- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
Androgen deprivation therapy (ADT) causes prostate cancer cells to undergo apoptosis and results in a clinical response in most patients. About 80% of patients with metastatic disease respond to hormone therapy, and the median duration of response is 2–3 years. In addition to its role in managing recurrent and metastatic prostate cancer, ADT has become standard adjuvant therapy for high-risk prostate cancer including those patients with locally advanced cT3/T4 tumours, high-grade tumours (Gleason score 8–10) or positive lymph nodes [1,2]. In randomized trials, 2–3 years of adjuvant LHRH therapy improves survival by 16% at 5 years in high-risk patients compared to those not receiving adjuvant treatment .
Prolonged and continuous ADT produces a constellation of symptoms that can compromise the patient's quality of life. This ADT syndrome is well documented and includes vasomotor side-effects, increased weight and cholesterol levels; and reduced sexual function, energy levels, and haemoglobin levels . Long-term effects of ADT include maintained loss of sexual function, reduced muscle mass and decreased bone mineral density (BMD) [4,5]. Because of the high incidence of prostate cancer, earlier use of ADT and prolonged duration of therapy, the preponderance of ADT syndrome has increased and cannot be ignored.
Intermittent androgen suppression therapy (IAS) is a cyclic therapy consisting of active treatment periods followed by observation periods. During the off-treatment interval, patients are closely monitored for disease control, and evidence of progression triggers initiation of the next active treatment cycle. IAS was introduced in an attempt to minimize the side-effects associated with hormone therapy and to delay the development of hormone resistance. Akakura et al. reported a significant delay in the development of androgen resistance using IAS compared to continuous ADT in the Shionogi carcinoma animal model.
The first clinical data on IAS were published in 1995; that study reported improvement in ADT-related symptoms and quality of life during the observation (off-treatment) periods . Subsequent studies confirmed the clinical feasibility of IAS  and, in some cases, there were reduced side-effects and improvements in quality of life during off-treatment intervals . It is not known if IAS will delay the onset of androgen independence in humans, and this is the subject of ongoing phase III trials conducted by the several organisations. There is also very little information documenting the impact of IAS on long-term side-effects.
In 1993, we initiated a phase II trial of IAS in patients with prostate cancer treated at the Ottawa Regional Cancer Centre. In a preliminary paper we characterized the duration of off-treatment intervals of the first 54 patients and the clinical factors predicting the duration of these off-treatment intervals . This report updates the experience in 95 patients and highlights the short- and long-term side-effects of IAS in that population.
PATIENTS AND METHODS
- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
This phase II single-arm study was approved by the Research Ethics Board of the Ottawa Hospital, and all patients gave written informed consent for trial participation. Patients were required to have biochemically recurrent disease, biopsy-confirmed local recurrence, minimal metastatic disease after radical radiotherapy, or newly diagnosed minimal metastatic disease. Patients with metastatic disease had to be asymptomatic with involvement of lymph nodes only, lung metastases only or fewer than six bone lesions on a bone scan. The minimum PSA level for entry into the trial was 10 ng/mL. Patients may have received previous ADT for ≤ 12 weeks as adjuvant or neoadjuvant management of the initial primary.
The IAS consisted of on-treatment intervals of 8 months followed by off-treatment intervals of variable duration, depending on PSA level. During on-treatment intervals, patients received total androgen blockade (TAB) using the LHRH analogue leuprolide acetate (7.5 mg i.m. monthly) and nilutamide, initiated at a dose of 100 mg orally three times daily (total daily dose of 300 mg) for 2 weeks before and 2 weeks after the first LHRH analogue injection. The nilutamide dose was then reduced to 150 mg daily for the remainder of the 8 months. To be eligible for an off-treatment interval at the end of the 8 months of treatment, the PSA level had to be < 4 ng/mL and not rising, with no evidence of clinical or radiological tumour progression. During the off-treatment intervals, patients had monthly PSA and testosterone level evaluations to monitor biochemical control. Treatment was reinitiated when the PSA level was > 10 ng/mL and/or in the presence of clinically or radiographically documented progression. A cycle is defined as one completed on- and off-treatment interval. Cycles were repeated until the treatment failed to achieve ‘normal’ PSA levels, or there was rapid progression (<2 months) during the off-treatment period; patients were then usually prescribed continuous TAB.
