Errata: Corrigendum Volume 116, Issue 5, E4, Article first published online: 9 October 2015
Stephen E. M. Langley, St Luka's Cancer Centre, The Royal Surrey County Hospital NHS Foundation Trust, Egerton Road, Guildford, Surrey GU2 7XX, UK. e-mail: email@example.com
Study Type – Therapy (case series)
Level of Evidence 4
What's known on the subject? and What does the study add?
• Prostate BXT is an established treatment option for low and intermediate risk localized prostate cancer. Long-term studies have confirmed similar oncological outcomes for low risk disease compared with radical prostatectomy and external beam radiotherapy.
• This study reports long-term toxicity for a large cohort of patients who had been followed-up after LDR BXT for up to 10 years, with a minimum follow-up of 5 years.
• Long-term toxicity results are very encouraging and provide useful date to help clinicians and patients make informed treatment choices.
• To report on the long-term toxicity outcome for patients with prostate cancer treated by low-dose rate (LDR) brachytherapy (BXT).
PATIENTS AND METHODS
• The study population comprised a cohort of men treated in our centre between March 1999 and April 2004 with LDR BXT for prostate cancer who had at least 5 years of follow-up post-implant.
• Patients who had died or experienced biochemical failure were excluded.
• We contacted eligible patients and asked them to complete a questionnaire to assess current urinary, erectile and bowel function.
• Urinary and erectile function was compared pre- and post-treatment and outcomes were assessed by treatment modality.
• Of the 226 LDR BXT-treated patients with >5 years of follow-up, 174 (77.0%) responded to the questionnaire.
• The mean International Prostate Symptom Score (IPSS) increased from 6.70 pre-BXT to 7.91 at follow-up (P= 0.003). Of the patients with mild symptoms pre-BXT (IPSS, 0–7), 64.2% retained mild symptoms at follow-up, 31.2% developed moderate symptoms (IPSS, 8–9) and 4.6% reported severe symptoms (IPSS, 20–35).
• A good or acceptable quality of life (QoL) secondary to urinary symptoms (IPSS QoL, 0–4) was reported by 98.0% of respondents.
• Of those patients potent (International Index of Erectile Function-5 ≥11) pre-BXT, 62.9% remained potent at follow-up. There were no differences in the proportion of patients who were potent when analyzed by the number of years post-implant.
• At follow-up, 51.7% and 45.4% of patients, respectively, had normal or mild bowel symptoms as indicated by the European Organisation for the Research and Treatment of Cancer questionnaire (QLQ-C30/PR25 scores, 4–8). Moderate bowel symptoms (QLQ-C30/PR25 scores, 9–12) were reported by 2.9% of respondents; none reported severe symptoms.
• The present study shows low morbidity after LDR BXT over the long-term for a large cohort of patients.
Since the late 1990s, prostate brachytherapy (BXT) has become an established method of treating organ-confined prostate cancer with an excellent outcome [1–3]. The 125I prostate BXT programme at the St Luke's Cancer Centre in Guildford, UK, began in March 1999 and, subsequently, we have treated more than 1700 patients.
Prostate BXT, conformal beam radiation and radical prostatectomy have similar disease-specific outcomes, as confirmed by studies of the biochemical relapse-free and overall survival rates of patients with organ-confined low-risk prostate cancer [3–5]. Concerns regarding side effects and long-term toxicity, however, ultimately may favour one treatment over others. In the light of those concerns, and as a result of the scarcity of long-term (i.e. ≥5 years) follow-up data, we reviewed the data obtained from patients treated successfully with low-dose rate (LDR) interstitial prostate BXT in our cancer centre. We assessed urinary and bowel symptoms, quality of life (QoL) secondary to urinary symptoms and sexual function for at least 5 years and up to 10 years after treatment and compared these data with pretreatment values where available.
PATIENTS AND METHODS
The initial study population comprised the cohort of patients who had been treated for prostate cancer at our centre between March 1999 and April 2004 (minimum follow-up of 5 years) with 125I LDR BXT, combined BXT and external beam radiotherapy (EBRT), combined hormonal treatment, BXT and EBRT or hormonal treatment with BXT (n= 270).
All eligible patients were sent a questionnaire, which was the same as that used at presentation for evaluating primary symptoms, and asked to return the completed questionnaire by mail. Urinary and potency scores reported at follow-up were compared with the pretreatment values that had been stored routinely on our prospective patient database. Bowel symptoms were assessed at follow-up.
Pretreatment assessment had included medical history, physical examination, TRUS, uroflowmetry, urinary function and potency. Presenting PSA level, Gleason score and clinical stage were also documented.
