Prostate-specific antigen bounce after curative brachytherapy for early-stage prostate cancer: A study of 274 African-Caribbean patients

Brachytherapy 14 (2015) 826e833

Prostate-specific antigen bounce after curative brachytherapy for early-stage prostate cancer: A study of 274 African-Caribbean patients N. Leduc1,*, V. Atallah1, M. Creoff1, N. Rabia2, T. Taouil2, P. Escarmant1, V. Vinh-Hung1 1

Department of Radiation Oncology, University Hospital of Martinique, France 2 Department of Surgical Urology, University Hospital of Martinique, France

ABSTRACT

BACKGROUND: Prostate cancer incidence in the African-Caribbean population ranks among the highest worldwide. We aim to evaluate the prostate-specific antigen (PSA) kinetics after brachytherapy, which so far remains unknown in this population. METHODS AND MATERIALS: From 2005 to 2013, 371 patients received 125I brachytherapy of 145 Gy for early-stage prostate cancer. Eligibility criteria were cTNM #T2c, Gleason score #7, and initial PSA #15 ng/mL. Pretreatment androgen deprivation therapy was allowed. PSA bounce was defined as an increase of $0.4 ng/mL, lasting $6 months, followed by a decrease without any anticancer therapy. We examined PSA kinetics during followup. RESULTS: For the 274 patients with at least 24 months followup, median age was 62 years old (range, 45e76). Initial PSA was !10 ng/mL in 244 and 10e15 ng/mL in 30 patients; 40 received androgen deprivation therapy. With a median followup of 50 months (range, 24e125), PSA declined continuously in 168 (61%) patients, bounced in 87 (31%), and initially declined and then rose in 22 (8%) patients. Among these latter patients, 18 presented clinical recurrence. Mean bounce intensity was 2.0 ng/mL (median, 1.2; range, 0.4e12.4). Bounces occurred in average 12 months after brachytherapy. Patients with bounce were significantly younger: mean age 59 vs. 63 years old in patients without bounce, p !0.001. Bounce was also significantly associated with the immediate post-brachytherapy PSA, mean 4.0 among bounce cases vs. 2.9 among non-bounce cases, p ! 0.001. Bounce was not associated with recurrence. CONCLUSIONS: PSA bounce in our African-Caribbean population seemed earlier and was more intense than described in other populations. Early increase of PSA should not be ascribed to treatment failure. Ó 2015 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved.

Keywords:

Prostate cancer; PSA bounce; African-Caribbean

Introduction The incidence of prostate cancer in French West Indies (Martinique, Guadeloupe) ranks among the highest worldwide with a raw level of 268 to 259 of 100,000 and a standardized level of 163.7 of 100,000 from 2008 to 2010 (1). These numbers keep increasing and do not seem to plateau yet. Incidence and mortality for this pathology are significantly higher in the French Caribbean than in French mainland (2). In Martinique, all patients presenting with Received 4 August 2015; received in revised form 9 September 2015; accepted 11 September 2015. * Corresponding author. Department of Radiation Oncology, University Hospital of Martinique, BP 632, 97261 Fort-de-France, France. Tel.: þ33-46-850-6750; fax: þ596 596755060. E-mail address: [email protected] (N. Leduc).

early-stage disease are treated according to French recommendations within our state-financed equal-access health care system. A wide range of curative treatments, including 125 I brachytherapy, are available. Prostate permanentimplant brachytherapy is a widely used, efficient technique for the treatment of early-stage prostate cancer (3). Prostate-specific antigen (PSA) levels have been commonly used as a sensitive surrogate for the measurement of outcomes after brachytherapy and the detection of possible failure of treatment. In contrast to PSA levels that can be measured after surgical removal of the gland, those after brachytherapy usually decrease over 2e5 years to their lowest detectable or undetectable level (nadir). It is of common knowledge among involved physicians that biochemical disease-free survival requires stable and low PSA levels (4, 5). As such,

1538-4721/$ - see front matter Ó 2015 American Brachytherapy Society. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.brachy.2015.09.005

