnon 46, CH-1011 Lausanne, Switzerland.