The patients were followed clinically every 4 months with a symptom assessment and physical examination, including a DRE. Patients were monitored monthly with a complete blood count, liver function tests (aspartate aminotransferase, alanine transferase, total bilirubin), alkaline phosphatase level, PSA level, and total testosterone level during on-treatment periods, and a complete blood count, PSA and testosterone level during off-treatment periods. In the second half of the study, electrolytes and creatinine levels were also evaluated monthly because two patients had developed renal failure during the off-treatment intervals, probably as a result of disease progression. Patients were evaluated with chest radiographs and bone scans at the beginning and end of on-treatment intervals, or for suspected clinical progression. Patients with nodal disease were followed with CT of the involved regions. Acute toxicity was coded using the National Cancer Institute of Canada Clinical Trials Group toxicity scale.
In 1999 we became aware of potential effects of ADT on bone mass [4,5]; therefore, in 2000, annual BMD measurements (using a scanner) were incorporated into the study to evaluate changes in bone density during IAS. BMD was assessed in the lumbar spine (L2–L4), left femoral neck, total hip and left ultra-distal radius. The T-score, which expresses the spread of the observed BMD from the expected value in the young adult population in sd units, is reported. Osteopenia is defined as a T-score below the mean for a young adult by − 1 to < − 2.5 sd, and osteoporosis is defined as ≥ −2.5 sd below the young adult mean. The Z score, which expresses the spread of the observed BMD from the expected value in an age-matched population in sd units, was averaged to estimate whether, overall, the BMD in this population was different from their age-matched counterparts.
Body weight was measured at baseline and followed at each clinic visit. The body mass index (BMI) throughout the study period was calculated from the patient baseline height and weight at each visit, and percentage changes in BMI from baseline were derived. Sexual function (potency sufficient for intercourse) was documented at baseline, and patients with adequate sexual function were re-evaluated at the end of each on- and off-treatment interval to document loss and recovery of sexual function.
The presence of anaemia during a treatment period, and its recovery during the following off-treatment interval, was assessed. Anaemia was defined as a haemoglobin level of < 130 g/L. On-treatment intervals were evaluable for the presence of anaemia only when at least one evaluation was documented ≥ 3 months after the start of the cycle or one evaluation of ≤ 130 g/L was documented. Similarly, off-treatment periods were evaluable for recovery from anaemia only when at least one evaluation was documented ≥ 3 months after the start of that period or one evaluation of ≥ 130 g/L was documented.
Total serum testosterone level recovery during the off-treatment interval was assessed. Similar criteria were imposed for establishing failure to recover as those for recovery from anaemia. The time to normalization (values ≥ 10 nmol/L) was measured from the last treatment day to the first measure of total serum testosterone level of ≥ 10 nmol/L. Cross-tabulation (Pearson chi-square) between testosterone normalization during the off-treatment interval and recovery from anaemia and from loss of sexual function was used to evaluate the relationship between testosterone suppression and these androgen-deprivation symptoms. Treatment failure was defined as withdrawal from the protocol for any reason. The median time to treatment failure and to testosterone level normalization were estimated using the Kaplan-Meier product limit estimate.
- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
Between November 1993 and April 2000, 95 patients were enrolled (median age at diagnosis 70 years, range 50–82). Clinical information on the patients at initial diagnosis, including PSA level, Gleason score and clinical tumour stage, is listed in Table 1; the median PSA level was 27 ng/mL and the median Gleason score 6, with most patients having clinical stage II disease at initial diagnosis. Seventy-three patients (77%) received radical radiotherapy as initial management, 35 (48%) of whom also received neoadjuvant hormone therapy. In patients treated with radiotherapy, the planning target volume-1 and -2 consisted of the prostate with a margin of 2 cm and 1 cm, respectively. Radiotherapy was delivered on a linear accelerator using 18 MV photons with a four-field technique. Patients treated with radiotherapy after 1995 were treated with a three-dimensional conformal technique. The total radiotherapy dose was 45–69 Gy (mean 65 Gy, median 66 Gy). Adjuvant radiotherapy after radical prostatectomy was given in two (pT3) patients (60–65 Gy). Four patients received radiotherapy consisting of 60–66 Gy for biochemical relapse after prostatectomy. For the 73 patients treated with radiotherapy as initial therapy, the mean (range) time from initial therapy to study entry was 41 (3–126) months. Sixteen patients received no treatment before entry on the IAS protocol, 15 presented with minimal metastatic disease, and one had refused treatment until bone metastases developed. Details on the disease status at study entry are also given in Table 1. Most patients (63%) had evidence of distant disease, 15 (16%) had disease confined to the prostate and 20 (21%) only had documented PSA level progression with no clinical evidence of local or systemic disease. The median pretreatment PSA level was 19.8 ng/mL; six had pretreatment PSA levels below the requirements, four of whom had levels of > 9 ng/mL.
Table 1. The clinical presentation at diagnosis and disease status before starting treatment
|Variable||N (%) of patients|
|Pre-treatment PSA, ng/mL|
| 0–5|| 4 (5)|
| >5–10|| 8 (10)|
| >10–20||22 (28)|
| >20–50||20 (25)|
| >50–100||16 (20)|
| >100|| 9 (11)|
| 2–4||12 (16)|
| 5–6||32 (42)|
| 7||18 (23)|
| 8–10||15 (19)|
| I|| 5 (5)|
| II||36 (38)|
| III||26 (27)|
| IV||28 (29)|
|Disease status at study entry:|
|Type of recurrence/disease|
| Biochemical only||20 (21)|
| Local||15 (16)|
| Regional/distant recurrence||45 (47)|
| regional nodes|| 5|
| lung|| 2|
| others|| 3|
| Initial metastatic disease||15 (16)|
|Pre-treatment PSA level, ng/mL|
| <10|| 6 (7)|
| >10–20||40 (44)|
| >20–50||30 (33)|
| >50||14 (16)|
At the time of analysis, 245 cycles were delivered and 55 patients (58%) had discontinued treatment for the following reasons: 46 (84%) had evidence of progression while receiving hormonal ablation therapy (42) or shortly after the end of an on-treatment interval (four), two (4%) discontinued treatment because of treatment-related toxicity, and seven (12%) for reasons unrelated to the study protocol. The mean (range) number of completed cycles amongst these patients was 2.2 (1–6). Amongst those still receiving treatment, the mean number of completed cycles was 3.1 (1–7). Study progress is shown in Table 2; 78 patients (82%) completed the first 8 months on treatment and were eligible for the off-treatment interval. Amongst the remaining 17 patients, treatment was discontinued during the first cycle because of poor treatment tolerance (two); biochemical (PSA) progression alone (11) or with radiologically documented progression (three); and an unrelated condition (one). Another two discontinued for unrelated conditions during the off-treatment intervals while five remain progression-free on their first off-treatment interval. Therefore, 19 of the 95 patients could not complete cycle 1; this group represented 25% of patients with disseminated disease (newly diagnosed or recurrent) and 11% of patients with local or biochemical disease only.