N. Leduc et al. / Brachytherapy 14 (2015) 826e833

noticeable increase of PSA level after treatment tend to be interpreted as a biochemical failure, leading physicians into increasing the frequency of clinical and biochemical followup and possibly the occurrence of additional invasive examinations or treatments. However, it has been widely reported since the first observation by Wallner et al. (6) that patients can present a benign and transient elevation of PSA several months or years after brachytherapy. The etiology and predicting factors of this ‘‘bounce’’ or ‘‘spike’’ remain largely unknown. This unexpected PSA bounce can be a cause of anxiety for the patient and his physician and could be the source of undue subsequent local or general treatment (7). This is all the more true that the patients in our care are almost exclusively of African descent. Indeed, it has often been argued that prostate cancer in African-American population has a poorer prognosis. Whether the disparities come from the differences of socioeconomic status and access to health care or from a physiological cause has been much discussed, and both phenomena are probably intricate (8e 12). As a consequence, and despite the equal-access system our patients enjoy, we still consider our patients to be at high risk of recurrence, and it is still in our habit to scrutinize PSA levels after brachytherapy, especially when a PSA elevation is noted. Therefore, establishing the incidence and levels of PSA bounce compared with biochemical failure is a question of importance for adequate management of our population. In this study, we evaluated the biochemical outcome and successive PSA levels of 274 consecutive patients treated by permanent-implant brachytherapy to determine the rate of occurrence of PSA bounce and its predictive factors and differentiating factors from biochemical failure.

Materials and Methods Patients We reviewed the records of all 371 consecutive patients who received 125I brachytherapy as a curative treatment for early-stage (localized) disease from 2005 to 2014 in the University Hospital of Martinique, France. All patients were treated according to the French Urology Association guidelines (13, 14). Selected patients presented with low risk, early-stage disease: initial PSA (iPSA) !10 ng/mL, clinical stage #T2c, and Gleason !7. Patients with intermediate risk of recurrence were also included on a caseto-case basis with PSA !15 or Gleason 7 (3 þ 4). All patients were clinically staged by medical history, clinical examination (including digital rectal examination), and initial PSA level determination. All patients received a systematic pretreatment endorectal MRI to detect local and regional extension. Patients with intermediate risk were also prescribed a bone scintigraphy and thoracic, abdominal, and pelvic CT scans. Patients with capsular or

827

seminal extension found on the MRI, and patients with systemic extension found on CT scans or bone scintigraphy, were excluded from brachytherapy. Pretreatment androgen deprivation therapy was allowed for 3 months for patients with high prostate volume (O50 cm3) or altered urinary functions (IPS score O 19). Brachytherapy technique Thirty patients from 2005 to 2007 received linked seeds I brachytherapy (IMC7000 RAPID Strand seed; Oncura, Amersham, Buckinghamshire, UK). For all other subsequent patients, a real-time, ultrasound-guided planning technology with loose radioactive permanent implants of isotope 125I was used (BEBIG Isoseed I125; Eckert & Ziegler BEBIG GmbH, Berlin, Germany). The prescribed dose was systematically 145 Gy. Intraoperative D90 was O145 Gy for all patients and mean V100 was 99.1%. Planning, treatment, and dosimetric calculation techniques were identical for all patients.

125

Followup All patients received long-term followup in accordance to the French recommendations: a first consultation 2 months after treatment, then every 6 months for 3 years, and once a year afterward. PSA level determination was prescribed for each consultation. A post-planning MRI was prescribed 1 month after brachytherapy. Followup was performed in our radiation therapy ward. We restricted the statistical analysis to the patients with O24 months of followup. Statistical analysis All data were obtained from an institutional registry. The insularity of Martinique provides low rates for patients lost to followup. Besides, the limited amount of laboratories provides consistent data for PSA values. We assessed the ethnicity of our patients using their phototype and place of birth. To this day, there is no consensual definition for PSA bounce. A variety have been used in scientific literature with threshold values from 0.1 ng/mL (15) to 0.4 ng/mL (7, 16), followed by a decrease of any value or a decrease to PSA values lower than pre-bounce levels. As the standard deviation of PSA test is usually measured around 0.1 ng/mL, considered threshold values should not be lower (17). As a consequence, we identified three PSA bounce definitions that seem more commonly used: - a 0.4 ng/mL bounce for O6 months, followed by any decrease (Definition 1) (7, 16). - a 0.2 ng/mL bounce for O6 months, followed by any decrease (Definition 2) (18). - 0.2 ng/mL bounce followed by a subsequent decrease to lower value than previous PSA nadir (Definition 3) (19e21).