All patients were treated by the TRUS-guided two stage (Seattle) technique using 125I RAPID Strand™ implants (Oncura, Arlington Heights, IL, USA) with modified uniform planning according to their disease classification at presentation: low-, intermediate- or high-risk disease. The risk grouping system used was based on Memorial Sloan-Kettering Prognostic Index . In general, patients with low-risk disease (stage ≤ T2b, PSA level ≤10 µg/L and Gleason score ≤6) had received BXT alone; patients with intermediate-risk disease (one unfavourable factor, i.e. stage ≥ T2c, PSA level >10 µg/L or Gleason score ≥7) had received 3 months of neoadjuvant androgen deprivation and BXT, and patients with high-risk disease (at least two unfavourable factors) had received a combination of neoadjuvant androgen deprivation for 3 months and EBRT followed by boost BXT. Reasons for patients not receiving treatment according to the usual protocol included the presence of perineural invasion, young age (<60 years) and a delay in the treatment decision or funding application. As part of our treatment protocol, the use of a phosphodiesterase-5 (PDE-5) inhibitor was encouraged routinely, particularly in the first year post-implant, to maintain optimal erectile function. In addition, an α-blocker was prescribed peri-implant.
Exclusion criteria included death from progression of the disease or other causes, or biochemical failure (defined as a PSA level greater than absolute nadir plus 2 µg/L; Houston definition +2) . The remaining patients (n= 226) were considered to have been treated successfully.
SYMPTOM SCORE EVALUATION
Urinary symptoms were categorized according to the IPSS (mild, 0–7; moderate, 8–19; severe, 20–35) . QoL secondary to urinary symptoms was also assessed, with scores of 0–2, 3–4 and 5–6 representing good, acceptable and poor QoL, respectively.
Potency was defined according to an abbreviated five-question version of the International Index of Erectile Function (IIEF)-5 [9,10], with scores in the range 1–25; men who were free from moderate-to-severe erectile dysfunction were defined as potent and, accordingly, an IIEF-5 score ≥11 was used as a threshold for potency [11,12].
Bowel symptoms were evaluated using the European Organisation for the Research and Treatment of Cancer questionnaire (QLQ-C30/PR25) [13,14]. Scores were stratified into normal bowel function (score of 4), mild bowel toxicity (scores of 5–8), moderate bowel toxicity (scores of 9–12) and severe bowel toxicity (scores of 13–16).
Pretreatment scores for urinary symptoms and potency were compared with post-treatment scores using a paired t-test. Frequency distribution graphs were plotted to illustrate potency both pre- and post-treatment.
IPSS scores at follow-up were compared with pretreatment values with an additional analysis of post-treatment IPSS stratified according to the patients' pretreatment IPSS.
Of the 270 patients who had met our initial inclusion criteria, 14 (5.2%) patients had died from progression of the disease or other causes and a further 30 (11.1%) patients were excluded because of biochemical failure. Of the remaining 226 patients, 174 (77.0%) responded to our questionnaire.
The mean (range) age of patients at presentation was 63 (47–75) years, the mean (range) PSA level was 8.4 (1.2–26) µg/L, the median (range) Gleason score was 6 (2–10) and the mean (range) follow-up time was 94.5 (72.9–124.4) months. Patient disease and treatment characteristics at presentation are summarized in Table 1. Almost half of the patient group (48.9%) had been categorized as low-risk at presentation; 12.6% of patients were high-risk. BXT alone was administered to 27.6% of patients, whereas 49.4% had been treated with hormones and BXT and 20.7% received therapy with hormones, BXT and EBRT; the remaining 2.3% had been treated with BXT and EBRT.
Table 1. Patient characteristics at baseline and treatment received
Data were available on urinary toxicity for 174 patients. The mean (sd) IPSS score increased from 6.70 (4.41) pre-BXT to 7.91 (4.93) post-BXT (P= 0.003) (Table 2). In the preoperative IPSS assessment, 62.6% (109/174) of patients had mild symptoms, 36.8% (64/174) had moderate symptoms and one patient (0.6%) had severe symptoms. In the postoperative follow-up, 52.0% of patients had mild symptoms, 43.0% had moderate symptoms and the proportion of patients with severe symptoms was 5.0%.
Table 2. Results of paired t-test for pre- and post-brachytherapy urinary symptoms and erectile function
IIEF, International Index of Erectile Function.
The severity of urinary symptoms at baseline (mild, moderate or severe) affected the patients' urinary symptom status at follow-up. Amongst those patients who presented with mild symptoms, 64.2% (70/109) retained mild symptoms at follow-up, whereas 31.2% (34/109) developed moderate symptoms and 4.6% (5/109) had severe symptoms. Amongst patients who had presented initially with moderate symptoms, 28.1% (18/64) improved and reported only mild symptoms at follow-up, whereas 60.9% (39/64) remained with moderate symptoms and 10.9% (7/64) developed severe symptoms. The one patient who presented with severe urinary symptoms before operation reported an improvement to mild symptoms at follow-up (Fig. 1).