Table 2. Number of patients at each stage of the cycle and the changes in haemoglobin during treatment
|Stage or anaemia status||Cycle|
|Starting that cycle||95||71||47||19||8||4||1|
|Progressing on OTI||14||12||12|| 6||1||1||–|
|remaining on OTI|| 0|| 3|| 9|| 3||3||1||–|
|Remaining on FTI|| 5|| 7|| 6|| 1||–||1||1|
|Removed for other reason|| 5|| 2|| 1|| 1||–||–||–|
|Completing cycle||71||47||19|| 8||4||1||–|
|Median (range) FTI, months|| 9.0 (1.7–50.6)|| 8.1 (3.2–16.0)|| 9.3 (3.5–31.4)|| 6.1 (2.5–16.7)||6.2 (4.1–9.1)||n/a||n/a|
|Changes in haemoglobin|
|Anaemia, n with anaemia/N evaluable (%) [total]:|
|during treatment||15/45 (33)||11/25 (44)||12/18 (67)|| 7/11||3/5||2/3||1/1 [51/108 (47)]|
|at baseline for each cycle|| 3/69 (4)|| 6/26 (23)|| 5/14|| 2/10||3/6||1/2||1/1 [21/128 (16)]|
|during treatment in patients with normal baseline|| 9/35 (26)|| 2/13|| 4/5|| 3/7||1/3||0/1||0 [19/64 (30)]|
The median duration of cycle 1 off-treatment interval was 9.0 months; 71 patients (75%) initiated cycle 2. The overall median duration of off-treatment intervals for all cycles was 8.5 (6–9) months (Table 2), although some patients maintained disease control for prolonged periods. Off-treatment intervals lasted > 1 year in 40 cases (27% of the 150 completed off-treatment cycles).
The median (range) follow-up was 44 (9–92) months; the median to time to treatment failure was 47 months for the entire group, 65 months for patients with local or biochemical recurrence only and 40 months for patients with regional nodes or metastatic disease. The estimated proportion of patients having no treatment failure at 3 and 5 years for the entire group was 62% and 48%, respectively.
The most common side-effects (any grade) occurring during on-treatment intervals included vasomotor symptoms (71%), fatigue (42%), impaired light-dark accommodation (23%) and nausea (12%). Grade 3 toxicity consisted of pneumonitis and hepatitis secondary to nilutamide (one case each) and one case of repeat episodes of syncope. Grade 2 toxicity (other than vasomotor flushing) included: insomnia, impaired light-dark accommodation, anorexia/nausea (one patient each). There were no grade 4 toxicities.
Two patients developed acute obstructive renal failure during an off-treatment interval, caused by locoregional progression. Although this is not directly related to the treatment, it is a consequence of withdrawing therapy. This observation prompted a protocol amendment requiring serum electrolyte and creatinine level evaluation in all patients during off-treatment intervals.
Testosterone recovery during off-treatment intervals could be evaluated in 117 cycles; amongst these, full recovery was documented in 71 (61%). The Kaplan-Meier estimate of the median time to normalization for the 117 cycles was 33 weeks. Amongst cycles with testosterone recovery, the median time to normalization was 23 (4–61) weeks. For patients who did not achieve normal serum testosterone levels, the median follow-up was 37 (4–120) weeks.
Protocol compliance for haemoglobin evaluation was poor; only 65 patients had sufficient data to evaluate their haemoglobin status during the hormonal therapy for at least one cycle. In all, 108 cycles were evaluable, and amongst these anaemia was recorded in 51 (47%) (Table 2). There was an apparent rise in the proportion of cycles with anaemia amongst patients receiving more cycles. During cycle 1, a third of patients were documented to have anaemia, whereas during cycles 2 and 3 that rose to 44% and 67%, respectively. There were too few evaluable patients in subsequent cycles to comment. The degree of anaemia was generally mild. Amongst the 51 cycles with documented anaemia, only 22 (20%) and eight (7%) were associated with haemoglobin levels of <120 and 110 g/L, respectively. When only cycles with normal pretreatment haemoglobin levels are considered (n = 64), the overall risk of developing anaemia during the cycle was 30 (0–80)%, with no apparent relationship between the cycle number and risk level. This suggests that the increasing proportion of cycles with documented anaemia with more cycles is, at least in part, a result of persistent anaemia.
In 26 cycles, anaemia was documented during the on-treatment interval, and sufficient information was available to evaluate recovery during the off-treatment period. Amongst these cycles, only 13 (50%) showed recovery after stopping treatment.
We evaluated the proportion of patients with documented anaemia at the start of each cycle. Anaemia was rarely present before the first cycle (4%), whereas in subsequent cycles the percentage of patients with documented anaemia before initiating hormonal therapy increased, with values of 20–50%, supporting the notion that recovery from the previous cycle was not complete.