828

N. Leduc et al. / Brachytherapy 14 (2015) 826e833

Our study bears strong similarities with the series published by Toledano et al. (7): identical national treatment and followup guidelines, brachytherapy apparatus, and technique. As a consequence, Definition 1, as used by Toledano et al. (7), will be considered in the following for comparison purposes, unless otherwise mentioned. The date for bounce was recorded as the date of the first rise of PSA after nadir. Biochemical failure was defined according to the ASTRO nadirþ2 definition (22) provided that PSA values do not recede under that level on subsequent measurements. The clinical, biochemical, and demographic parameters of bouncing patients (Group B) and not bouncing patients (Group A) were analyzed. Chi-square analysis, Fisher’s exact test, and Student’s t test for Pearson coefficient were used to assess the significance of the univariate comparisons. Logistic regression multivariate analysis was performed to determine independent predicting factors of PSA bounce. Statistical significance was set at p !0.05 for all analyses. Matlab R2013a (MathWorks Inc, Natick, MA) was used for all statistical computations.

Results Patients Three hundred seventy-one patients were treated with brachytherapy in Martinique from 2005 to 2014. Among them, 274 have been followed for O24 months. Mean age was 62 years (45e76) at the time of implantation. Clinical stages were T1c (70%), T2a (15%), T2b (12%), and T2c (1%). Two hundred forty-four patients (89%) presented with an iPSA !10 ng/mL and 30 patients (10%) with an iPSA 10e15 ng/mL. Initial PSA was inferior to 15 ng/mL for all patients. Overall, 215 of our patients (79%) presented with low-risk prostate cancer, 55 (20%) presented with intermediate risk, and 4 (1%) with high risk. All four patients with high risk of recurrence were classified as such because of a hardened gland at the time of initial digital rectal examination. Forty patients were prescribed pretreatment androgen deprivation therapy (ADT) for 3 to 6 months to decrease the prostate volume. The median followup was 50 months (range, 24e126). The median number of PSA values per year for each patient was 7 (mean, 8.2; i.e., one value determination every 5 months in average). Mean gland volume was 30.2 cm3 (median, 30; minimum 12.1, maximum 59 cm3). Throughout followup, 168 patients (61%) showed a continuous decrease of PSA (with possible intermittent elevation !0.4 ng/mL); 87 (31%) showed a transient PSA increase of at least 0.4 ng/mL followed by any decrease (Definition 1), 106 (39%) showed a transient PSA increase of at least 0.2 ng/mL followed any decrease (Definition 2), and 85 (31%) patients showed a transient PSA increase of at least 0.2 ng/mL with a subsequent decrease to former

values (Definition 3). A decrease to nadir followed by a continuous increase to this date occurred in 22 patients (8%). Among them, 18 proved clinical recurrences have been found; others are still in followup and do not meet the ASTRO criteria of biochemical failure. Thirty-nine (14%) patients met the usual ASTRO criteria of biochemical failure: increase of PSA value over nadir þ2 ng/mL. Of those patients, 18 presented a longlasting PSA elevation, true biochemical failure, and finally showed clinical or image failure (17 patients O18). The 21 others subsequently showed a continuous decrease of PSA and met the criteria of benign PSA bounce. Three clinically failed patients had previously met the criteria of PSA bounce. There were less biochemical failures among patients with PSA bounce than among patients without bounce (three over 87 patients 5 3% vs. 15 over 187 patients 5 8%, p 5 0.15), but this result was not statistically significant. After exclusion of patients receiving neoadjuvant ADT, 78 (33%) showed a transient increase of PSA value according to Definition 1, 94 (40%) according to Definition 2, and 76 (32%) according to Definition 3. No significant difference could be found with bounce rate in the full pool ( p 5 0.8). Bounce characteristics The mean bounce peak value was 2.0 ng/mL (median, 1.2; range, 0.4e12.4) for definition 1 (0.4 ng/mL threshold), 1.7 ng/mL (median, 1.1; range, 0.2e12.4) for definition 2, and 1.6 ng/mL (median, 1.0; range, 0.2e12.4) for definition 3. Initial bounce was observed at a mean 13 months (median, 12 months; minimum, 6 months; maximum, 37 months), whereas first PSA increase before biochemical failure was found at a mean 36 months (median, 35 months; minimum, 14 months) as plotted on Fig. 1. The difference was statistically significant ( p ! 0.001) and those results remain accurate within 1 month, whatever the definition for PSA bounce. The peak times were observed at a mean 15 months after brachytherapy (median, 15 months; range, 0.6e45). iPSA was not associated with bounce peak values ( p 5 0.7). Age was not associated with PSAB peak values ( p 5 0.6) and the PSA bounce peak times did not differ significantly ( p 5 0.8) before or after 65 years old. Excluding patients who received ADT did not significantly modify those results, including time to bounce: the mean bounce peak value was 2.0 ng/mL (median, 1.2; range, 0.4e12.4) for definition 1. Initial bounce was observed at a mean 12 months (median, 12 months; minimum, 8 months; maximum, 37 months). Figure 2 plots the kinetics after implantation of the mean PSA value for five different groups of patients depending on their bounce and ADT status. Mean PSA values for identified biochemical failures were also plotted until the first of them started increasing again. The difference and ratio between iPSA value and first PSA value after implantation are summarized in Table 1