Good, acceptable or poor QoL secondary to urinary symptoms was reported at follow-up for 77%, 21% and 2% of patients, respectively.
Amongst patients treated with BXT alone, the proportion of patients with mild urinary symptoms changed from 70.8% at baseline to 50.0% post-implant (Fig. 2). A greater proportion of this treatment group had moderate symptoms at follow-up compared to baseline (39.6% vs 29.2%) and 10.4% (five patients) developed severe symptoms. For patients treated with BXT and hormones, the proportion with mild urinary symptoms changed from 60.5% at baseline to 54.7% at follow-up, the proportion with moderate symptoms increased slightly (from 39.5% to 41.9%) and three (3.4%) patients developed severe urinary symptoms. Patients treated with BXT, hormones and EBRT had a greater proportion with moderate urinary symptoms at baseline (41.7%) compared to the other treatment modality groups described, and one patient had severe symptoms. At follow-up, the proportion of these patients with moderate symptoms had increased to 47.2% and four (11.1%) patients reported severe urinary symptoms. Too few patients had been treated with BXT + EBRT to warrant meaningful analysis.
Overall, 120 of the 174 patients included in the present study had completed the IIEF-5 questionnaire both pre-BXT and at follow-up. The mean (sd) IIEF-5 score fell from 13.08 (9.99) at baseline to 8.04 (8.59) post-implant (P < 0.001) (Table 2). The frequency distribution in the IIEF-5 scores for patients pre- and post-treatment is shown in Fig. 3. Of those patients (n= 62) who were potent (IIEF-5, ≥11) pre-implant, 62.9% remained potent at follow-up. We found no difference in the proportion of patients potent when analyzed by the number of years post-implant, suggesting that increasing time post-implant is not associated with a decline in potency. There were 48 patients reporting that they used a PDE-5 inhibitor. Because the PDE-5 inhibitor user group included patients considered as impotent and patients considered as potent by the criteria applied, the association between potency and PDE-5 inhibitor use post-BXT is unclear.
Follow-up data on bowel toxicity indicated that 51.7% of patients had normal bowel symptoms and 45.4% had mild bowel symptoms. Only 2.9% of patients reported moderate symptoms and none had severe symptoms.
Bowel toxicity at follow-up stratified by treatment modality is shown in Fig. 4. Of those men who received BXT alone, 60.4% had normal bowel function at follow-up; only 2.1% reported moderate bowel toxicity. Amongst men treated with BXT and hormones, all reported either normal bowel function (55.8%) or mild bowel toxicity (44.2%). Just over half (55.6%) of men treated with BXT, hormones and EBRT reported mild bowel toxicity at follow-up. Again, too few patients had been treated with BXT + EBRT for meaningful analysis.
Studies comparing the outcomes of prostate BXT with those of radical prostatectomy or EBRT have reported equivalent overall survival and biochemical relapse-free survival rates for each of the three treatment modalities in patients with low-risk disease [3–5]. In view of that equivalence, post-treatment QoL secondary to urinary symptoms, maintenance of sexual function and normal bowel function are becoming increasingly important to both physicians and patients when deciding which modality is probably the most suitable. Several studies have reported toxicity outcomes within the first few years after treatment [11,15–27] but there have been fewer longer-term studies of those considerations [3,28–33]. In the present study, we assessed various aspects of QoL and sexual function up to 10 years after prostate BXT, with a good response rate to the request for follow-up information (77%).
We found a small but significant increase in mean IPSS at follow-up. A smaller proportion of patients had mild urinary symptoms at follow-up compared to baseline (52% vs 62.6%), and 5% of men reported severe urinary symptoms at follow-up. Conversely, Gutman et al.  reported urinary morbidity for a similar cohort of patients treated with BXT and found that the proportion of men with mild symptoms increased from 71.5% at baseline to 85.6% at 5-year follow-up, with a marked fall in the proportion of men with moderate symptoms (from 27.8% to 13.4%). In another study, Schafer et al.  found that, after a median follow-up of 51 months post-BXT, only 10.4% and 1% of men reported moderate or severe urinary symptoms, respectively. It is possible that the less favourable urinary morbidity at follow-up in the cohort of men in the present study is a consequence of the increased duration of follow-up (up to 124 months), namely the effect of age  and the possibility that any long-standing mild obstruction would upgrade the patient symptom score into the moderate category. Ash et al. , however, reported that mean IPSS returned to within one or two points of the pretreatment level at 1 year post-BXT, and that no subsequent deterioration in urinary symptoms was observed up to 9 years after treatment (median 4.9 years).