Haemoglobin levels and testosterone recovery during the off-treatment interval were both documented in 48 cycles. Anaemia was more frequent in off-treatment intervals during which serum testosterone levels did not normalize (47% vs 14%; P = 0.01).
Sexual potency before the start of therapy was reported in 65 patients; 44 (68%) reported being impotent before initiating the study treatment. The 21 patients who were sexually potent received 57 cycles. Patients universally lost potency during on-treatment intervals. In nine cycles, insufficient information was available to determine whether sexual recovery occurred during the off-treatment period. In another 10 cycles, a determination could not be made because the patients had not completed the off-treatment interval at the time of analysis, and were therefore not re-evaluated. Amongst the 38 evaluable cycles, patients regained potency in 18 (47%). When sexual potency was regained in the first cycle, the patient's sexual function tended to recover in subsequent off-treatment intervals. Information on testosterone normalization was available for 23 of the 38 cycles evaluable for the recovery of sexual function. Men in whom testosterone levels had normalized were somewhat more likely to also have recovery of sexual function (64% vs 50%), but this was not statistically significant (P = 0.51)
The change in BMI was compared from baseline over the course of the study period; in all, 1103 weights were recorded in 89 patients. The largest percentage changes in BMI from baseline during the entire study period in each patient are shown in Table 3; most had no significant change in BMI. Only five patients had a decrease in their BMI of > 10%; increases in BMI were more notable, as in 13 patients (14%), weight gain during the study period caused a change in BMI of > 10%. However, there was no consistent trend in BMI changes during the entire study period. The mean and median overall weight change from the first to the last evaluation was − 0.6 and − 0.1 kg, respectively, suggesting that weight gain which is normally associated with ADT is offset by weight loss during the off-treatment interval.
Table 3. Maximum changes in BMI per patient during the study period
|% change in BMI||N (%) patients|
| 0–5||67 (75)||61 (69)|
| 6–10||17 (19)||15 (17)|
|11–15|| 5 (6)|| 9 (10)|
|16–21|| 0|| 4 (4)|
BMD was assessed in 41 patients (59 evaluations); as all patients had already been enrolled in the study at the time the BMD protocol was amended, we did not obtain baseline measurements. BMD was evaluated at a mean of 43 (8–90) months after the start of therapy, revealing osteopenia and osteoporosis in a significant proportion of patients. Table 4 shows the classification of lowest T-score per patient recorded during the study period for the lumbar spine, left femoral neck, total hip, and ultra-distal radius. Overall, 15 (37%) of the 41 patients evaluated for BMD had documented osteoporosis in at least one site, with osteoporosis documented in 8–31% of each of the four sites evaluated. Most patients in the study received previous pelvic radiotherapy, which in all cases included the hips. However, the hip had the lowest prevalence of osteoporosis (8%). The highest prevalence of osteoporosis was in the ultra-distal radius (31%).
Table 4. BMD status, derived from the lowest T-score for each patient
| ||N evaluable||Osteopenia (−1.0–< −2.5)||Osteoporosis (−2.5+)|
|Lumbar spine||41||12 (29)|| 6 (15)|
|Left femoral neck||41||16 (39)|| 4 (10)|
|Total hip||40|| 9 (23)|| 3 (8)|
|Ultra-distal radius||36|| 8 (22)||11 (31)|
Eighteen patients had two BMD assessments ≈ 1 year apart (mean 1.1, range 0.9–1.5 years); there was a mean decline in the BMD of 0.36, 0.13, 0.09 and 0.42 sd for the spine, femur, hip and radius measurements, respectively. However, when the lowest Z scores (which measure the deviation from age-matched controls) were averaged across all patients, there was no directional shift in the mean deviation. Mean sds for the Z scores were + 0.30, − 0.44, + 0.71 and −0.37 for spine, femur, hip and radius measurements, respectively.