N. Leduc et al. / Brachytherapy 14 (2015) 826e833

829

Fig. 1. Mean PSA value during followup; 274 patients stratified according to their bouncing/not bouncing status and ADT/ADT-free status. Also plotted: mean PSA value of failed patients from initial treatment time to first rise of PSA of any of them (no subsequent statistical significance). Method: piecewise cubic Hermitte interpolating polynomial. ADT 5 androgen deprivation therapy; PSA 5 prostate-specific antigen.

for patients receiving or not ADT and failing patients. The mean PSA difference is significantly lower for bouncing, ADT-free patients compared with not bouncing (3.4 vs. 4.7, p ! 0.001) and failed patients (3.4 vs. 3.9, p 5

0.05). The curves for failed patients and bounce-free patients were not significantly different ( p 5 0.6). Results of the same magnitude can be showed from mean ratio. Predictive factors of PSA bounce Table 2 summarizes the characteristics of the treated population stratified by the Bouncing/Not Bouncing status. Statistical analysis is performed to assess the significance of the observed difference between the two groups. Patients experiencing PSA bounce were 4 years younger than patients with no PSA bounce (median, 59 vs. 63 years old, p ! 0.001). Androgen deprivation therapy (ADT) was given to 40 patients for 3 to 6 months before brachytherapy to decrease the prostate volume. Bounce occurred in 9 of 40 (24%) patients receiving ADT compared with 87 of 274 (31%) who did not receive ADT. The difference was not statistically significant ( p 5 0.7). Table 1 ‘‘iPSA value/FP value ’’ ratio and ‘‘iPSA value  FP’’ value difference iPSA  FP

iPSA/FP Subgroups of patients No ADT, no bounce No ADT, bounce p value Failed Fig. 2. Distribution of magnitude of PSA bounce, thresholds 0.2 and 0.4 ng/ mL, any subsequent decrease, n 5 106). PSA 5 prostate-specific antigen.

Mean 4.7 3.4 !0.001 3.9

Median 2.4 1.6 2.3

Mean 4.0 2.0 !0.001 4.3

Median 3.8 1.8 3.8

iPSA 5 initial prostate-specific antigen; FP 5 first PSA after treatment.

830

N. Leduc et al. / Brachytherapy 14 (2015) 826e833

Table 2 Population characteristics and univariate predictors of bounce All patients N 5 274 Factor Age Mean Median Minimum Maximum iPSA Mean Median Minimum Maximum Prostate volume Mean Median Minimum Maximum Androgen deprivation Yes No Gleason score !7 7 missing cTNM cT1c cT2a cT2b cT3a missing pTNM pT2a pT2b pT2c missing Dosimetric values D30 Mean Median D90 Mean Median V100 Mean Median IPS score !7 !19 O19 Missing Mean Median