Chen et al.  have suggested that baseline status is the most powerful predictor of QoL outcomes for men with prostate cancer  and that analysis of post-treatment outcomes stratified by pretreatment functional status allows for a better understanding of the probable impact of treatment choices on patients' QoL. We found that around two-thirds (64.2%) of men with mild urinary symptoms pretreatment reported mild symptoms at follow-up, whereas 28.1% of men with moderate symptoms at baseline reported improvement in urinary symptoms. Urinary ‘flare’ is a common but transient problem after BXT [24,36] and the time to resolution of IPSS post-treatment has been associated with baseline IPSS . We have found that urinary morbidity over the long-term is also related to baseline IPSS, which is in agreement with the findings of a previous study from our centre .
For most of the patients in the present study, we were able to compare preoperative erectile function with potency 5–10 years later; 62.9% of patients who had been potent before operation retained their potency. The finding is more favourable than the potency outcomes reported in other studies. Merrick et al.  reported BXT-induced erectile dysfunction for more than 50% of patients at 3 years post-treatment and a recent study by Taira et al.  reported that, of 124 patients potent at baseline, 69 (55.6%) reported 7-year actuarial potency preservation. The finding of the present study was similar, however, to that reported by Stone and Stock  who found a potency preservation of 61.5% at ≥5 years follow-up for men treated with BXT. More recently, Merrick et al.  reported a potency preservation of 75.6% (without mechanical or pharmacological support) at a mean follow-up of 5.6 years after BXT amongst men aged ≤50 years. There were too few patients aged ≤50 years in the present study cohort to allow meaningful analysis of potency stratified by age.
The patients in the present study were treated using the conventional BXT technique whereby seeds are implanted in a modified-uniform pattern. We have reported previously that patients treated using this modified-uniform pattern BXT technique have a 2-year potency of 61.7% . We have now modified our BXT implant technique so that stranded seeds are placed peripherally and loose seeds are placed centrally, thereby reducing the dose to the penile bulb. Using the new technique, we have shown that potency is improved to 83.3% at 2 years , and we expect that it will be equally durable. The results obtained with the new technique confirm previous observations of an association between the dose delivered to the penile bulb and improved potency outcome [31,38]. We consider that the new technique will offer even better potency outcomes for patients over the longer term.
The use of a PDE-5 inhibitor can improve potency outcomes for men post-BXT and men treated at our centre are prescribed a PDE-5 inhibitor routinely. Merrick et al.  reported that pharmacological support can improve potency outcomes to 92% post-BXT. Tsui et al.  reported that, with pharmacological intervention, only 9% of potent patients who underwent BXT report a loss of potency by 18 months.
We did not find that potency preservation declined as the time since treatment increased and the potency preservation data currently available for BXT are generally encouraging. However, although the use of the IIEF questionnaire is recognized as an accurate and reliable method for collecting QoL data , the patient is required to have been sexually active at some time during the preceding month to achieve a ‘potent’ score,. Thus, an overestimation of erectile dysfunction may arise by counting abstinent patients as impotent, which is particularly relevant for patients susceptible to a natural age-related decline in sexual activity. Indeed, Miller et al.  compared outcomes at a median follow-up of 6 years amongst men treated for prostate cancer and controls and reported deterioration in potency amongst controls (P= 0.014), as well as amongst patients treated with EBRT (P= 0.002), whereas Stock et al.  reported a fall in the actuarial preservation of potency with BXT from 79% at 3 years to 59% at 6 years.
For the present study, we did not collect data on comorbidities and lifestyle variables (e.g. diabetes mellitus, cardiovascular disease and smoking) that could impact negatively on erectile function. It is possible that the presence of any of these factors amongst the patients described in the present study could have influenced the potency rates observed and we recommend their inclusion in future analyses.
The interpretation of data of bowel symptoms at follow-up is limited by the fact that we did not have baseline data available with which to make any comparison. Nevertheless, bowel toxicity at follow-up was low: only 2.9% of patients reported moderate bowel toxicity and none reported severe toxicity. These findings are consistent with those of previous studies [21,23,26–28,32].
Overall, the results obtained in the present study confirm that BXT has a favourable side effect profile over the long-term with regard to potency, urinary and bowel toxicity. Although urinary symptoms will worsen slightly during follow-up, this does not appear to affect the patients' QoL. These findings will help clinicians to counsel their patients on the probable long-term QoL outcomes of treatment for their prostate cancer. Improvements in BXT techniques (i.e. real-time planning and modification of the dose to the proximal penile structure) should allow even better outcomes for patients treated with BXT in the future. These expectations will be addressed in forthcoming studies.
The authors would like to thank Dr Helen Seaman for her assistance with the preparation of this manuscript for publication.