- Top of page
- PATIENTS AND METHODS
- CONFLICT OF INTEREST
Hormone ablation is the standard therapy for metastatic and recurrent prostate cancer. In addition, long-term adjuvant hormone therapy has become the standard of care for patients with high-risk prostate cancer . The median survival for patients with hormone-refractory disease is 9 months, and there is currently no effective therapy to prolong survival in hormone-refractory disease.
Intermittent hormone therapy has the potential to delay the onset of androgen independence. In animal models of IAS, tumour cells surviving androgen withdrawal are forced into a normal pathway of differentiation by androgen replacement; therefore, the apoptotic potential is restored and androgen independence is delayed. During tumour progression there is a marked increase in the proportion of stem cells and a 500-fold increase in the proportion of androgen-independent stem cells. Laboratory studies using the Shionogi mouse mammary tumour show that androgen withdrawal alters the ratio of stem cells in the tumour population, suggesting better tumour control. IAS was reported to significantly delay the development of androgen independence compared to continuous hormone therapy in the Shionogi carcinoma . Randomized clinical trials are underway to determine whether IAS can delay the hormone-refractory state in patients with prostate cancer.
Androgen blockade therapy can significantly compromise the patient's quality of life. Because of the long natural history of prostate cancer and widespread use of long-term hormone therapy, attention to side-effects and quality of life become paramount. The side-effects of prolonged ADT include diminished sexual function, energy levels, muscle mass, and haemoglobin level, as well as increased weight and cholesterol levels. Long-term use of ADT is associated with reduced BMD [3–5]. Several of these symptoms are directly attributable to suppressed androgen levels. IAS was first introduced clinically in the 1980s to help to minimize the side-effects of antiandrogen therapy . Phase II studies in prostate cancer show the clinical feasibility of IAS; 60–92% of patients were eligible for off-treatment intervals and could therefore potentially benefit from a break in therapy [9,11–13]. Studies report improved quality of life during off-treatment intervals, with an improved sense of well-being and recovery of sexual function in most patients . Serum testosterone levels return to normal after withdrawing antiandrogen therapy in patients treated with neoadjuvant therapy . The time to normalization varies among studies and is reportedly higher in patients with longer treatment periods .
Androgen blockade and surgical castration are associated with an apparently accelerated loss of BMD, resulting in osteopenia, osteoporosis and osteoporotic fractures. Studies show that patients with prostate cancer who have had orchidectomy were significantly more likely to develop an osteoporotic fracture than those not castrated , and that the use and duration of ADT are significantly associated with the loss of BMD in men with prostate cancer [4,5]. In the present study, 37% of patients had osteoporosis in at least one of the four sites evaluated, and there was a loss in BMD over time in some patients. These estimates were obtained from evaluations at a mean of 3.6 years after the start of therapy. However, the median age of the present patients was 70 years, and we expected some deficiency in BMD compared to the normal population. Age-matched comparisons showed no obvious deficiencies. There were too few patients to make inferences about the contribution of ADT or age to the change in bone mineral loss. Interestingly, the incidence of osteoporosis was lower in the total hip region than in the lumbar spine, despite 76% of patients receiving hip radiotherapy in the original radiation portals. The association between IAS and bone loss is still unknown. Earlier studies of IAS did not report changes in BMD; in one study, seven of 17 patients in whom BMD was measured had levels lower than normal for their age, and four of the 35 patients in the study had osteoporotic fractures . Effective strategies for managing osteoporosis include dietary change, exercise and bisphosphonate therapy. Recent studies in postmenopausal women showed the efficacy of oral bisphosphonates in improving bone mass and reducing the risk of osteoporotic vertebral fractures . There were similar improvements in BMD using bisphosphonates in osteoporosis of varying causes. A similar efficacy can be anticipated in men with osteoporosis related to ADT. In a primary prevention study by Smith et al., i.v. pamidronate reduced bone loss secondary to ADT in patients with prostate cancer. We are currently conducting a phase II study using the oral bisphosphonate risedronate in the treatment of established osteoporosis secondary to ADT.