n

No PSA bounce n 5 187 %

n

PSA bounce n 5 87 %

n

%

Univariate p value !0.001

62 62 45 76

_ _ _ _

63 63 48 76

_ _ _ _

59 59 45 75

_ _ _ _

6.8 6.3 0.2 15.0

_ _ _ _

6.9 6.6 1.6 12.1

_ _ _ _

6.5 6.0 0.2 15.0

_ _ _ _

30 30 12.1 59

_ _ _ _

30 30 15 51

_ _ _ _

30 29 15 59

_ _ _ _

40 234

14 86

31 156

16 83

9 78

10 89

241 32 1

88 12 0

165 21 0

88 12 0

76 10 1

87 13 0

194 42 32 1 5

70 15 12 1 2

134 24 23 1 5

71 12 11 1 3

60 18 9 0 0

69 20 10 0 0

108 72 93 1

39 26 34 0

71 51 64 1

38 27 35 0

37 21 29 0

42 24 34 0

185.5 186.3

_ _

186.4 186.9

_ _

183.3 183.7

_ _

181.5 181.5

_ _

181.3 181.8

_ _

182.0 181.4

_ _

99.1 99.4

_ _

99.1 99.4

_ _

99.0 99.4

_ _

230 34 0 11 3.2 3

84 12 0 4 _ _

151 28 0 8 3.5 3

81 15 0 4 _ _

79 6 0 2 2.7 2

90 7 0 2 _ _

0.049

0.73

0.73

0.65

0.23

0.53

0.01

0.44

0.56

0.16

0.03

iPSA 5 initial prostate-specific antigen; cTNM 5 clinical TNM evaluation of tumor; pTNM 5 histological TNM ranking of biopsy; D30, D60 5 dose received by 30%, 100% of prostate; V100 5 volume of prostate receiving 100% of prescribed dose; IPS 5 International Prostate Symptom.

Prostate volume was not significantly different between patients with or without bounce (mean 30.4 vs. 30.2, p 5 0.73). Dosimetric parameters did not show significant differences for main parameters V100 or D90, but D30 was lower in bouncing patients than in bounce-free patients (183.3 vs. 186.4, p 5 0.01).

No significant difference was found regarding initial tumor Gleason score (76 [87%] ! 7 vs. 165 [88%] ! 7) or cTNM, but median iPSA for bouncing patients was slightly lower (6.5 vs. 6.9, p ! 0.05). Two hundred sixty-three patients provided an initial International Prostate Symptom (IPS) score, whose median value

N. Leduc et al. / Brachytherapy 14 (2015) 826e833

831

Fig. 3. Distribution of time to first PSA rise: bouncing patients (n 5 87)/failed patients (n 5 18); normal interpolation of data. PSA 5 prostate-specific antigen.

was 3. In univariate analysis, altered urinary functions (IPS score O 7) were not associated to a significantly lowered bounce rate ( p 5 0.15). However, bouncing patients had a lower mean IPS score (3 vs. 3.7, p 5 0.03). This association was not significant anymore after adjustment to age at the time of treatment ( p 5 0.07).

Discussion We present new findings in the African-Caribbean population regarding PSA bounce. Bounce characteristics The median increase above nadir is significantly higher than what was expected reading from the scientific literature: with 1.2 ng/mL (Definition 1) or even 1.1 ng/mL (Definition 3), it is higher than the median 0.4 ng/L computed by Merrick et al. (21), the 0.76 ng/mL (0.5e0.9 ng/mL) calculated by M
adan et al. (23) in their review from five studies (24e28), or the highest occurrence of 0.8 ng/mL found by Toledano et al. (7). Mean time to occurrence is also shorter, with a median value of 12 months to the first increase (Definitions 1, 2, or 3). Values recorded in literature range from 14 (27, 28) to 19 months (7). As can be inferred from Fig. 3, bounce magnitude ranges from 0.2 to 12.4 ng/mL, describing very different clinical situations to unaware physicians.

Unlike what has been found for instance by Crook et al. (25), a non-negligible number of PSA bounces occur very early (2 months after brachytherapy) in our population, perhaps because of inflammatory reaction just after implantation. Because our team prescribes early determination of PSA after brachytherapy, we might incidentally record a higher number of early bounces than other authors. It is, however, reasonable to assume the bounce distribution in time to be normal (Lilliefors test for normal fitting value 5 0, p ! 0.05). Under that assumption, we observed that the median first PSA rise for failed patients occurs later than the median PSA bounce (median, 36 vs. 12 months, p ! 0.001). As already found by Ciezki et al. (19), median PSA bounce occurs no !2 years before a typical biochemical failure. The time span to the first rise in PSA is, therefore, a good discriminator between a benign PSA bounce and a possible future biochemical failure. Three outliers, however, presented a first rise of PSA bounce before showing a second PSA increase to a true biochemical failure. Besides, as can be inferred from Fig. 1, PSA kinetics of subsequently failed patients can overlap PSA kinetics of bouncing patients. Differences between PSA bounce definitions Results in terms of bounce rate are significantly different depending on the level of PSA increase specified by the definition. The frequency of a 0.2 ng/mL PSA bounce in