Lack of serum testosterone in men has been associated with increases in body fat and loss of lean body mass . In the present patients, there were fluctuations in body weight and corresponding changes in BMI during the study in both directions. However, a > 10% change in BMI was associated with weight loss in 5% of patients and weight gain in 14% of patients, but overall there was no net weight gain at the end of the study period. An evaluation of the percentage changes in body fat and muscle mass during on- and off-treatment intervals would have been a more appropriate measure of the impact of IAS on body composition. We are currently evaluating the effects of ADT on lean body mass, and assessing the role of progressive resistance exercise training on lean body mass, muscle strength, patient weight and quality of life in a rehabilitation programme.
Amongst patients who were potent and sexually active before starting therapy, sexual potency was regained in 47% during off-treatment intervals. Patients who reported the recovery of sexual function during the first off-treatment interval usually maintained their potential to recover in subsequent off-treatment intervals. There was no significant association between sexual recovery and testosterone recovery. Two other studies found that patients who reported normal or near-normal sexual function before the start of the therapy recovered potency and libido during off-treatment intervals .
Anaemia has been previously reported as a common side-effect of ADT. The anaemia is normochromic, normocytic, temporally related to the initiation of continuous TAB, and usually resolves after therapy is discontinued . In the present study patients there was anaemia (a haemoglobin level of < 130 g/L) during on-treatment intervals in 47% of cycles. Only seven cycles were associated with haemoglobin levels of < 110 g/L. These results suggest that ≈ 30% of patients with normal baseline haemoglobin levels will develop anaemia during the on-treatment interval, and that the overall number of cycles with anaemia increases with more cycles. We estimated that half of patients recover from anaemia during the off-treatment interval, and postulate that incomplete recovery from previous anaemia results in a higher prevalence of anaemia in subsequent cycles. The improvement in haemoglobin during off-treatment intervals probably contributes to improvements in the sense of well-being and energy in these patients. In the present cohort, recovery from androgen suppression during off-treatment intervals was significantly associated with normal haemoglobin levels.
Two patients had tumour progression in off-treatment intervals that lead to a progressive deterioration in renal function. One patient developed local prostate progression resulting in obstructive uropathy, and the second developed bilateral hydronephrosis secondary to progressive nodal disease. It is unknown if these locoregional complications would have occurred if the patients had been on continuous ADT; however, based on our experience we recommend follow-up during off-treatment intervals that includes an assessment of serum creatinine levels to detect impaired kidney function.
As patients are off active therapy for a comparable duration to the on-treatment intervals, IAS has the advantage of a lower cost than continuous medical castration. The present median duration of off-treatment intervals was 8.5 months. The likelihood of progression during the first cycle appeared higher for patients with metastatic disease (25%) than in those with local or biochemical recurrences only (11%). Overall, the treatment failed 4 years from the start of therapy. Because of the diversity of patient groups treated in phase II trials, their small sample size, and non-randomized setting, it is not possible to determine if IAS can delay the development of androgen independence in the clinical setting. These early-phase studies have been the impetus for ongoing randomized controlled trials evaluating IAS in patients with biochemically recurrent prostate cancer and for patients with metastatic disease.
Strum et al. reported on factors associated with prolonged off-treatment intervals. The presence of biochemical disease only, undetectable PSA levels (<0.5 ng/mL) on-treatment, and slow testosterone recovery during off-treatment intervals, were predictors of a better outcome. We are currently evaluating factors predicting disease control in the present cohort and will report on this later.
In conclusion, data from the present phase II study show the feasibility of IAS and the potential for reducing side-effects, including recovery of haemoglobin levels, return of sexual function and absence of weight gain at the end of the study period. Our study also highlights the need for clinicians to be aware of acute and long-term side-effects of hormone therapy, including potentially treatable side-effects such as osteoporosis. We await the mature results of phase III trials of IAS to determine if it delays the onset of androgen independence and improves survival in recurrent and metastatic prostate cancer. Should IAS be shown to be equivalent to continuous hormone ablation in survival and time to an androgen-independent state, IAS may still be a useful therapy if evidence of reduced duration of side-effects, improvement of quality of life and cost reductions is confirmed.