832

N. Leduc et al. / Brachytherapy 14 (2015) 826e833

literature varied between 13% and 50% with an average of 35.8% in a review of previously published series of PSA followup (23). Our observed 23% rate lies within this range. It is, however, more relevant for us to compare with the study of Toledano et al. (7). Indeed, they performed curative brachytherapy according to the same French guidelines, selection rules, and with identical setup and techniques as the series we present here. The relevant and interesting difference is that they consider a population that is mainly Caucasian. The frequency of PSA bounce, defined as a 0.4 ng/mL elevation for Toledano et al., reaches 32% in their population against 31% in ours. The definitions for PSA bounce do not cover all situations. For instance, PSA level curves for several patients showed a rise well over the required levels for PSA bounce but stabilized for many years around PSA levels higher than the previous nadir. Therefore, they did not meet all the criteria of Definition 3. But they did not meet the criteria of biochemical failure either because PSA level did not rise over nadir þ2 ng/mL. Some other patients showed a PSA bounce in the sense of Definition 1 but ended up later with a biochemical and clinical failure. In that sense, no PSA bounce definition seems to be perfectly adequate to discriminate against biochemical failure. Androgen therapy Androgen deprivation therapy was not associated to an increased PSA bounce rate with 0.4 ng/mL cutoff, but intermittent elevation under this threshold was seen with almost all patients (87%). It has been argued by Merrick et al. (21) that post-ADT median bounce is of low amplitude (median, 0.1 ng/mL), but we find on the contrary that median PSA increase (if any) after ADT reaches 1.1 ng/mL (mean, 2.2) against 0.2 ng/mL only in ADT-free population. This result is not statistically significant because of the low rate of bounces in ADT patients. This discrepancy in literature could be ascribed to the flexible time span between ADT injection and the actual brachytherapy. Hormone level and, therefore, post-brachytherapy PSA values may differ depending on the length of time since ADT last injection. Time to bounce was not lower for patients with ADT (12 vs. 12 months in ADT-free population) with a 0.4 ng/mL threshold but was significantly decreased with a 0.2 ng/ mL threshold (9 vs. 12 months, p 5 0,04). As explained before, almost all ADT patients show discreet increase of PSA levels just immediately after brachytherapy. This is probably a consequence of the early wearing off of the anti-testosterone effect: almost all our patients (82 of 90) received a 3-month ADT O4 months before implantation. Testosterone levels usually increase to their former levels shortly after ADT wears off, but this time span could be shortened in the African-Caribbean male population, who exhibits higher testosterone levels than Caucasian population (29). PSA rises linked to ADT remain small and

limited in time. They cannot clinically be mistaken for biochemical failure. Predicting factors of PSA bounce Discriminating early PSA bounce from future biochemical failures is of crucial importance in the followup of our patients. A few factors measured in our patients are strongly correlated to the occurrence of PSA bounce. Age As already shown in Caucasian populations by Toledano et al. (7), Stock et al. (16), Patel et al. (30), and Merrick et al. (21), PSA bounce is significantly associated with age younger than 65 years. This could be explained by the reactivation of epithelial cells in younger patients because of increased testosterone levels. Other authors argued that younger patients are more likely to have experienced recent ejaculation. Other volume and dosimetric or staging parameters do not exhibit meaningful association to the occurrence of PSA bounce, which is consistent with the findings of most authors. Initial PSA decrease as a predicting factor of future PSA bounce Merrick et al. (21) have already explored the early decrease of PSA values, showing that the mean first postimplant PSA level was significantly higher in the spike than in nonspike patients (1.2 vs. 0.7 ng/mL, p ! 0.001). Our observation is similar (4.0 vs. 2.9, p ! 0.001) but performed at an earlier date (2.7 months after implantation in average vs. 3e6 months). As plotted on Fig. 2, bounce-free, ADT-free patients exhibit first-order PSA kinetics. PSA values for ADT patients increase from almost zero at implantation time. As can be inferred from Fig. 2, the initial decrease of PSA after seed implantation exhibits peculiar patterns: until roughly 10 months after implantation, PSA levels for failing and bounce-free, ADT-free patients decrease much faster than bouncing, ADT-free patients. As explicated on Table 1, initial PSA difference and PSA ratio seem to be good predictive factors for the appearance of future bounce during followup. However, a number of outliers prevent us to use this parameter as a differentiating factor between bounce and biochemical failure. PSA bounce as a predicting factor of biochemical failure. True biochemical failure (with proved clinical failure and/ or without subsequent decrease of PSA values) seems to be slightly more frequent in the bounce-free group (5% vs. 4%, p 5 0.4). A few authors (30, 31) have argued PSA bounce could be a factor predicting success of brachytherapy. Although our results tend to the same direction, it is necessary to be cautious in conclusions, for our series have short followup and rare occurrence of biochemical failure.

N. Leduc et al. / Brachytherapy 14 (2015) 826e833

Conclusions To our knowledge, there is no large study specifically exploring the PSA kinetics of brachytherapy in patients of African descent. Some have explored these patients as small subgroups of limited numbers. In our study, African-Caribbean patients seem to exhibit a posttreatment PSA profile quite similar to what can be found in Caucasian populations. PSA bounce rate does not differ significantly. However, bounce is clearly more intense and seems to happen earlier. This could be an intrinsic characteristic of African-Caribbean patients. The results found in our study do not call for a modification of followup frequency for African-Caribbean patients. However, keeping the PSA bounce phenomenon in mind when confronted to very early and intense rising of PSA value should be advised in this population to avoid further unnecessary explorations or treatment. References [1] Joachim C, Macni J, Veronique-Baudin J, et al. Epidemiologie du cancer de la prostate aux Antilles-Guyane: Donnees des registres generaux des cancers. Bull de veille sanitaire Antilles-Guyane 2013;3e5. [in French]. [2] Binder-Foucard F, Belot A, Delafosse D, et al. Estimation nationale de l’incidence et de la mortalite par cancer en France entre 1980 et 2012. Partie-1 Tumeurs solides. Saint-Maurice, France: Institut de veille sanitaire; 2013. p. 122. [3] Zelefsky MJ, Kuban DA, Levy LB, et al. Multi-institutional analysis of long-term outcome for stages T1eT2 prostate cancer treated with permanent seed implantation. Int J Radiat Oncol Biol Phys 2007;67:327e333. [4] Cox JD, Gallagher MJ, Hammond EH, et al. Consensus statements on radiation therapy of prostate cancer: guidelines for prostate rebiopsy after radiation and for radiation therapy with rising prostatespecific antigen levels after radical prostatectomy. American Society for Therapeutic Radiology and Oncology Consensus Panel. J Clin Oncol 1999;17:1155. [5] D’Amico AV, Whittington R, Malkowicz SB, et al. Biochemical outcome after radical prostatectomy, external beam radiation therapy, or interstitial radiation therapy for clinically localized prostate cancer. JAMA 1998;280:969e974. [6] Wallner KE, Blasko J, Dattoli MJ. Evaluating cancer status. In: Wallner KE, Blasko J, Dattoli MJ, editors. Prostate brachytherapy made complicated. 2nd edition. Seattle, WA: SmartMedicine Press; 2001. [7] Toledano A, Chauveinc L, Flam T, et al. PSA bounce after permanent implant prostate brachytherapy may mimic a biochemical failure: A study of 295 patients with a minimum 3-year followup. Brachytherapy 2006;5:122e126. [8] Chornokur G, Dalton K, Borysova ME, et al. Disparities at presentation, diagnosis, treatment, and survival in African American men, affected by prostate cancer. Prostate 2011;71:985e997. [9] Hoffman RM, Gilliland FD, Eley JW, et al. Racial and ethnic differences in advanced-stage prostate cancer: the Prostate Cancer Outcomes Study. J Natl Cancer Inst 2001;93:388e395. [10] Moul JW, Douglas TH, McCarthy WF, et al. Black race is an adverse prognostic factor for prostate cancer recurrence following radical prostatectomy in an equal access health care setting. J Urol 1996; 155:1667e1673. [11] Ward E, Jemal A, Cokkinides V, et al. Cancer disparities by race/ethnicity and socioeconomic status. CA Cancer J Clin 2004; 54:78e93.

833

[12] Yamoah K, Stone N, Stock R. Impact of race on biochemical disease recurrence after prostate brachytherapy. Cancer 2011;117:5589e 5600. [13] Soulie M, Beuzeboc P, Richaud P, et al. Bulletin de synthese de veille 2005 Recommandations pour la pratique clinique: Prise en charge therapeutique du cancer de la prostate non metastatique. Prog Urol 2007;17:801e809. [in French]. [14] Salomon L, Bastide C, Beuzeboc P, et al. Recommandations en oncourologie 2013 du CCAFU: Cancer de la prostate. Prog Urol 2013;23: S69eS101. [15] Critz FA, Williams WH, Benton JB, et al. Prostate specific antigen bounce after radioactive seed implantation followed by external beam radiation for prostate cancer. J Urol 2000;163:1085e1089. [16] Stock RG, Stone NN, Cesaretti JA. Prostate-specific antigen bounce after prostate seed implantation for localized prostate cancer: descriptions and implications. Int J Radiat Oncol Biol Phys 2003;56:448e 453. [17] Pruthi RS. Editorial: prostate specific antigen bounce after radioactive seed implantation followed by an external beam radiation for prostate cancer. J Urol 2000;164:2031. [18] Cavanagh W, Blasko JC, Grimm PD, et al. Transient elevation of serum prostate-specific antigen following (125) I/(103) Pd brachytherapy for localized prostate cancer. Semin Urol Oncol 2000;18:160e165. [19] Ciezki JP, Reddy CA, Garcia J, et al. PSA kinetics after prostate brachytherapy: PSA bounce phenomenon and its implications for PSA doubling time. Int J Radiat Oncol Biol Phys 2006;64:512e 517. [20] Caloglu M, Ciezki JP, Reddy CA, et al. PSA bounce and biochemical failure after brachytherapy for prostate cancer: a study of 820 patients with a minimum of 3 years of follow-up. Int J Radiat Oncol Biol Phys 2011;80:735e741. [21] Merrick GS, Butler WM, Wallner KE, et al. Prostate-specific antigen spikes after permanent prostate brachytherapy. Int J Radiat Oncol Biol Phys 2002;54:450e456. [22] Roach M, Hanks G, Thames H, et al. Defining biochemical failure following radiotherapy with or without hormonal therapy in men with clinically localized prostate cancer: recommendations of the RTOGASTRO Phoenix Consensus Conference. Int J Radiat Oncol Biol Phys 2006;65:965e974. [23] M
adan V, Aungurenci A, R
adulescu A, et al. Predictive factors for PSA bounce in 125I brachytherapy for prostate cancerdA literature review. Revista Rom^an
a de Urologie nr 2014;13:16. [24] Mazeron R, Bajard A, Montbarbon X, et al. Permanent 125I-seed prostate brachytherapy: Early prostate specific antigen value as a predictor of PSA bounce occurrence. Radiat Oncol 2012;7:46. [25] Crook J, Gillan C, Yeung I, et al. PSA kinetics and PSA bounce following permanent seed prostate brachytherapy. Int J Radiat Oncol Biol Phys 2007;69:426e433. [26] Kuban DA, Levy LB, Potters L, et al. Comparison of biochemical failure definitions for permanent prostate brachytherapy. Int J Radiat Oncol Biol Phys 2006;65:1487e1493. [27] Mitchell DM, Swindell R, Elliott T, et al. Analysis of prostatespecific antigen bounce after I 125 permanent seed implant for localised prostate cancer. Radiother Oncol 2008;88:102e107. [28] Zwahlen DR, Smith R, Andrianopoulos N, et al. Prostate-specific antigen bounce after permanent iodine-125 prostate brachytherapydan Australian analysis. Int J Radiat Oncol Biol Phys 2011;79:179e187. [29] Giton F, Fiet J, Cornu JN, et al. Serum sex steroids measured in middle-aged European and African-Caribbean men using gas chromatography-mass spectrometry. Eur J Endocrinol 2011;165: 917e924. [30] Patel C, Elshaikh MA, Angermeier K, et al. PSA bounce predicts early success in patients with permanent iodine-125 prostate implant. Urology 2004;63:110e113. [31] Bernstein MB, Ohri N, Hodge JW, et al. Prostate-specific antigen bounce predicts for a favorable prognosis following brachytherapy: a meta-analysis. J Contemp Brachytherapy 2013;5:210.